/ohm/ - electronics general

>>1138806

last thread:
>>1133263

link to this thread in the last thread before it 404s

I get the feeling I should bump this thread. Anyone here have experience making or modding RTL-SDR dongles?

I'll start with saying I have no experience with any electronics project other than wiring a laser diode to a switch and some batteries.
I have an art project I'd like to get started on that requires a 1"x1" screen that can display images stored on a micro sd and cycled through with the press of a button. There wouldn't be much room behind the screen either, maybe allowing for 3/4" behind the screen.
I have no clue where I could even get the components for something like this. I'm sure I could figure out how to put it all together quickly enough. Any suggestions?
Oh and as far as a power source..it would have to be small and fit behind the screen as well, maybe something rechargable via usb.

>>1138886
I'd probably go with a Raspberry Pi 0, $5. There also might be a Blueberry Pi at this range too, I'm not sure. I don't know what else you can get that displays an image, especially from a microSD card. The device has a microSD slot, but it is used for the OS (linux of some sort) and all other memory too, so if you want to swap the cards, you'll need to copy your configured OS onto every card before putting on the photos. Second problem is configuring the OS to use with a 1" screen. Unless you get one with a miniHDMI port on it, you'll need to get drivers to run it of the GPIO headers. And installing drivers on Linux is a lot harder than on a Windows machine. For the button press mechanic, I'd go full budget and dumpster dive for a broken USB keyboard, remove the chip and all the electronics, and attach a mechanical button to the right arrow key pad. With a little work and maybe some rebuilding, you should get it to work.

https://www.adafruit.com/products/2088 looks like a pretty good choice, not many colours, but it has an extra memory card slot and is almost exactly 1"x1". Adafruit also should have instructions and downloads for the drivers. Touch screens are much harder though, I tried to make a portable SDR all-band radio with a dongle, a pi, and a touchscreen, but getting the touchscreen to work was beyond me.

>>1138886
do you want the viewing area to be 1x1 or the maximum product dimensions to be 1x1?

here's a screen:
http://www.hotmcu.com/index.php?main_page=index&cPath=6_16
and a board you can fit behind it:
https://www.arduino.cc/en/Main/ArduinoBoardMini
and an sd card breakout:
https://www.adafruit.com/products/254?gclid=CK_HpYqKudICFcRAhgodUtQLGw

given your constraints you'd be better off getting some dumb EE at your school to make a pcb for you.

i'd surrender the notion of an onboard battery charger if you're just doing this as a short term presentation. just get a baby lipo cell and a separate charge board that you can clip on with hook leads when it needs to be charged.

>>1138904
I want the viewing area to be 1"x1" roughly. Could be a touch smaller or bigger. Also it could be 1"x3/4" or something like that. The bezel around the screen can't be too big either, so I can't have a tiny screen on a huge board.
I guess an easy way to picture it would be...imagine I'm making a very small scale model television set. It doesn't have to be able to display animations (although that would be pretty neat), just static images. The television set visual is just to give you an idea of the dimensions I need to work with.

>>1138908
you'll be hard pressed to find a 1x1-.75 screen on a breakout board close to 1x1. the ones that anon and i linked are representative of the minimum space that bezel and board will consume in addition to the viewing area. if that's fine for you then you're golden.

also it seems like my screen link is broken. you'd want the top result on the list it gives. note that i've bought two screens from there before and one was DOA so buyer beware.

File: adfhdajfgjfsgjhfs.jpg (33KB, 618x239px) Image search: [Google]

33KB, 618x239px

>>1138911
Thanks man, I really appreciate the help you're giving me here. I'm clueless when it comes to this stuff but starting to get really interested in it.
I made this in paint to give a sort of feel for what I'm looking for I guess. What I sort of had pictured in my head that it would be. I forgot to add in visual queues that the usb and microsd slot would be open on the back side though.
I might just have to rework my project a bit with the stuff you've provided links for. That seems pretty doable. I can't thank you enough for the help.

File: Screenshot 2016-11-16 01.17.33.png (101KB, 1920x1080px) Image search: [Google]

101KB, 1920x1080px

what do you think of my solution to a problem i had recently.

I'm rebuilding a tube stereo i made many years ago.
when rebuilding the controlcard i decided to ditch µC and go Cmos for shits and gigs.

Problem:
how do I cycle channels(1-4) with a rotary switch while being able to cycle channels with a remote?

solution:
rotary switch, 2 bit binary counter and 4x 4way muxes.

each step of the rotary switch goes to one of the muxes common input.

the muxes are adressed by the binary counter.
the counter counts pulses from the remote receiver.

the first mux is connected output A to relay driver 1, B to 2, C to 3 and D to 4.
the next mux is advanced by one A to 2, B to 3, C to 4 and D to 1.
the following mux is A to 3, B to 4.... you get the idea.

meaning pulsing the counter advances channel by 1, and allowing the rotary switch to select channels normally.


this was the best and easiest solution i could think of. with the stuff i had available.
might be a little over-engineered, is there a smarter solution that doesn't involve rotary encoders and discriminators and such?

its built now, so I won't change it, but I'm curious.

File: Screen Shot 2017-03-03 at 12.51.37 pm.png (164KB, 483x467px) Image search: [Google]

164KB, 483x467px

>>1138919
If you want a screen with a premade board, this is probably the best your gonna get in regards to board size/screen size.
That said it is a 1.4" screen.
The smaller 1" screens around have bigger breakouts.
If you do want it flush you'd have to make your own board, which isn't exactly hard if you know what you're doing, and then everything could be hidden behind the screen.

File: Screen Shot 2017-03-03 at 12.54.24 pm.png (200KB, 392x365px) Image search: [Google]

200KB, 392x365px

>>1138919
Also thi, oh. You want a 1x1" screen, not a 1" diagonal screen? Then the last one I replied with is what you want.
This is the only 1" color diagonal screen with a breakout around.

>>1138928
Seems pretty simple to me anon, not sure there'd be a much cleaner way.

File: ZX-HOVER-MTR-WHL-lg.jpg (133KB, 550x450px) Image search: [Google]

133KB, 550x450px

I'm trying to build a small gokart powered by four "Hoverboard" (Self balancing scooter) hub wheels, and am trying to decide the best way to run them.
I got a broken hoverboard to scavenge parts from, so I thought I'd have a play around before going and buying more wheels.
They're just three phase hub motors, but unlike most I've seen, where feedback is given on the unused wire, in this case it has a separate five pin cable I'm guessing are connected to hall effect sensors.
Before I just start taking readings from the things and doing everything from scratch, I thought I should ask around to see what other anons suggest.
If I was really lazy I could just use some high amp hobby ESC's, but I thought I'd build my own triple half h-bridges to run each wheel.
Has anyone else played around with making their own ESC's?

>>1138960

They are almost certainly hall sensors. Almost any ebike/scooter controller would work fine.


>but I thought I'd build my own triple half h-bridges to run each wheel. Has anyone else played around with making their own ESC's?

Honestly, it wouldn't be much effort for me, personally, on top of already having everything needed lying around. But I still only MIGHT hassle with it only because off the odd "4 separate motors" requirement.

Which brings up the other issue of "is it even worth using these motors?" I can't imagine ripoff Hoverboard motors have much power, I'd guess in the neighborhood of 250W each. The cooling and power characteristics of hub motors are also garbo compared to other options.

If you're just doing this as a learning exercise, go for it, but if you actually want to use it now and again after you're done, you might want to consider using some else as your power plant.

>>1138985
I live in Australia, we have a 200w limit on non pedal vehicles, so I have, and am used to the power of, a 200w ebike.
The thought of four 250-350w motors hooked up to custom, non limited controllers makes me feel fuzzy inside.
Really, I just wanna see how fast I can get this thing. Even as a massive RC car.

>>1138997
>I live in Australia,
Stopped reading there.

>>1139000
No booley plz

>>1138655
Embedded doesn't seem to get paid as much for whatever reason. I suggest getting into low-level Android stuff.

>>1139000

Check'd.

>>1139001

Respec't.

>>1138997

>200W limit

I still can't figure out why the fuck ANY legislative body thinks power limits are a good idea, much less why almost ALL of them do. I was astounded when I found out California dropped the power limit in favor of a speed limit only. Why there are 3 classes, and why it's okay to go 27MPH with a motor set to 99% pedal assist but not okay to go 27MPH with a throttle makes no god damn sense (nevermind the seemingly-arbitrary 27MPH in the first place), but it's more than I was expecting.

Anyway, you wouldn't need any custom controller to do it. Standard Chinese fare would do, though you'd obviously need 4 of them. As cheap as they are, price might become a problem due to that issue. Might be worth making up a special one if you're already familiar with designing electronics.

my 315mhz reciever:
>works when the scope probe is inserted into the breadboard
>prefers the probe to be on the vcc, data, or gnd pins
>still works with more dropped packets when the probe is plugged into a nearby isolated row
>will transmit intermittently when i fondle the antenna or move the receiver close to it
>doesn't work even with the probe when i hold it under the table
i think i'm done with rf

>>1139011
I don't know where you're getting that. Me and the other embedded programmers I know are payed very handsomely. Just don't go work for a startup with some IOT crap. A lot of them run out of money before getting anything on store shelves.

It's also important to understand that embedded programming requires a high level of electronics competency or you're almost useless.

>>1139018
I've designed a few low voltage microcontroller-oriented things before, so I don't think this'll be too much of a challenge for me, learning experience blah blah. Also, I want total control and to be able to tune the thing to the absolute max speed, and I've found a lot of chinese controllers are pretty limited (pun not intended but I'll take credit for it).

File: 1458770793474.png (20KB, 510x619px) Image search: [Google]

20KB, 510x619px

>>1139053
give this high side nmos gate driver a shot, i don't have any excuse to make it

>>1139035
Well, that was my impression when looking at entry-level jobs a while back. Not my subfield though.

CS and applied math major here. embedded device programming fascinates me, is this still a viable career in the 21st century?
Will I have to immigrate to China?

>>1139136
Question answered @
>>1139011
>>1139035
Thank you.

Power supply fag from a few months ago.
I'm still curious, esp if EEs are about is there a smarter better way (whilst still using LM317) to adjust voltage in a linear fashion by driving a voltage into the adjust pin? op amps hav ~30V supply range, so basically anything above that is out of reach, isn't it?
its really bothering me

>>1139024
A probe is effectively a tiny cap + a big ass resistor to ground, try adding those to your circuit and see what it does.

>>1139149
An LM317 is just a fixed 1.25V regulator, it'll output whatever's on the ADJ pin + that 1.25V, assuming the input is at least 3V above desired output.

Recommended operating conditions for the LM317 say output shouldn't be over 37, so you're not really losing a lot of range if you can only put 30V onto VADJ

If you don't mind having a lower minimum output, you could use a +5V or +12V regulator instead of the +1.25V of the LM317, that'll get you a little extra at the top.


If you want to be able to drive a wider range than that, then you start having to design your own opamp + power transistor/mosfet arrangement to do the regulation.

What are some good beginner experiments?
I've been mucking around with silly shit, hooking up piezos to solar panels and playing them like a theremin to annoy my girlfriend, I'm looking for a project that doesn't blow out the scope but is a bit more to chew than something I can throw together in 30 minutes

>>1139158
nah can't do it with op amp because rail to rail 30V (plus a bit more to get a -1.25V Vs-)op amps are extremely uncommon or expensive.. usually both (i haven't been able to find one anyway)

I can 'design' a basic op amp transistor setup.. output of op amp connected to base / gate of the transistor. output is connected to negative feedback through resistor network.. surely its a bit more complicated than that though?

I've also wanted to make a switching preregulator that tracks voltage and ensures its always 3 - 5V above the output too.. so there's another obstacle

If i have a ground plane for my antenna do i have connect it to my circuit ground?

File: 20170303_143032.jpg (3MB, 5312x2988px) Image search: [Google]

3MB, 5312x2988px

What is this little thing on the end of some wires called? I need to buy some wires with these on them and don't know what they're called.

Not the rubbery plasticy insulation. I mean the thingy you slide onto another thingy to make a connection.

>>1139270
a crimp connector. alternatively you can just buy a bunch of those connectors and crimp the wires yourself. should be cheaper imho

>>1139271
I might do just that! Thank you for the quick response!

I'm designing my first pcb that I plan to etch and use. I notice that connections can be drawn on both the front and the back, which seems handy for avoiding a tangle of pathways; Do connections still work no matter what side they're made on?

>>1138928
Am on bus currently but I was thinking of just using a counter ic and preloading it, it's a little hard to explain but will draw something up when I get home

>>1139312
Yes, but to connect between two (or more) sides of a multi-sided board, you need what I assume to be some sort of conductive plug that fuses between the two sides. Useful for connecting the ground planes together. If you happen to not own 200um drill bits, then just solder a wire between them, or try soldering a spare lead on either side of a through-hole. And you should be careful with through-hole components to make sure they don't connect to the other side of the board if you don't want them to.

File: 20170228_200420-1.jpg (1MB, 3679x1359px) Image search: [Google]

1MB, 3679x1359px

it's a good thing the only thing I can do well is solder

File: resistometer.png (4KB, 640x400px) Image search: [Google]

4KB, 640x400px

Ok, so I want to build a simple little automatic "resistometer" that powers on when I put in a resistor and displays what the resistance is on a 7 seg display. I'll probably use an Arduino. My plan at the moment is a circuit like pic related. I plan on having it be able to read a resistance anywhere from 0 to 100M (just in case). The 50Ω resistor is to limit the power used by both the 50Ω resistor itself, (1/2W max running at ~1/4W at 0Ω test resistance/short circuit) and used by the test resistor (0.05W max). I'll be able to calibrate the imperfections in the 50Ω resistor by software, possibly through an external pot, and calculate the resistance through "R=50*V/(3.3-V)". I'm using the FET to turn on the electronics, and hopefully it doesn't matter how little current is going through it, from 66mA to 33nA, as long as there's a voltage across the circuit. It's better than using fifteen cascaded BJTs. Anyway, my question to you is can I get an ADC that can measure across such a long range of resistance values. Because as far as I see it, It's going to be precise for values similar to 50Ω, and vastly imprecise for huge and tiny resistances. Do I need to, and how would I make it auto ranging?

>>1139364
The low resistance side isn't bad, with an 8 bit ADC the lowest non-0 resistance comes out as ~0.2Ω, which I could probably live with as long as I'm not trying to measure current shunts, otherwise I'm going to have to decrease my voltage. But the highest non-infinite resistance comes out at 12,700Ω, which is pathetic. I could put the ADC across the 50R, but the effect would be exactly the same. Manual ranging would be fairly simple, use a rotary 2-pole switch that changes which resistor the signal goes to and also tells the microcontroller how to calculate for this resistor, but manual ranging takes away from the ease of use. The rotary switch might also have significant resistance, especially in the order of single ohms or less, which I will have to calibrate for. I hope the FET doesn't have a significant resistance, because if it does it will be difficult to calibrate for.

I have an idea for autoranging, that is whenever the current range shows a resistance of 0 or infinite ohms (0 or 3.3 volts respectively), the arduino will switch a series of FETs/BJTs, closing contact with the next resistor and opening it with the old one, but at the MΩ range I've no clue if the open-circuit transistors will have a finite and significant resistance, because this will also throw a spanner in the works and make more calibration issues. Am I going to have to use solenoid relays?

>>1139316
i thought of that, but the problem is that you need additional circuitry to only preload when there is a change in the rotary switch.

also the rotary switch would not necessarily add up with the channel you are on.
if the counter is on channel 4 and the switch is on channel to. going up would preload channel 3 and down would preload channel 1, so you it would switch to the wrong channel.

could be I'm missing something or misunderstood you.

File: 7SEG.png (10KB, 640x400px) Image search: [Google]

10KB, 640x400px

>>1139364
>File: resistometer.png

i kek'd

something about it reminds me of leggo batman

>>1139381
why complicate it that much?
why not have a current sense that enables the display?
like a base-emitter to ground in series with a shunt, having the collector pull 7 seg common cathode to ground?

getting a transistor that can handle the trivial currents of measuring resistance should be easy.

also, get real and drop the adc and µC

>>1139364
Typically you'd use a programmable current source instead of fixed series resistors. (Or a bunch of current generators with individual enables.)
Required ADC resolution obviously depends on the desired display resolution and the number of ranges. 24b converters are pretty cheap nowadays, but you aren't going to get 24 real bits out of them.

>>1139391
I just figured not running out of battery would be a good idea, especially if I was going to run this off a coin cell or something else tiny. And the current sensor has to be able to detect nanoamps and milliamps, if you can point me in the direction of one that would be great. I'm not sure how I'd convert the current/voltage read by whatever I should use into the resistance, I'd need either a specialised IC with room for an external calibration pot, or use a microcontroller and do the "R=50*V/(3.3-V)" calculation. Are you suggesting that I replace the FET with a current sense or replace the ADC with one, because I don't know how I'd calculate the resistance without one. If you happen to know a less complicated way of doing this, enlighten me.

>>1139393
That makes sense, would I be able to scan through a continuously variable range of currents searching for the most accurate resistance? Any other sort of auto ranging would also be good. And I hope these current sources can produce in the nA range, because otherwise the actual voltage across a 10MΩ would be significantly high.

>>1139396
darlington transistor, loads of gainz

>>1139398
How is that better than a FET in a voltage-driven circuit? I can understand using one with a current supply.

>>1139397
So the cheapest 24bit ADC for SMD is $3, while the cheapest through-hole is $30. I don't get it.

File: 20170303_214711.jpg (2MB, 3120x4160px) Image search: [Google]

2MB, 3120x4160px

>>1138806
Ohm I got a motor to replace my tablesaw motor, I have some issues tho. I got one that said it could run on a 120 circuit. but reading the manual, it says I need like fuckin #4 gauge wire?! from the breaker to the outlet, is about maybe 5 feet. its in the garage, but I'm pretty sure that gauge inside the house [120v 15amp] is MAYBE just 12 gauge, its pretty whimpy. then I need to run a heavy duty extension cord, maybe 30 feet..

pic related the motor.

File: 20170303_214720.jpg (4MB, 3120x4160px) Image search: [Google]

4MB, 3120x4160px

>>1139419
Pic related the manual page...

/ohm/ help me unfuck this, I don't really have the money to hire someone to install a breaker and outlet.. got a quote for a 220v install, and was told 600 dollars.. for literally like a max of 6 feet. to the left of the breaker.

>>1139419
so this motor is higher power than the motor it replaced?

File: 20170303_215932.jpg (4MB, 3120x4160px) Image search: [Google]

4MB, 3120x4160px

>>1139421
the motor it replaced was 1HP. I tried to keep the RPM the same as the OEM motor. Pic related. the plate.

but im pretty sure the motor was fried before I plugged it in at my house when I got the saw. so I never knew the motor to work correctly, when I got the saw home and tested it, it felt very weak, weaker then my skillsaw.. so pretty sure it was fried when I got it. hence why I needed a replacement.

>>1139419
>[120v 15amp] maybe 12 gauge

It also may be 14 gauge

The plate gives 14.2A a full load rating
>allow for 125%

Have someone put a 20A outlet where you need it.
#10 will be fine

>>1139444
>Have someone put a 20A outlet where you need it.
>#10 will be fine

Even better is a 230V outlet. - same hp from fewer amps

>>1139445
Even even better, get three phase installed.

>>1139404
Some bits are more 'real' than others.

>>1139444
but I need to get a fucking switch added to the panel, and run wiring, so its gonna cost me 600 dollars no matter fuckin wut? so I would just end up getting the 220 installed because its better in every way.

>>1139450
I still don't get it.

>>1139419
>from the breaker to the outlet, is about maybe 5 feet. its in the garage
>>1139419
>then I need to run a heavy duty extension cord, maybe 30 feet..
>>1139457
>but I need to get a fucking switch added to the panel, and run wiring

Just pull the outlet and check the wire going back to the panel.
If it's #12 it's good for 20A
Make up a 30' extension cord of #10 wire.
If you can't do any of this - hire a pro

>>1139460
ADC architectures are not made equal. For example, NAU7802KGI (which is $2.27 in DIP?) has ENOB of < 24b.

>>1139462
that nigger just told me that I needed to install #10 to the panel... nothing about the extension cord. I got the extension cord down, I can make that up EZPZ. the only reason I don't dick with the panel is insurance shit.

>>1139463
That sounds like false advertising to me, or is it half bits like a multimeter count?

>>1139391
>get real and drop the adc and µC
How the fuck are you supposed to turn an analogue voltage into a digital integer without an ADC?

how reliable are gate delays when being used to pulse something? I have made a simple astable multivibrator and would like to output a single pulse to trigger a t-latch.
circuit simulators i have tried seem to assume there are no gate delays it seems

>>1139464
>that nigger just told me that I needed to install #10 to the panel.
I'm that same nigger. It would be better to go with #10 all the way but 5' of #12 is still fine.

all except
>>1139421
>>1139447
have been my replies

I apologise if this is the wrong place but /diy/ has been good to me so I figured you could help.

I work in apartment maintenance. I'm fairly new and still learning a lot. I have a bedroom that is, for lack of a better word, pulsing. The lights (both fixtures and lamps) dim slowly and then brighten, seemingly at random. I'm almost sure something is pulling too much amperage but all of the switches and outlets on the circuit are wired tightly and good. Any ideas what's going on?

>>1139467
Oh, you get all the bits. It's just that the bottom few might not be usable.

>>1139481
well it may be more then 5 feet idk, it may go up into the attic, about 3 or 4 feet, across, 6, and down another 4-5 feet.

im still unsure of the gauge used. this is why I wish I had a shed hooked to its own panel, at least I'd be willing to fuck with shit cause a whole house wont burn down, and insurance gives fuck all.

>>1139495
I'm just a physics major and novice electrical bumblefucker, but an explanation for that is you've got a choke point somewhere. Assuming your country isn't in a population or power crisis, I don't think the power stations will be unable to keep up with the city's load a varying rates, so that should be out of the question, but whether it be in your apartment/house's internal wiring, just the wiring in your room, or even wiring back in the substation or house connection, there could be some wiring with too high resistance. Suppose you've got a resistor R0 somewhere in the wiring that your room's lighting and other devices draw from, like pic related. R1,2,3 are the devices. the voltage V1 across those devices is equal not to V0, but to R1*R2*R3*V0/(R0(R1(R2+R3)+R2*R3)+R1*R2*R3). In extreme examples, if R0 is 0, then the circuit behaves normal and you get full voltage across your appliances. If R0 is infinite, then you will get 0 voltage across your appliances. If the resistance of your appliances are all very low and they're all drawing shit-tons of power, then the resistance of the wire R0 will be comparatively high and will have a significant and very high voltage drop across it. If they're all open circuit then there is no voltage drop across R0 because there's no current.

TL;DR, when you increase the power through the appliances, the voltage across them all decreases. You might think that you're not changing the power being pulled by any appliances and so shouldn't be noticing a difference in brightness and therefore in voltage, but many things in a house do not draw power continuously. It could be a heater with a thermostat that toggles the elements, a fridge, a hot water tank, a soldering iron, etc. If it's a slow, 30s-10min oscillation, then it's probably something temperature controlled. If it's a momentary dip every 7-20min, it might be commercial breaks across the country causing people to put the kettle on and use more power. Cont.

>>1139495
>I work in apartment maintenance.

>The lights (both fixtures and lamps) dim slowly and then brighten, seemingly at random.

Locate the breaker for that room.
Pull the breaker and inspect the contacts on the breaker itself and the mating contact in the panel.
Check the tightness of the screw on that breaker.
While you're cinching screws in the panel, check all neutrals and grounds too.

>>1139510
>inspect the contacts

sorry - you're looking for damage or corrosion

File: household wiring.png (7KB, 640x400px) Image search: [Google]

7KB, 640x400px

>>1139509
Cont. At this large scale, even the minuscule mΩ of resistance in wiring and probably substation losses become significant, and there's a small documentary on youtube:
https://youtu.be/slDAvewWfrA
Now unless you have a particularly beefy fridge, hot water cylinder, heater or god forbid, soldering iron, this shouldn't be terribly significant unless you've got some bad wiring somewhere. So I'd check which rooms/sockets this happens in and locate the breaker for this section and check that it's properly installed and that there isn't any thin, frayed, or partially detached wiring behind it. If your problem isn't there, then you'll have to go crawling in your walls or ceiling or wherever wires go. Probably something to ask the electrician, because digging through plasterboard without knowing where your wiring might be wouldn't be too fun. Either way I'd get it checked out because it probably means your wiring might not be safe, especially when you draw a lot of current through it, as it might heat up enough to start a house fire if you aren't careful.

>>1139425
>Pic related. the plate
The old motor draws more current than the new.
14.8A / 14.2A
There may be nothing wrong with the old motor except trying to run it on too small wiring.

>>1139053
>I've designed a few low voltage microcontroller-oriented things before

Should be fine, just remember the various problems involved in routing high-power signals around a board. And look up the details of driving a FET, if you haven't already.

>Also, I want total control and to be able to tune the thing to the absolute max speed

You won't do any better than a Chinese controller, in this regard. Top speed is ENTIRELY dependent on the characteristics of the motor and the voltage it's being supplied with. In a nutshell, the back EMF puts a hard limit on how fast a permanent-magnet motor can spin. Unless you get a phenomenally shitty controller that can't even commutate the motor properly, the only way to get the top speed up is to use a higher-voltage battery, or to run the controller/motor off of a power supply instead of directly off the battery.

And a power supply of the power requirements you'd need would be pretty expensive. Plus, I would have no idea as to what the physical limits on those motors are, and those kinds of failures tend to be a lot more...catastrophic...than electrical ones.

>>1139065

Standard charge pump gate driver, no reason it shouldn't work.

>>1139471

Technically possible, practically inadvisable.

But I think he meant the combination of ADC and micro. A lot of microcontrollers already have ADCs built-in.

File: resistometer_2.png (11KB, 770x616px) Image search: [Google]

11KB, 770x616px

>>1139536
I didn't think an Arduino Nano had any analogue inputs. It has 8, but they're only 10 bits, so I'd only have a ~1000 count meter unless I could somehow chain the multiple analogue inputs together to get a higher resolution. Though this is almost certainly possible by putting the same signal into each of the 8 inputs and adding an 8th of the minimum detectable voltage step to each pin, I don't think that level of low voltage control is particularly easy. I'd just buy the 24 bit (or less) ADC, but I don't know how to put 24 bits (or any more than 12) of data into an Arduino Nano without a bit of shitty multiplexing, which I guess I'll have to do.

I found a neat circuit that uses a 10V voltage reference IC and an opamp to produce a very stable constant current source, which I might well do, though if I can use a 3.3V(5V?) voltage reference IC I can use it as the 'duino's power supply too, and hopefully the 7seg display too. The circuit uses a resistor as the current chooser, so hopefully I can use one of those digital potentiometers accurately enough to define the auto ranging.

I think that's all the problems worked out except the transistor(s), for which I'll probably just look for a BJT (or darlington) that triggers at the minimum current and hope it doesn't explode at the maximum. I'll have to calibrate for this transistor, but that's a lot less calibration than the auto ranging many-FET design iteration would have required. From then on it's all programming.

Also what does your "technically possible" refer to? Just making an ADC out of transistors?

Ok, it looks like I can't buy a digital pot with value 100k+ and 24 bits. Is there some way to make an analogue electronic pot that I feed a DAC'd voltage into?

File: Class D.jpg (151KB, 881x1283px) Image search: [Google]

151KB, 881x1283px

So, I just bought this (very cheap, Chinese) class D amplifier board. I'm unsure of how to wire it, though. Does anyone have any insight? I'd like to use it to amplify some computer speakers.

Pic is my best guess. I don't really know what the upside-down T is, but I'm assuming its the ground for the input(?). I don't have much hope of finding the number or I would look up the datasheet. It hasn't arrived in the mail yet.

>>1139630
I'd check the - on the speaker connections before wiring it like that.
On some amps the speaker can't be connected to ground/common
If you look on the board and traces go to both from a common source it's OK but if they go to an active component the speaker connections may be looking at differential outputs.
The - and + are connected to outputs of opposite phase to get more power from low voltage.

File: bridging2.gif (7KB, 360x342px) Image search: [Google]

7KB, 360x342px

>>1139630
>>1139645

File: 3.jpg (106KB, 1346x614px) Image search: [Google]

106KB, 1346x614px

>>1139645
So, I'm pretty sure I've found it. Pic related. Good call, I don't think I can connect the output negatives.

Unfortunately, the speakers I want to amplify just use a 3.5 mm jack for input.

I'd probably have to sever the connections at the back of the speakers to make this work with them. So maybe I'll find something else to do with this amplifier. It was still worth a dollar, though.

>>1139649
I suppose I could bridge the outputs like that, although I'd only get mono output by the looks of it.

>>1139654
>I suppose I could bridge the outputs like that,
I posted that just to show what some IC amp outputs are like.
That's a diagram of what each [driver] block on your pic is.
It wasn't meant as a solution.

>>1139659
Ohh ok then, thank you!

>>1139514
The old motor gave up funky smoke from the windings. Pretty sure it was kill. Also had a thermo overload switch built in. And the switch worked cause after so much smoking it would pop itself. And the breaker.

>>1139559
>It has 8, but they're only 10 bits

Technically, it only has 1 of them, which is multiplexed between numerous pins.

Anyway, I didn't say I necessarily agreed with dropping the ADC. It would cause problems if the micro's ADC peripheral did less than 12 bits or so.

>but I don't know how to put 24 bits (or any more than 12) of data into an Arduino

Use an ADC with serial output. SPI is pretty common. You'd have to go a bit out of your way to find a high-resolution ADC that used anything but serial, actually. Barring the odd niche case, parallel communication is really only used for very high-speed communication, where it's difficult or flat-out impossible to get the bandwidth required over a single line.

>I found a neat circuit that uses a 10V voltage reference IC and an opamp to produce a very stable constant current source

Not sure why you'd want to do this. As far as I'm aware, most actual multimeters use a precision voltage reference and calculate current (and therefore total resistance) across a resistor. Eliminates the need for the current source, instead using just the reference. You don't really see current sources being used until you get into tools made specifically for low resistances. Kind of impractical to try and get 1,000A+ out of a voltage reference to measure a 1mΩ connection.

>Also what does your "technically possible" refer to? Just making an ADC out of transistors?

Making a basic successive approximation ADC out of discretes really wouldn't be difficult. Making it accurate enough to be reliable at greater than 4-6 LSB is another matter entirely.

File: 02.png (361KB, 800x762px) Image search: [Google]

361KB, 800x762px

I am looking for a cheap ass digital oscilloscope with also logic analyzer to mess around.
I found about the bitscope micro (~100 eur), what do you think about it?
Specs:
http://www.bitscope.com/product/BS05/?p=specs

>>1139766
I guess it makes sense that a 24 bit ADC doesn't have 24 output pins. That eliminates need for any (manual) multiplexing, assuming that's what serial means. I learnt all my digital logic theoretically at university, not by experience, and I'm definitely noticing things I wasn't taught.

When it comes to constant-current/constant voltage, >>1139393 convinced me that current was the way to go. The reasoning behind "not needing a bunch of toggle-able resistors" threw the argument in current's favour, but with at least this design, I need an array of resistors/digital pot anyway. With a current source I can easily measure low resistances but run into trouble dealing with high resistances, and it's the opposite for a voltage source. For my current source, the current is "I=Vref/Rpot", so for the voltage to be at maximum of 5V or lower, the external pot must reach accurately 100MΩ for it to be able to measure a 100MΩ resistor, and the most I can find is a 1MΩ digital pot. Assuming a digital current source can't produce a voltage higher than the voltage source it was made from, that is. That's not happening, so constant voltage is the key.

With presumably 24 bits of resolution, I can read a ~100MΩ resistor to ~2 significant figures with a 100Ω dropper resistor, which would also be fine for measuring anything down to 0Ω. It looks like I can auto-range with only two ranges/resistances as long as I get an accurate 24-bit ADC (which is pretty overkill), a 50Ω like previously wanted, and probably a 50kΩ or higher for the big resistors. No need for a pot with these, I'd just switch between enabling the two with a transistor or two, hoping that the transistors don't have a non-linear effective resistance with respect to voltage/current (they probably do). Maybe I should use logic-level physical relays for switching the resistors instead, but I can't do that with the auto-power supply because the coil will act as a resistor initially, killing initial accuracy.

>>1139781

You should consider buying these tools for when you need them

If you are doing digital logic you need a logic analyzer. I find that an oscilloscope is a more powerful but more niche tool than a logic analyzer. If you think about it, the only time you would ever use a scope is when you're building a powersupply or doing something that involves opamps (purely analog work). Hobbyist projects tend to follow the pattern of feeding analog data into an ADC or just staying digital the entire time.

As for that product it looks overpriced. You aren't even getting a case with the thing.for 100 euros I would expect a low tier FPGA. I would buy a used (and better) one on ebay.

>>1139824
>staying digital the entire time
Tell that to our amplifier brothers. Don't listen to this guy, everyone around here got their russian military 1970s surplus oscilloscopes by dumpster diving, and you can too.

why are analog multiplier ics so expensive

>>1139849
what do you mean analoge multipler?
something opamps and OTA's cant do?

>>1139857
the application (electronic load concept, constant P and R) can't tolerate the presumably excessive nonlinearity of a discrete solution or the low bandwidth of a simple digital one.

>>1139824
>I find that an oscilloscope is a more powerful but more niche tool than a logic analyzer.
Visit any not-totally-shit electronics lab and there will be oscilloscopes. Finding a lab with a logic analyzer typically requires much more work.

>>1139857
Most likely something like AD633.
>>1139849
Niche products which have largely fell out of fashion due to rather limited performance.

>>1138806
what are some good components to keep in stock? Things like 2n3904 transistors.

>>1139837

I've got one too and I end up using it to read digital signals 95% of the time

>>1139866
>10pack of irf9540s or some other power fet
>555 or 556 timers
>a few $2 cheap ch430 arduino nano knockoffs
>terminal blocks
>a variety of op amps
>uln2003
>10pk small toroidal inductors for power conversion
>adjustable linear regulators (lm317)
>misc LEDs
>one or two esp8266s
>10pk chinese blue relays
>male/female headers
>potentiometers
>assorted zeners

any one of those can be had for $2 on ebay and may or may not be useful. not all components but things that you'll wish you'd bought when you get to postpone a project for a month.

>>1139824
You can use a scope as a logic analyzer in a pinch for most hobbyist stuff. Personally I don't find logic analyzers that useful.

>>1139895

Usually they are very useful because they allow you to read/store digital data

If you're trying to reverse engineer or even debug a bus/protocol you have to be able to do that

>>1139899
Just my opinion. I have debugged from serial comms to PCIe, and the logic analyzers were generally more trouble than they were worth.

>>1139866
depends on what you do.
like I work with modular synth stuff so i keep a healthy stock of assorted opamps and cmos logic.
aswell as shrouded headers, jacks, pots etc.

i find it very prudent to just buy 100 pieces or more of the most common resistors.
resistor kits are useful for most values, but 1k,10k and 100k fly out of the drawer so buying a lot is nice.
same for caps i guess, i expend a lot of 100nF, for example.

File: PicsArt_03-05-05.06.03.jpg (172KB, 1200x715px) Image search: [Google]

172KB, 1200x715px

I bought this Akai amp for my turntable. Turns out it was designed to also power the turntable, in order for it to select "phono" out of the three other options (tuner, tape, CD) the power supply needs the power going out. It works well if I just stick a wet finger into the supply but that would not be very comfy to keep my finger in the power supply while listening to my records. Not comfy at all.
So I'm just wondering here, how do I make it think it's powering the turntable?

>>1139874
why a variety of op amps?
Like some cheapo 741's, some 324's, and rail to rail op amps?

>>1139959
the cheap ones all have something terribly wrong with them. low supply range, non r2r, 5mv offset voltage, 100na input leakage, etc. if you buy a few you can pick which one's best for your application.

>>1139950
>DC 12V at 100mA
Easy, just get a power resistor or two. Assuming that's the standard power draw of the turntable, the power discharged will be "12V * 0.1A = 1.2W", so to use a standard 1W resistor, you'd have to install a couple in series. The total resistance needed will be "12V / 0.1A = 120Ω". Of course if you're able to get multi-watt power resistors you'll be fine.

If this 100mA is the minimum current (unlikely because your wet finger worked) value needed to allow your turntable to work, then use a slightly lower total resistance, 10% should do. If this is the maximum current value (unlikely because that's not much power for a turntable), then use a slightly higher total resistance, again 10% should do.

But since your wet finger did work, consistently at that, then a ~100kΩ resistor should work, and the power dissipated will be in the low milli Watts / micro Watts, so no need to worry about that.

You could install this externally or internally, but if you don't have a spare plug to fit in that socket, then I'd open it up and twist/solder the resistor into the inside of the power socket.

>>1139961
I bought 4 OPA2134s for $11.50 each because I thought I kept losing or ruining them, only used one of them and I know where the other three are now.

>>1139969
It works. Gee thanks anon

>>1140038
Not gonna lie, that's a first for me.

Anyone have the ohm project list?

is analog IC design good career choice?

File: 20170305_144927.jpg (4MB, 4160x3120px) Image search: [Google]

4MB, 4160x3120px

Edited out the quote.
This is right? And if so where does the switch tie in?? I know green goes to green as its ground. Black goes to gold because niggers love gold and white goes to sliver to keep it classy.

>>1140158
Only if you are good at it.

>>1140181
What does "good at it" mean though? I am a straight A student so far but I won't know if I am going to be a good designer til I become one and then its too late for change of field.

>>1140194
Well, that's the fundamental difficulty, isn't it?

I would say though if you have a decent amount of common sense, you should be okay. IC design tends to weigh academic success more highly, so grades are good. And the highest degree you can stomach.

File: Screenshot_2017-03-05-16-00-53.png (349KB, 1440x2560px) Image search: [Google]

349KB, 1440x2560px

>>1140178
Wiring diagram from the manufacture.. Idk what to do with this. I understand some of it. But how to tie in the switch I'm still lost at.

>>1140196
I want Masters for sure but I don't see myself doing PhD desu. Thanks though. I am studying for my bachelor's thesis now, I'll probably try to design some CMOS voltage regulator or something. We'll see how it goes. The professor I'll be working with says I should get internship easily with my grades but I don't believe him, this particular field seems like one of the most hardcore ones in EE and I don't see how an IC company could benefit from that... this field of engineering is really weird, all the civil engineers and mech engineers are already doing part time jobs but I feel like I still need several years to be actually usable. Such is life.

>>1140211
Definitely go for an internship if you can, don't sell yourself short.

If you can get an IC design job with a MS that's great, but for example, where I work the IC department is pretty stuck up and only hire PhDs for the most part.

can anyone explain characteristic impedance and its significance in high speed digital circuits? I think I've wrapped most of my head around it, but if anyone can explain it somewhat simply it would be really helpful

>>1140214
lol what company is that

File: 5jPMj.jpg (191KB, 640x750px) Image search: [Google]

191KB, 640x750px

>>1140194
>then its too late for change of field.

you can always lower your goals. if you go into EE and you have trouble with the second year electronics design and things like small signal transistor models, you will know in time that you might be a good EE but you will not be designing analog ICs, probably. but if you still like electronics and computers you can stick it out and get some other job.

Look up Bob Widlar on wikipedia or somewhere. He is a god of analog. His wikipedia page says that at one point in the 60s he was the designer of 80% (i think) of the analog chips being manufactured at that time.

but back to your career. when I was still in school I could not understand classmates who looked forward to working for the local power utility. I now realize that they probably had a job for life, and the engineering they do is just as valid and rewarding as what most engineers do. The top guys there probably get the challenging projects, and the droids do the grunt work.

engineering is a great career if you have the aptitude and the mindset for it.

Got some cheap LED lightbulbs, wanting do do a little project with them, mabye make a few flashlights.

>>Would i have to get resistors and heatsinks for them all.

>>could they be powered by battery or battery pack.

File: 1486062606904.jpg (5KB, 300x300px) Image search: [Google]

5KB, 300x300px

>>1140263
afaik socketed LED bulbs have conversion circuitry inside them. those that do won't operate off lower voltages because they're meant to accept a high voltage input (ac, but some will run on dc too). you couldn't practically run those off a battery.

>>1140272
So taking the bulb apart and salvaging LEDs is my best bet?

>>1140263
>>1140272
>implying he wants to screw a big light-bulb into a flashlight

I'll assume that you're willing to take these apart and put them in a different housing, and if so you'll need to include a fair bit of heat-sinking because they pack a fair punch. Compare the brightness of a normal torch to the brightness of your bedroom light.

Now these things sometimes have a capacitive dropper in them instead of a resistor, so it's not quite as easy to figure out the correct dropper resistor value for use with a lithium ion cell, but even using dropper resistors in this instance is probably a shitty idea anyway. You'll need an LED driver board, and a beefy one at that. You'll just have to look around, try looking for an automotive one, they run off 12V but should be able to take 3 stacked Li-Ion cells without a problem, and they should be built for high power in mind. The problem is headlights are a bit too bright and anything else on a car is too dim, so either you've got a choking (or melting) tiny power supply or an oversized one. And that's not even considering that they might be constant current (which I think LED drivers are).

So I'd take the wattage of your bulb, use that as a slightly bad approximation of the LED's wattage, and search for a board for that. The colour of your LED might also matter, whether it be warm white, cool white, or PULSATING CRIMSON.

Is there a minimum frequency to use for an SPI interface? I need a DAC, and I'll only need to set it less than once a second, so 100hz is literally the highest I would possibly go, bit-banging the data I need to set the output.

I know SPI itself has a "minimum" frequency of 1MHz or something, but I since I don't need ANYWHERE near that, I want to know if I could drive it slow. I'd expect my chip to only respond on the clock cycles (there'd be only the one clock, no data would be sent back, etc) but I'm trying to find any resource on running the interface slow and I can't find any.

I'm already looking at bit-bang chips which don't follow a protocol, so I'll buy those in addition (this is time sensitive), but if I can use cheaper ICs, the better.

>>1140210
Can anyone understand this shit? where do I put my on an off switch between? holy fuck its been almost all fucking day.

>>1140320
>Is there a minimum frequency to use for an SPI interface?

No digital communication has a minimum frequency based on technical reasons. As long as the parts can actually be configured to run slow enough, it'll work.

>>1140210
>>1140325

Without looking at the diagram (and assuming you're in the US), on/off switches go on the incoming black (live) wire. Period. It's always the live wire, since it's the one that actually provides voltage; the neutral is just a conductor needed to complete the circuit and the ground is there for safety and EMI reasons.


Looking at the diagram and the picture you provided earlier (and again, assuming US 120V mains and associated color-coding), you're doing it wrong. The relevant box is the one labelled "LOW VOLT. CCW ROT."

P1 connects to either hot or neutral.

White, yellow, and black should be connected to either hot or neutral.

Blue and T1 should be tied together.

P2, orange, and red should be tied together.

Ground (green) should be electrically connected to the case. Either properly, with a crimp-on ring terminal and a dedicated ground screw, or ghetto style, just mooshing it under whatever screw is convenient.


Admittedly, I can't see where P1, P2, or T2 are in your picture, or even know what they're actually supposed to be (I assume P/T are "post" and "terminal", but I'm not sure).

File: 5b5dc1d2-ebeb-4679-9af5-badd8e374525.jpg (114KB, 768x1024px) Image search: [Google]

114KB, 768x1024px

>>1140341
Thats the problem I'm having with the piss poor labeling. There are 7 wires. 1 being ground. So 6 all together. Black white and yellow go together? And orange red and blue go together. According to the lable on the motor. One says P1 an the other P2. I'M ASSUMING. p stands for power? And that black white and yellow are the live. And the others are the neutral.

But here's the kicker.. I'm reusing the switch on the tablesaw. It has 2 wires. A black and a white. What I was thinking was connecting the white wire to the 3 black white and yellow combo. And attaching the black side of that switch wire to the live end of the cable that plugs into the wall. Then taking the white wire from the wall cable and plugging it up with the orange red and blue wires. An green going to the earth wire on the motor? That sound about right?

>>1140350
>I'm reusing the switch on the tablesaw. It has 2 wires
Two wires in addition to the power cord or just two wires?
Show a picture of the switch.
It may be a double-pole switch.

File: 20170305_204156.jpg (4MB, 4160x3120px) Image search: [Google]

4MB, 4160x3120px

>>1140357
Its just a toggle switch..

Switch on the right with its female plug end. And new power cord on the left.

>>1140361
>Its just a toggle switch..

Then yes. What you described >>1140350
should work.

File: 1470158897484.jpg (49KB, 500x442px) Image search: [Google]

49KB, 500x442px

>>1138806
Not quite sure where on /diy/ to ask this.

For my class design project, we are building "drill powered vehicles," capable of propelling a rider along. We have the option to use a drill (Which is pretty easy, all things considered), but we also have the option to use an electric motor. We thought it would a pretty fun experience with all the electronics involved, so we've been looking into it. It would be fun to build a speed controller and power management system.

Unfortunately, we're struggling to find a DC motor powerful enough to do this from a company that doesn't only deal in bulk sale. We're limited to a battery voltage of 20 volts.

I don't suppose anybody here would have any suggestions? Otherwise this project really doesn't seem like it's all that big of a deal short of picking the most power efficient and powerful drills based on the testing criteria we have to meet. (Efficiency tests and torque tests)

>>1140492
You could be a cheating shitlord and use a 1 hp AC motor off an inverter, but it's probably a grey area in terms of rules. To get a powerful 18V-20V battery drill it will cost you an arm and a leg, though searching for a broken one isn't a bad idea, and for what you pay for one of those you'll have quite the value for your money, provided the motor isn't what's broken. You could try PWMing your way through a car starter-motor, and that could pack quite the punch, but it might be expensive and prone to overheating. Get one from a wreck and you should be ok on the price tag.

Problem is, drill motors spin quickly, and you'll need to gear them down significantly, provided this rider is human. I can't imagine a human being pushed far by a battery drill, so it sounds like an odd task. Have any dead-reckoned equations to tell us what kind of acceleration you're predicting from a drill/other motor?

Also try this for a great way to break a few teeth!
https://youtu.be/C5lp3fP1WKo

>>1140492

What's budget? Large motors made for RC can handle absurd amounts of power, as long as they have adequate cooling.

Otherwise...scooter motor. Duh.

>>1140320
>SPI itself has a "minimum" frequency of 1MHz or something
Look at the datasheet. Generally it says "DC" or something like that.

>>1140329
There are considerations like PLL lock ranges for some interfaces.

>>1140492
>>1140496
If you want to go the starter-motor route, here's a little guide I found relating motor torque to cylinder size and displacement, so if you know how much torque you need you can see what model engine's starter motors will work for you:

As a very general guide the stalled starter torque required per litre of engine capacity at the starting limit temperature is;
(i) 12.5 Nm/1 for twin cylinder engines.
(ii) 8.0 Nm/1 for four cylinder engines.
(Hi) 6.5 Nm/1 for six cylinder engines.
(iv) 6.0 Nm/1 for eight cylinder engines.
(v) 5.5 Nm/1 for twelve cylinder engines.

I guess the 1 means 1 litre. Just as an example, a 1.8L 4 cylinder Toyota Corolla's starter motor will produce at least 14.4Nm torque. If you put a 30cm diameter wheel straight on that, the radius will be 0.15m, and so the force will be 14.4/0.15 = 96N. If the mass of the cart is 120kg with a person, then you will ideally accelerate at a modest 0.8ms^-2, or 0.1g. Not even too bad, but I'd half that at least for starting out.

>>1140492
http://www.surpluscenter.com/Electric-Motors/DC-Motors/DC-Motors-Base-Mount/1-HP-12-VDC-1800-RPM-MOTOR-56C-10-2574.axd

+ jackshaft.

i need a cheap rough 0.7-2.0v reference. do some kinds of diodes have a more stable forward voltage than others? no zeners i've seen have reverse breakdowns that low.

>>1140633
If you literally want "diodes with stable forward voltage", they they're called stabistors. Despite the name, they're quite shitty.
IR and red LEDs are also occasionally used for that purpose.

Are IC references like LM4041 and TLV431 really too expensive?

>>1140647
expensive enough that i'd just pay a few cents more for a pic with a 5 bit dac. i just need a reference that i can divide down into a value from ~.28-.34v for motor stall detection. 3v supply rail.

i could just read the supply rail and average a pwm signal to get about that value but a diode vf would be easier.

File: Screen Shot 2017-03-07 at 12.46.42 PM.png (320KB, 1356x1052px) Image search: [Google]

320KB, 1356x1052px

It's resistoscope anon again, and I was wondering what the maximum power I could put through a 0.5W resistor without it producing a significant amount of heat, and the same goes for the resistor being tested, with minimum wattage of 0.25W. That's assuming 0.25W is as low as through-hole resistors get. I'm using 0.5W resistors because they have an accuracy threshold of 1% and not 5% from my local supplier.

The spreadsheet is when using a 5V voltage reference and the reference resistor "RR" in series with the test resistor "R?". The reference resistor has a maximum power use when the unknown resistor has a low resistance, while the unknown resistor has a maximum power use when it is equal to the reference resistor, hence the two columns. The sub-columns refer to current and resistor power respectively. Red is above rating, and orange is above half rating. Is the bottom of the orange range fine, or should I go deeper? I also made charts for 3.3 and 10V references, but the concept is the same.

>>1140806
And just so it was clear, I want to be able to leave the machine on for a couple of minutes and not have the measured resistance change because of heat buildup.

>>1140651
>>1140633

How cheap is cheap? Cheap as in "I'm a hobbyist with limited funds" or cheap as in "I'm a major manufacturer making a literal million of them and an extra $0.02 is another $20,000 to the bottom line"?

If the former, +/- 0.5% references can be had for $0.25/ea at qty. 10, even at DigiKey (shit like eBay will be even less). If the latter...IDK, use a diode and have fun going bankrupt lol.

>>1140651

Nevermind, if you're looking for "under a few cents", you're not going to do better than some random diode. You don't need to, either, if all you're doing is stall detection.

>>1140496
>>1140531
I should have been a tad more specific in my information. The vehicle will be under 200 lbs after factoring in the rider. So it doesn't need to be huge or anything, butmore is better. I'd totally go find a car motor but probably a bit overkill.

Appreciate the information.
>>1140503
Haven't finished budgeting but I hopefully don't wanna spend any more than 100-150.

>>1140552
Looks to be exactly what I was looking for.

>>1140839
200lb is about 120kg, that's what I was going for. I know a starter motor is way overkill, but if you can get one cheap, smoke 'em if you got 'em, am I right?

>>1140843
Kek. It's not actually a bad idea, I went poking around. But the issue I was having was that a lot of the specs we need for our calculations are available for a lot of them, so it makes it difficult

>>1140877
Find one cheap enough and you've got nothing to lose, but they do like pulling hundreds of amps, so either protect your circuit and battery, pull some mad PWM with big smoothing caps on the battery side, or watch your LiFePOs explode.

File: IMG_0827.jpg (964KB, 3264x2448px) Image search: [Google]

964KB, 3264x2448px

Ah fuck guys, I ran out of alligator clips.

>>1140916
use chip clips (as in cheeto bag clip)

>>1140923
Oh yeah, I guess bread tags will do as well.

>>1139404
nobody uses outdated TH components for things that don't need isolation or deal with high impedances, that's why you pay a premium for them

>>1140930
I get that, I just assumed they could use the same silicon dies for building both SMD and through-hole components, and they'd just do so to meet demand. They're probably deliberately ripping off the hobbyist market. Can I do SMD soldering with a 30W iron, and should I get that flux solder paste? I also don't have anything to hold the chip down by, so I'd better get good pliers.

>>1140934
Rosin core will work fine. Just tack down a corner or two first and it'll stick.

File: SP22G.gif (13KB, 589x212px) Image search: [Google]

13KB, 589x212px

I'm reading a continuous rotation potentiometer with a microcontroller. The track of the pot is not completely 360° however, so it produces incorrect values when the wiper is around the 0°/360° position (I assume because it's floating). Is there a way to fix that?

>>1141015

no, there's no way to fix that. it has to have a discontinuity unless you short the high end to the low end, making for a weird response curve that looks like the boy scout tent you had your first gay experience in.

>>1140806
You don't need that small resistors. For example, if you feed a 1 ohm resistor from 5V via a 1k resistor, you'll still see 5mV across it. That's easy to measure and getting 0.1 ohm resolution wouldn't be difficult either. Considering normal test leads add more than 0.1 ohms, that would be a reasonable resolution. If you aren't happy with that, you should use 4-wire measurement.

>>1140934
They don't really give a shit about the hobbyist market. Breakout board makers are there to meet this demand for fancy components in through-hole friendly format.
Toothpick is a cheap, common way to hold SMDs in place when doing what >>1140965 said. A fancier alternative would be (good) tweezers.

>>1141015
A very weak pull-up (or pull-down) from the wiper to some suitable voltage would be the simplest option. For example, if you have a 10k potentiometer, you could connect its one end to GND and the other end to +1V and then pull the wiper to +2V via a 1M resistor. The drawback is the induced nonlinearity.
Another option would be to test the pot wiper resistance after each measurement.

File: 20170307_183215.jpg (101KB, 691x389px) Image search: [Google]

101KB, 691x389px

I want to make a jacobs ladder and I have obtained a few microwave transformers, however before I start with the big boys, I thought I would start small and ripped this transformer out of the microwave. Unlike the big one however, this one has 3 pins on the primary and 4 on the secondary, how can I tell which is which? Google gave me nothing except for prices on them.
pic is little guy next to the big one I hope to move on to.

>>1140923
wtf, you amerifats have a dedicated clip to preserve potatochips?

>>1141025
>Another option would be to test the pot wiper resistance after each measurement.

What do you mean? Testing the pot wiper resistance is the same as doing a measurement.

>>1141053
Reading the position would usually mean that you connect the pot to some suitable power supply and then measure the wiper voltage.
Wiper resistance measurement would mean that you would feed current to the wiper (maybe by activating MCU's pull-up) and seeing how much effect it has. If it pulls the wiper voltage to VCC, then the pot isn't making contact. (Assuming that the pot feed voltage is less than VCC.)

>>1141025
I guess I'll do the resolution calculation to see if I need multiple ranges (reference resistors) first, then figure out what resistors I need. Thanks.

>>1141041
They'll probably be centre taps, so a good bet would be that the furthest ones from each other will have the most coils. A half-decent approximation would be to see which pair of pins has the most resistance, and you should be able to get a model of the tap scheme. I suppose you don't have an AC power supply, but if you did you could just put it across and measure the resulting voltage. You could still put it in the wall once you figure out which side is which and if any of them can take mains voltage.

File: 4n37.jpg (259KB, 759x903px) Image search: [Google]

259KB, 759x903px

trying top protect my 3.3v and 5v i/o from 12v damage , but i got lost at breakdown voltage

what are my usable voltage ranges with the 4n37?

>>1141274
1. The LED can take 6V reverse bias and the transistor 7V. Do not reverse bias them pointlessly.
2. Assuming you leave the transistor base open (like people usually do), the transistor can switch 30V.

File: opto.png (63KB, 532x410px) Image search: [Google]

63KB, 532x410px

>>1141296
ok, so can i do something like this to protect my low voltage inputs from higher output signals with something like this? and do i need resistors like in the pic related but mine will be 4n37

>>1141328
Not that anon but you do need the resistors.
The 10K is a "pull-up" resistor because the transistor acts as a switch to ground. Without it, it would switch between grounded and open. Alternately you could enable a pull up resistor on a uC.

The 560 limits current through the LED.
The value of it depends on the LED and Voltage
Get forward Voltage and forward Current from the LED specs.
Resistor = (Voltage-Vforward_led) / Iforward_led
(round up to the next highest avalable value)

Is there a reason you can't just use a voltage divider?
Opto's are generally used for noisy shit like motor spikes that are hard to supress or when the input wires are long and may generate noise.

This would turn a 0-12V signal into a 0-3 volt one:
12V->10Kresistor-> newSignal ->3.3Kresistor->GROUND
in other words, hook the 3.3V input (pin?) to ground through a 3.3K resistor and use a 10K resistor to the 12V input.

>>1141341
the 560 then is because its an led so need to accomadate, and the 10k is because its an NPN, and then math

i dont remember why im not using a voltage divider , something about the industrial shit im using, and signal speeds, and backwash through ground

>looking for voltage references on TI site
>see good one
>total chip area = 12 mm^2
>largest one is barely 90 mm^2
This shit really isn't made for hobbyists is it?

if i change my 3volts source to 9volts, would i also need to change my resistors? if so, how would i compute for their new values?

>>1141451
more details needed

File: LFR-page-0.jpg (130KB, 1286x1119px) Image search: [Google]

130KB, 1286x1119px

>>1141451
my bad, forgot pic

>>1141452
>>1141453

>>1141454
only the TX resistor absolutely needs to change. the current through it right now is (Vbat-Vf)/R=(3V-1.5V)/390 ohms = 4mA. if you want it to stay at 4mA use a (9V-1.5V)/4mA=1.875k ~= 2k to 1.5k resistor. i'd also recommend roughly tripling the RX resistor.

i don't know why but that circuit pisses me off. what's its purpose?

>>1141456
thanks m8, its for a line follower robot. would you mind telling me the reason for tripling the 22k resistor?

>>1141456
You probably don't like the look of the backwards LEDs. I've no idea what I'm looking at and I assume he know's what he's doing, but it still pisses me off like chaining a NOT gate to itself.

the LED in reverse bias is an IR photodiode m8. i forgot to indicate

>>1141473
Same with the diodes in forward bias too? I guess that's what the IR stands for, but what kind of voltage drop do you need that you use LEDs instead of normal diodes or a resistor?

>>1141478
the LED in forward bias is an IR LED, the one in reverse bias is a photodiode. they form an rx-tx pair. basically they track a black line on the ground . i could also use a white LED->LDR pair but infrared is better in terms of mitigating the noise from ambient light

>>1141465
the resistor pulls the PNP's base high thus holding it in cutoff until enough IR LED reverse current flows to pull the base below the PNP's Vbe. if it wasn't there you'd always have current leaking through Q1, though probably not enough to run the motor.

Vbe~=0.7v
0.7V/22kohms=32uA to start turning on the PNP.

now that i've actually stepped through it that resistor is fine as it is. it seemed too low to me.

>>1141484
doood thanks for showing your computations!!

File: photodiode.png (40KB, 2000x1340px) Image search: [Google]

40KB, 2000x1340px

>>1141483
You could try blaming it on your software, but the symbols are not the same thing.

>>1141491
dayum sry for the confusion m8

>>1141434
90mm^2 is nearly one square centimeter. How hamfisted you are?

File: not_actually_from_pinterest.jpg (28KB, 480x360px) Image search: [Google]

28KB, 480x360px

>>1141592
The average is about 20mm^2, and I need to sort by criteria more important than size. Please not that I'm living in a student flat while soldering out of a briefcase with only a pair of bulky $20 pliers to hold a component without burning myself with a $30 iron; I'm not exactly a pioneer of precision technique. I need a PCB holder with some sort of articulated SMD hold-down mechanism similar to those on a radial arm saw, like pic related.

>>1141599
Plenty of references are available in SOT-23 and they're quite easy to solder even with shitty irons and mediocre skills. You have at least 3 options:
- learn to solder bit better (practice with cheap transistors or diodes)
- adjust your requirements downwards so that you can use bigger (or through-hole) references
- ask someone more skilled to solder your reference of choice on a suitable adapter board and use the board like a through-hole component.

will 120/240vac fuses work on 12vdc? so if i have a 30 amp 120vac ceramic fuse will it blow at the same or similar amperage at 12vdc?

>>1141715
The fuse wire will melt at the same current. The problem is that the initial arc is more difficult to extinguish at DC.
The safest option would be to check the datasheet of your fuses to see whether they're also rated for DC use. That said, typically your random 240VAC fuse will work properly at 12V.

>>1141734
thanks

>>1141734
But that's assuming they've got no resistance and so have no voltage across them, when in reality a 12V fuse would typically have more voltage across it than a 240V. The resistor in series with a fuse in 240V would have to be 8Ω or less to blow it, while in 12V it would be 0.4Ω, so if the resistance of the fuse is the same then there will be a higher voltage across the fuse in 12V than in 240V. If the fuse has enough resistance for this to matter, it will blow at a lower current in DC. This effect might only be noticeable at much lower terminal voltages, but if the resistance of the fuse is in the range of 0.4Ω, cold or otherwise, you will notice a distinct difference.

>>1141748
Different fuses with the same current rating generally have quite similar resistances and thus voltage losses. If you're worried about the voltage/power loss, better check the datasheet first.
Whatever the resistance of the fuse is, it is designed so that the specified current will heat the fuse wire to the melting point. The voltage of your circuit has no effect on that. Higher fuse resistance lowers the momentary peak current just before the fuse blows, though.

>>1141748
>The resistor in series with a fuse in 240V would have to be 8Ω or less to blow it, while in 12V it would be 0.4Ω,

Get a couple of 240VAC 20A and 12VDC 20A fuses and measure the resistance of them.
Please report back with your finding.

>>1141757
But it's the power dissipated by the resistor that causes it to heat up and blow, not the current. It serves to reason that unless you're running then with a current source, voltage will make a difference. But again, the actual voltage drop across the fuse would be minimal from 240 to 12V. Apparently an AC fuse should be rated for twice the DC voltage you're using, but I'm assuming you're not using more than 50VDC anyway.

>>1141763
I just tested a single 0.5A 250VAC fuse from my (other) shitty multimeter and it gave a resistance of 0.5Ω, so through P = I^2*R, the power dissipated by it should be 0.125W. In 12VDC, for the power to be the same and the fuse to blow at the same speed (or at all), I = SQRT(P/R) so will be 0.5A. I don't get it. I just proved that the same current though the same resistor makes the same power, by using the same formula twice, what the fuck am I doing.

I guess the point of my earlier post >>1141748
was to state that your circuit will draw less load if you put a higher resistance fuse in there, but since there isn't any reason for a DC fuse to have lower resistance for the same rating whatsoever, I guess everyone else is right. At least I learnt something, even if it was about my own stupidity.

>>1141715
Also you might want to be sure you have a, sand-filled (always ceramic, but not all ceramic fuses are sand-filled) fuse because that should extinguish the arc. Otherwise don't do it.

>>1141599
Honestly you're best off getting an assembly microscope. Maybe a USB one would be cheap.

File: 81848614.jpg (27KB, 846x480px) Image search: [Google]

27KB, 846x480px

>>1138806
Trying to solder some SMD passives on a board.

I lay down some solder flux on the pads, place the component on the board, tin my iron tip with a little bit of solder and attempt to tack it in place. However whenever I try to solder it in place, the component just slides over to the solder on the iron and consequently gets stuck to the iron.

I dunno if I'm ruining my passives like this. But is there a better method to soldering SMD passives? I have managed to solder ICs with no issues, but passives seem to be a bit of a bitch.

File: resistameter.png (13KB, 1080x720px) Image search: [Google]

13KB, 1080x720px

Resistotestameter guy here, I ran some rudimentary significant figure calculations through, planning around a 24b ADC. Since the highest count 24 bits goes up to is around 16M (2^24-1), and the inverse of this is 6E-8, I rounded that to 1E7 and decided I had 7 decimal places of accuracy. I converted a few test resistances (powers of ten from 1 to 10M) at a few series reference resistances (50,500,...5M,50M), and converted them to voltage fractions read by the ADC through V/V = RR/(RF+RR), rounded them to the seventh decimal place, and converted them back through R = RR*(1/(1-V/V)-1). I hope this working is sound.

Since I want to have a 6-digit display (plus LED indicators for m,k,M) I found each reference resistor only covered a range of 2-3 orders of magnitude, and I'd need a 50M resistor(s) to accurately measure in the 100M range. I guess I'll get a current shunt too so I can measure really tiny resistances. They'll all be toggled with relays so I don't get the transistor parasite currents messing me up. The highest value resistor will have to be always enabled and I'll be compensating for that in the calculations, but it has to be like that otherwise the auto-power-on function won't be able to activate to turn any of the relays on in the first place. You don't need protection diodes on relay coils, do you? Just posting this to check that I'm in the right ballpark.

I'm still not sure if a single FET (or BJT or Darlington or FETlington or whatever) will be able to switch/trigger completely with a 100+MΩ resistor in series with the gate, be able to handle the current when there's just a 50Ω resistor (or even a current shunt, I'll look into high-current Li Ion protection circuits) in series, and produce no voltage drop?

Since the voltage doesn't actually matter as long as the ADC gets the same reference voltage that's put across the two resistors, I don't think I'll even need the voltage reference, and I'll just stick with a battery potential of 3.7V or so.

File: SMD vice.png (12KB, 1080x720px) Image search: [Google]

12KB, 1080x720px

>>1142025
You need some sort of really tiny bench-mounted clamp, like described in >>1141599 , or a good pair of tweezers, or really fine pliers I guess.

For now, glue a pencil perpendicular to a length of wood or metal and hinge it on your desk at the end, and use some sort of weight between the load and the fulcrum to put force on the pencil tip.

>>1142029
Tweezers, why didn't I think of that. And here I am trying to free hand this shit bumping it with a screwdriver.

Anyway, I tried repairing an 8051 dev board I got from my university. I recently fried the thing accidentally hooking up 10V to it. I had a spare PCB lying around so I soldered on new ICs and took all the passives from the old board to the new one. Tried hooking it up to my computer, usb device not recognized. Would this imply the passives are fried or that I incorrectly soldered the ICs or both?

>>1142037
Caps might be fried, especially over-voltage protection ones, test them with your multimeter, but otherwise I'd just go point-to-point checking logic-level voltages and checking that every chip is live.

File: 2017-03-09 11_57_50-casiotone mt-100.fzz_ - Fritzing - [Breadboard Weergave].png (298KB, 1153x796px) Image search: [Google]

298KB, 1153x796px

Im currently working on a replacement control unit for a casiotone mt100. I found a paper online wich dealt with the sound generating chip. Now im able to load my own waveforms in to it and thus unlocking a lot of mod potential. After making a control unit for all the standard functions im going to make a sequencer/drum computer and midi in out functions.

Any ideas as to what else i should implement?

>>1142025
>put some solder on one of the pads
>grab passive with tweezers on one end
>put some solder on the other end
>hold part in place and use the heat of your iron to reflow the solder and make a connection
The part is now locked in place and doing the other side is trival.

>>1142028
>Since I want to have a 6-digit display
Sure, you can display 6 digits, but keep in mind that the last digits generally mean jack shit.
For example, your average metal film resistor has a temperature coefficient of 0.01%/K. This means that if you use such resistor as your reference and your meter is displaying 5.00000k now, it will display 5.00050k (or 4.99950k) when it's one degree warmer. And this is just one source of error.

>I'm still not sure if a single FET will be able to switch/trigger completely with a 100+MΩ resistor
While this is doable, such "resistor detector" will be obnoxiously sensitive to everything. This can include things like nearby lights, you touching the device or just moving near to it and so on.

What are some good side projects that I can finish in a week or so that will beef up my resume? I have access to a power supply, a soldering iron, and a very shitty 20MHz scope I picked up at a physics department garage sale (which I suspect doesn't really work well).

I'm hunting for jobs in EE, and every employer I've spoken to that has turned me down has told me that personal projects are very important when you don't have experience. However I'm having trouble coming up with something that would look impressive to an employer but not take me months to finish.

>>1142084
lipo buck boost portable power supply - as complicated as you care to make it. can just be three ics. one for lipo charging, one switcher, and a uc to sense/display outputs and manage the other two ics.

iot home sensor - stick harmful gas detectors, humidity, and temp sensors on a board with an esp8266 and interfacing uc. find more dumb shit to bloat it with for bonus points.

both should be on a proper pcb and as smd as possible.

>>1142060
The extra digits would be more for if I've got a 470 and a 1M in series, but I do see your point that the values reach insignificance at that point unless I'm dealing with constantan resistors or something. Makes you wonder the point of a 60,000 count multimeter, huh?

If I shield the FET and wires connecting to it completely I should be able to stop that, but if the nearby light problem would do photoelectric effect wizardry with the metal contacts that I'm to put the resistors themselves into then yes I have a problem. I wonder if temperature differentials and different metal junctions could be peltier enough to also induce a tiny voltage?

Anyway, I thought of an auto-on function by putting a tiny inductor with an insignificant time constant in series with the resistors and using a highly sensitive hall effect sensor to pick that up, but that's probably either stupid or unrealistic. I guess I'll do what the other anon suggested and just toggle power to the 7-seg display.

Hello /ohm/

I want to make a 5V PWM signal generator with a 4% duty cycle and a 2 KHz frequency. I saw a couple of guys use a 555 timer IC and another one with a LM324 quad Op-Amp - but none of them showed how to change anything on their circuits to make the PWM I want.

Sources :
http://www.ece.utah.edu/eceCTools/ECE1250/ECE1250_Lab04_OpAmp_PWM.pdf

http://www.pcsilencioso.com/cpemma/pwm.html

http://www.pcsilencioso.com/cpemma/555pwm.html

Help me out guys - Or just reference some material on how these circuits are designed - I'll post the completed code and schematic here for everyone to see / use.

File: pwmcirclel.png (12KB, 573x354px) Image search: [Google]

12KB, 573x354px

>>1142336
I'd use a 555 timer in astable configuration. If you need a 4% duty cycle, you can stick an inverter at the output and have the 555 timer spit out a 96% duty cycle. Changing the pwm is a matter of playing with the passive values. I built a circuit a while back whose pwm I could modify with a potentiometer (pic related).

http://www.daycounter.com/Calculators/NE555-Calculator.phtml

>>1142084
Anything worth talking about is probably going to take longer than a week, just saiyan.

>>1142153
What anon means is that your digits are worthless unless you've verified that your apparatus can actually deliver accurate results. Lab multimeters with 6 digits ain't cheap.

Is there a specific reason the pasta suggests The All New Electronics Self Teaching Guide instead of the newer Complete Guide With Projects?

File: 92fec25eee387a7109f6839e477beaf06efc96920754778dacafc3df00f1e4da.jpg (49KB, 500x480px) Image search: [Google]

49KB, 500x480px

What <$100 soldering station would /ohm/ recommend? I am sick of my $15 shit iron.

>>1142361
Hakko FX888(D)
Weller WLC100 if you can handle only working with analog.

I personally use the Kendal 862D+, mostly because it has a hot air gun. It could be better for it's price point, but for the hot air, I'd say it's worth it.

>>1142355
I mean, I can get a 24 bit ADC (or even a 32 bit one for 10x the price) and calibrate my microcontroller for the resistance. I could even attach a digital thermometer to account for temperature variations if I wanted (in the reference resistor and ADC), but the actual tolerance of the test resistor would be sought to get right, and I'd need a stupidly low voltage to prevent my resistors from heating up. ADCs that work in the mid-low mV range don't exist, especially not at high resolution. But even if I could correct for all those the resistor being tested might be wire wound or carbon or any different sort of resistor, all of which would have different temperature co-efficients and are accurate to within no less than a percent. Unless I'm testing high-precision reference resistors there's no point.

>>1142361
Is that an ironing-board covered in mince? Also get a weeb-iron, the Goot PX-501 temperature controlled soldering station looks ok.

>>1142361
And here I thought wrapping a strip of bacon around a soldering iron was the most autistic way to cook.

Is it save to draw 3amps from a AA battery for like 5 seconds? I want to test something and don't have a psu that can handle it.

>>1142340
Wow thank you ! You are a life-saver. Thats a neat circuit too.

For the inverter, should I get an IC or what ? I googled inverters and I get 5v to 220v inverters lol.

>>1142361
Also this, I got a soldering station with all sorts of temperature settings - what settings do you use for normal SMD soldering ? I never saw a difference but I dont want to burn my kit out,

>>1142427
AA batteries wont blow up if that is what you are asking, I doubt it will give you the 3A you want. NO GUTS NO GLORY

File: 311277.image0.jpg (10KB, 536x378px) Image search: [Google]

10KB, 536x378px

>>1142429
OK, for a 5V VCC , what might be the optimum values for the 2 resistors ?

R1 depends on the load. It needs to supply enough current to keep Anot high. (also needs to be high enough to keep transistor current down to acceptable levels.)

R2 depends on the transistor.
Figure max collector current, or just estimate it at 5V/R1.
Divide that current by the transistor DC gain to get base current.
subtract ~0.7V(or lookVbe) from the input voltage then divide it by base current to get the resistor value. round down to ensure enough base current to switch the transistor fully on.

Or for general purpose switching just try a couple 4.7K.
If the voltage on Anot doesn't go high enough, drop R1 to 2.2K.
If the Voltage on Anot doesn't go low enough, drop the base (R2) to 2.2K

>>1142441
I cannot thank you enough for helping me anon ! Really if there is anything I can help you with - let me know - I am a programmer by profession.

>>1142441
>Divide that current by the transistor DC gain to get base current.
You should overdrive the transistor rather heavily, meaning typically 5-10 times higher base current.
If both input and output are 5V stuff, then R2 = 10*R1 would be a rather conservative choice.

>>1142427
what the other anom said. batteries have internal resistance that, besides dropping voltage, increases as you draw more current. (not sure if that's true for all chemistries)

google says an AA's esr is approximately .15 ohms so you'll see about a .5v drop which might fuck up whatever circuit you're driving.

File: filter.png (44KB, 700x348px) Image search: [Google]

44KB, 700x348px

If I have an oscillator and a separated analog and digital supply and ground like pic related, should I hook up the oscillator to the analog or the digital side?
I obviously don't want any noise on my analog side, but I don't want my oscillator suffering from noise from another source either.

>>1142427
http://www.candlepowerforums.com/vb/showthread.php?64660-Alkaline-Battery-Shoot-Out

found this.

>>1142486
If you're paranoid, power the oscillator via a similar filter you used for your analog stuff.
If the oscillator is sensitive to noise, but you can expect that it won't disturb the analog side (much higher operating frequency, for example), you can use the analog supply for it. Or you can take it via a simplified filter from the analog supply.
If you trust on your ground (solid, properly routed), omitting the ground filter might be worth considering. Doubly so, if the analog, digital and oscillator sides need to communicate with each other.

In short: it depends.

>>1142429
The quality of your of your station will determine your temps. A really fancy one will be able to work fine at 550c, I keep my 888d on 650 personally. Dwell time and tip size make a larger difference in heat transfer anyways. Stay below 800 tho, it gets easy to rip smd pads close to there.

>>1142492
That's the thing, I'm not sure if the oscillator is sensitive to noise. If it's not, then there's no downside to using the digital supply. If is is though, i think it'd ruin everything since it's the reference to a PLL. I guess I'll add some pads for a filter. Thanks

>>1142518
If it's a crystal oscillator and you're using it for digital side's PLL, then normally you'd power it from the digital supply. A simple ferrite bead + ceramic capacitor filter can be added for oscillator's positive supply, but that's usually not needed. Ground should go directly to PLL's ground.

>>1138806
Are IR detectors the same as Ir receivers?

>>1142541
a detector is a single componant that reacts to IR light

A receiver cantains a detector and other circuitry that amplify, filter, and convert the detector signal and give it meaning (with a protocol)

>>1142427
>3A
>5s
>1.2V - 1.5V
Easy, just get a 32F super capacitor, put it parallel with the battery and wait till it's charged. Might take a minute or two. They cost 6-20 dollars online, go for it.

Is it just me or do links to other posts no longer show properly in this thread? Can't refresh at the bottom of the page either.

>>1142626
be warned that some supercaps have up to an ohm esr. i was just looking at a 10F with 80mohms though.

>>1142647
You can get them as low as 18mΩ for $9, I'd say that's as good an idea as any, besides buying a cheap DC-DC converter and running it off a wall adapter.

>>1142626
He could just buy a pack of AA's and wire them in parallel for less than the cost of the supercap. Minimum four, eight would be better.

>>1142367
>ADCs that work in the mid-low mV range don't exist, especially not at high resolution
You know by definition, if a 12-bit ADC has an input of 5V, then it's LSB is ~1.2mV. At 1V max it'd go down to .2mV. 16-bit 5V goes down to .076mV LSB, 16-bit like THE FUCKING MICROPHONE IN YOUR COMPUTER AND EVERYTHING ELSE

>>1142675
That is a far better idea. I also did some mathematics and realised that less than half-way through, the cap would no longer be outputting 1.5-1.2V. To account for this, it should have a capacitance of at least 55F, and you can get a 60F one for about $12 NZ, but a bunch of batteries in parallel would be far easier.

>>1142682
That's not what I meant, sorry for the confusion. What I meant was there aren't any 1-50mV MAX ADCs which specialise in taking an LSB from those of a few nanovolts, which I need to be able to measure high resistance values. At least not the way I understand the datasheets. They all state they have a reference voltage input of 2.5-5V or so, so I m limited to using a voltage source in that range, instead of using a stupidly low voltage so that I can measure current shunts without blowing everything the fuck up, not to mention very low value resistors.

File: 19fig1.gif (2KB, 356x375px) Image search: [Google]

2KB, 356x375px

>>1142429
Yeah you can get an IC for the inverter, pretty sure that's how they all come packaged, most digital circuit elements will be found in ICs. Try getting something that will output a 5V signal. If the inverters you find output weak signals, consider using a bjt to amplify the output back to 5V.

Heck if finding an inverter is being a problem, scrap that all together and just stick the 555 output onto a transistor to invert the pwm signal (pic related)

File: jt.png (839KB, 862x633px) Image search: [Google]

839KB, 862x633px

Soldering practice: "Joule Thief" with an SMD blue led riding on top of an sot23 transistor, with the torroid core from an ethernet port.

No clue what the battery is or where it came from, but it reads 0.9v

>>1142121
That portable power supply sounds like a good idea, I could try experiment with some wireless charging methods too.

>>1142355
I had a sinking feeling that would be the case. I'm a neet anyway so I have all the time in the world right now... you know of any worthwhile projects that I could tackle that would get me hired for a hardware design position?

File: sDFSD.png (2MB, 1006x1244px) Image search: [Google]

2MB, 1006x1244px

These are just as /sqt/ as they are /ohm/, please excuse my possibly dumb questions.

I'm making a device with stepper motors. Most will require cable runs up to a meter from the "control box" to each motor. Is it better to run long cables from the mosfets to the steppers, or long signal wires from the control box to the mosfets (reads, should I place my mosfets in the control box or as close as possible to each stepper?). Pic related is the (ribbon) cable I'll be using.

Speaking of pic related, I need to work out how many amps each conductor can take at 12v, meaning if I run high loads through it I'll know how many conductors need to be paralleled for a given max. current. Is there a more elegant way of determining this than simply wiring it across the outputs of a big fuck off power supply turning the amp knob up till the smoke come out?

File: trailer cable.jpg (20KB, 515x515px) Image search: [Google]

20KB, 515x515px

>>1142766
Those look like data cables, so if they're not for speaker purposes, they probably shouldn't be handling more than an amp each, probably around 500mA or less. Steppers need 5 or so wires right? Trailer cable looks to be able to handle 5A each conductor and comes in 5 or 7 strands at my local Jaycar, not sure what you'll be able to find locally or what kind of continuous currents you'll be facing. Also it comes in 10m lengths for $50 so you might want to look somewhere cheaper if Jaycar is your local shop.

>>1142766
>should I place my mosfets in the control box or as close as possible to each stepper?

wont make any difference. the same cable will have to carry the same current the same distance, so the cable loss will be identical either way.

>I need to work out how many amps each conductor can take at 12v

for standard wires there are lookup tables. your cables look like PCB traces, and there's lookup tables for those as well. problem is, you dont know how many ''ounces'' of copper those cables are equivalent to. so, you can either do some destructive testing, or look for the maker's documentation.

>>1142737
You have a BSEE? More and more companies are only hiring MS and up for design, you might have to settle for New Product Introduction level test/manufacturing stuff.

What do you want to do? Personally, I like FPGAs, test equipment, and radio, so I do hobby stuff in those areas. Robotics? Drones? What you often can do is find a more technical angle to a common hobby project and drill down a few more layers so that you're implementing more of the stack, so to speak.

File: gears.png (680KB, 599x603px) Image search: [Google]

680KB, 599x603px

Where can I find a cheap solution to building the chassis for custom robos?
I got a cheap set of Meccano at the trash and treasure market but it wasn't nearly big enough so I got more gears from aliexpress (pic related)
Now I'm looking for something similar to Meccano to build my chassis out of, perhaps something made of plastic for cheap.
Maybe I shouldn't be posting this on ohm but I figured someone here might know and it saves creating a 1 trick pony thread

>>1142839
This might help
https://www.youtube.com/watch?v=XhtnTBbiJTI

>>1142839
Make a rough sketch first, maybe show us what you want. What I posted maynot be the best given you want a 'custom' chasis and I have 0 idea what sort of a configuration you want.

I have no idea how you will use those gears and can anyone tell me how to make a differential for a RC car ?

File: 17238162_434961146837349_568543261_n.jpg (89KB, 800x507px) Image search: [Google]

89KB, 800x507px

>>1142844
It's a little hard to follow

There's two motors, one that powers the wheels on a single axle, and another that rotates a doll head and arms on a vertical axle. I need to make a body that's strong enough to hold all the bits up but also not overly heavy and not going to blow out budget. Wood is out of the question and metal may be too beause I can't weld like in the video you linked. Metal might work if I can find strips of it and drill holes into it though now I think about it

>>1142851
He didnt weld - he said he used brass and used a soldering iron and solder to 'weld' the brass joints together. He even said the brass is easier to shape with minimum tools.

Follow what the guy did - get brass strips and a handsaw to cut them to size. It wont make it super tough to make and it will remain light.

Are you making one of those 'dynes' from Terminator 3 ?))

>>1142854
Oh right I'll admit I didn't watch beyond a minute or so.
I didn't realise solder could make for reasonably strong joints, I suppose it is metal?

I'm just making a silly robot from parts I found at the market, some guy was selling creepy broken doll parts for 50c so I grabbed some arms and a head and another dude was selling motors taken from drills for a dollar each

>>1142856
Dude its all good - I mean - maybe you'll learn a thing or two new from making this doll - maybe just scare dogs and critters with it - who knows ?

>>1142856
To be strong it shouldn't be normal electrical solder, but probably silver solder or something like it. It melts about 250°C higher so an anaemic 30W iron won't touch it. Joints this way are done with a gas torch instead, and the process is similar to brazing.

>>1142859
I have used electical solder on steel wire that had a galvanized coating. Shit sticks REALLY good. I used plumbers flux to clean it first too. Makes strong enough parts for RC stuff.. Has to be galvanized tho.

File: battpinout.jpg (59KB, 840x888px) Image search: [Google]

59KB, 840x888px

1 year old laptop battery is not charging but the AC adapter is able to power the computer. The adapter is also able to charge a different laptop.

I've replaced the battery with a new one and I'm still getting the same problem. I've checked the connection of the charger to the plug on the laptop using a multimeter and it's passing the voltage to the motherboard fine. The battery pins on the other hand are different:

I have 0V between pins 1 and 8 (+ and -) on the new battery when the adapter is not plugged in.

On the new battery: 0V when testing pins 1-7 in reference to 8 without charger. With charger (reference is 8):
1: 2.5
2: 2.5
3: 3.2
4: 3.2
5: 0
6: 0
7: 0


On the old battery: without charger (reference is 8):
1: 8.8
2: 8.8
3: 0
4: 0
5: 0
6: 0
7: 0

With charger (reference is 8):
1: 9.3
2: 9.34
3: 3.2
4: 3.2
5: 0
6: 0
7: 0

What exactly is the problem? Other than the 0V on the new battery without charger the other pins seem to be acting normally. Did I just get unlucky and get a dead new battery?

>>1142692
There are, or more accurately, there are ADCs with built-in amplifiers like ADS1242.
Not that it really helps you. You aren't going to measure nanovolts at any accuracy. Even microvolts are really difficult to measure with a setup like yours.

>>1142894
>with a setup like yours
What are you implying? If you are referring to the lack of fixed voltage reference, which I removed due to apparent obsolescence, I'm pretty sure it doesn't matter because either way the voltage entering the reference-voltage input on the ADC and the voltage across both resistors will be exactly the same, especially now that I've decided to remove all semiconductors along the way. In the case that a (two 18650s in parallel actually) Li-Ion has significantly more equivalent series resistance than a voltage source, I retract this statement. If you've any other ideas on how to improve my accuracy, I'm all ears, but judging from the existence of ADCs with voltage multipliers within, I guess I don't have to look for a 30W 0.5Ω resistor/shunt with low tolerance AND low temperature co-efficient. I'm still doing 6 digit.

If I can use a couple of OPA2134s I own as voltage multipliers/dividers I could just do that, but I guess their ESR is too high and I'll need specialised components instead.

>>1142884
There's a lot of circuitry in between a battery and the charger. On some laptops they dumbly pass the charge voltage to the battery and let the battery's circuitry manage charging/discharging, on others the battery's circuitry has to talk to and pass tests before the laptop will send or even try to drain it.

With that said, test pins 8-7, 8-6, 8-5 for continuity. same with 1-2, 1-3, 1-4, 3-4. I'm thinking that pins 5-8 are all grounds, 1 and 2 are both positive, and 3-4 may be a ttl signal or connected to a thermsistor.

>>1142834
Well I intended to do a MS after I got some work experience. So I could really see what I excel at and what I enjoy. At this point, my overall gpa was below average and wouldn't be competitive for grad school. I honestly don't mind settling for some bullshit job that pays shit, I figure once I get my foot in the door I can move up from there. Ultimately I want to run my own company and employ myself, but I have yet to come up with a good idea.

Anything in designing circuits is something I really enjoy. I have been reading up on power electronics on my spare time just cause circuits interest me. So hardware design, SoC design, PCB design, that sort of thing. I have entertained the idea of making my own electric motor and use them for something like a skateboard that operates on the same principle of a segway. Also entertained the idea of making wireless power. So really, anything in EE will get me going.

>>1142903
I'm implying that microvolt level DC signals are difficult to measure, even with professional instruments. Some usual reasons which make it difficult are thermal voltages, amplifier/ADC offset drifts, parasitic rectification and plain old noise.

>OPA2134
The datasheet says that your typical amplifier at room temperature can have 500µV offset voltage and that it can change 2µV per degree. They aren't well-suited for this application.

And since we are talking about an ohm meter with 6 digit resolution, at least you could include 4-wire measurement.

File: Capture.png (23KB, 533x525px) Image search: [Google]

23KB, 533x525px

>>1142790
A criteria on my last two rentals has been "proximity to Jaycar". It's right up there with proximity to exchange. And yeah, they originally ran data from position (TTL) sensors. The other anon suggested PCB track lookups and I found pic related, so looks like you're on the money with 500mA.

>>1142812
> cable loss will be identical either way.
Really? I assumed V=I/R and all of that, adjusting V would change the other parameters for a given length of conductor. Therefore, a TTL level signal would drop more than 12v 2A across a given conductor. R never changes?

Also, I assumed 1 ounce because default, I bumped it up to 2 and the width dropped to 1mm external, make of that what you will.

>>1138806
YouTube Collins Lab.

File: ohm.jpg (692KB, 2080x1560px) Image search: [Google]

692KB, 2080x1560px

>there are people that haven't read through chapter 5

also have an OP image

File: 1480406127900.png (294KB, 1240x795px) Image search: [Google]

294KB, 1240x795px

what's the best resource to learn how this antenna matching network and filter were designed? the 50 ohm load on pic is an 868mhz antenna. chip is an mcu+txrx

i'm designing an IOT stepper motor pcb with an si106x as a learning exercise.

https://www.silabs.com/documents/public/data-sheets/Si106x-8x.pdf
https://www.silabs.com/documents/public/application-notes/AN791.pdf

File: 4 wire.png (5KB, 640x400px) Image search: [Google]

5KB, 640x400px

>>1142932
>4 wire sensing
I guess the fact I had to look that up shows how much I know about what I'm doing. At least I'm learning. I'll assume that all ADCs have a near-infinite ESR, but are the circuits in pic related equivalent? Because otherwise I'm not sure what the semiconductor equivalent of an Ammeter is, and what its ESR would be. If they were equivalent then I guess calling it 4 wire sensing world be a little stupid, but that would make two of us. I'm also wondering how easy auto-ranging this thing would be.

Anyway, assuming the depicted circuits are equal both theoretically and practically, the only difference between this and my circuit is that there would be another ADC across the reference/shunt resistor. I can understand how this would help remove systematic error, but the idealisation of the shunt-resistor+voltmeter is troubling, because I don't think a 0.00000Ω ammeter that can measure minuscule and large currents has ever existed. To measure resistances higher than 100Ω with a 24b ADC and shunt resistor of 0.005Ω, I'd only get 4 digits of accurate resolution at best, and that's the same as with my setup. As I see it, 4-wire sensing only decreases error, not increases ideal accuracy.

Anyone know any good resources on analog video signals? I can barely find anything on the topic.

>>1142962
>Really? I assumed V=I/R and all of that
maybe you're confusing things. the control signals to turn on the FETs are low-current, and so not really affected by cable losses. (they may be susceptible to noise, tho, so would be better kept short or non-existant.) however, you also need to send a power line (presumably 12V) to the motors, at substantial current, and its this wire where cable losses will occur. and since the motor current is the same whichever way you wire it, then the losses are the same.

Just bought that The Art Of Electronics book - Student Edition ofcourse - ITS WORTH ITS PRICE IN GOLD.

Buy it. Really, I am pissed I never read this before.

>>1143232
>ITS WORTH ITS PRICE IN GOLD
Y-yes?

>>1143274
It is - the Student edition is 600 pages but it has so many examples. I am actually pissed that no one talks about it. Really get it from 'other means' if you doubt me. Have a read.

>>1142812
>>should I place my mosfets in the control box or as close as possible to each stepper?
>wont make any difference. the same cable will have to carry the same current the same distance, so the cable loss will be identical either way.

Yes, it will. Not as far as resistive losses are concerned, but you always want to keep the conductor to the gate of the transistor as short as possible. Lengthening it makes ringing a more pronounced effect, which, if not accounted for, can pretty easily exceed the voltage limit on the gate and destroy the transistor. At best, you need additional damping (resistance) to quell the ringing effect, which limits your switching speed and forces the transistor to spend more time in a semiconducting state.

Even worse if it's a long track next to an electrically noisy conductor, which the switched power line going to an inductive load most certainly is.

>>1142766
>I need to work out how many amps each conductor can take at 12v

Current carrying capability of a conductor is independent of voltage. And how much current your conductor can handle is dependent on how hot you can allow it to get and/or how much voltage drop is acceptable. I've run 40A through 14AWG speaker cable before, but that was in free air on an ebike, meaning the cable would get exceptionally good cooling. That same shit would melt the insulation off a wire running through some fiberglass in a wall in minutes, tops.

As pointed out, ribbon cables are usually only expected to carry signals or relatively low-power lines. I wouldn't expect more than an amp per conductor, and that's assuming decent cooling (open air or ventilated location) and only a few of the conductors are carrying power.

Only way to get an accurate idea of what the limits will be is to test it, ideally in the environment they'll ultimately provide service in, or at least a half-ass simulacrum.

>>1143232
I don't see a student edition? Just the standard 3rd edition and the "A Hands-On Lab Course".

There are some really old "student manual" versions, is that what you're referring to?

Hey /ohm/, mind taking a look at my conundrum?
>>1143306

>>1143307
Given the properietary nature of Apple, I doubt you will find a diagram. Show us the full length of the wire , whatever is on there - the earphones will ofcourse be like any other analog headphone. Just have to see how the connector fits to them

>>1143277
I'm afraid you don't quite understand. The phrase is "worth it's weight in gold" because gold has a high value per it's weight. By saying "worth it's price in gold", you are saying that a $1 worth section of that book is worth the same as $1 worth chunk of gold, which is a tautology because money has the same instantaneous value no matter what it's applied to. Unless you're implying that it's a good investment, which kind of works, but I'm calling your bluff in that case.

>>1143313
Nah mate, people do teardowns on them. I already posted there if you want the details.

>>1143320
The levels of autism in this post.

>>1143350
Thank you for replying on my behalf lol.

See, I already said - read a preview of the book online through 'some means' and then purchase it if you want. I really liked this book and I recommend it to everyone.

>>1143300
Yes I got the old Student version for 10$.

Are there any low cost methods for making custom cases for homemade electronics?

>>1143655
An old favourite here is the cigar tin, pretty rare nowadays mind you. Any rectangular tin like that should do, even a sardine tin could work in a pinch with some painting and you could use an ordinary round can if you're desperate. Nice thing about metal is it provides nice shielding for RF/audio electronics.

If you want something larger then you won't care quite so much about the mass/area of the case because the mass/volume of the electronics within scales quicker, so wood becomes an option. If you've access to carpenters tools this is a rewarding choice. If you've got an accurate tablesaw with good depth of cut or a bandsaw, you should be able to rip some timber down thin enough for a small box, I tried it with a handsaw and the few mm of gap gives it a certain amateur charm.

Small rigid plastic cases are great insulators if you might have the odd live wire loose or just want a double-insulated box, but the only boxes rigid enough tend to be the purpose-built ABS enclosures. They can be a bit of a pain to work with when it comes to screwing into them and stripping the fuck right out of the threads, but you can stand on one no biggie. They make cast aluminium versions of the same, and they're pretty skookum too. But if you've got an entire workshop full of grey bland ABS enclosures, it'll look like a greyscale snapshot of the 70s.

If you aren't too fussed about strength then you can just chuck some slices of acrylic together with a little cyanoacrylate, and you can give yourself a window if you want. Suspended PCBs are pretty trendy for vintage/minimalist electronics like Nixie clocks, and they do look pretty nice if you've got elegant standoffs and a nice base, but don't bother if your board's a clusterfuck of ICs or, god forbid, copper spaghetti.

>>1143655
Not really "custom", but I've used:
cassette tape cases
VHS cases
toothbrush cases

File: color temperature.jpg (82KB, 610x963px) Image search: [Google]

82KB, 610x963px

Does anyone know how these 12v Led light replacements work (https://www.aliexpress.com/item/q/32784145199.html)? I want to see if I can use these as cheap replacements for 8v lamps in old receivers, but it doesn't seem that there is a resistor in them.

The cheapest drop in option is 50 pcs for $20, but I don't need that much.

File: sardine can radio.png (650KB, 1265x464px) Image search: [Google]

650KB, 1265x464px

>>1143676
>even a sardine tin could work in a pinch

It used to be common.
Now, not so much.

File: bread pan radio.jpg (214KB, 800x1200px) Image search: [Google]

214KB, 800x1200px

>>1143676

I've actually used a cake pan as a chassis for a tube audio amplifier.
I remember it being a little 'flimsey' (from the thin tin) and barley supported the transformers without bending.

Way too many years ago to have pics but here is something I found similar when googling.

>>1143750
>all your electronics smell like fish

Sounds like some really fucked up dystopian future.

>>1143741
Unless they're running the LEDs in series with a 4V drop across each one (unlikely), they'll have some sort of resistor, even if it is just to drop 3V / 1V each. You can't see the other side of the board so they'll be there. I guess these are the same fitting as the 8V lamps you're replacing, but if they're in series you'll have to disassemble the thing, even if it is just to put a same-sized board in with different traces on it. Shame they don't show the underside of the board. These LEDs typically run off 3V each, so in series they'd need around 9V total without a resistor, more with, so your 2.67V probably wouldn't be enough. You could just hook a Joule Thief up to them, but that would probably be more work than redoing the board if you have etching materials on hand. I'd try scratching off the traces connecting the LEDs if they're in series and adding external wires to put them in parallel, then putting an appropriately sized resistor in. If they are in parallel, you'll just need to install a bigger resistor. The existing one will be an SMD though. And those LED things look craptacular.

>>1143114
>I'll assume that all ADCs have a near-infinite ESR
You mean input resistance? It is pretty common for 24b converters to have not-so-high input resistance, something between 100k and 10M. Better check your ADC's datasheet.

While current input ADCs with low input resistance exist, there's no need for one here. You can use your reference/current limiting resistor for sensing the measurement current. An alternative would be to replace the series resistors with current sources (you don't need to measure current, when you know it already). Or if you really want to, you can build a nearly zero-ohm current to voltage converter using an opamp.
The basic/minimum requirement for 4-wire measurement in this case would be an ADC with 2 differential inputs.

Yes, the point of 4-wire measurement is to eliminate/reduce certain kind of measurement error. It's kinda moronic to measure 100 ohm resistance to 6 digits, when the digits below 1 ohm will vary depending on contact pressure or even test lead temperature.

Is there a good, concise source for reviewing the mathematics necessary for electronics? Been about 10 years since I took calculus...

>>1143827
>arithmetic
>complex numbers
the end. you can use trig if you really want to.

File: Screen Shot 2017-03-13 at 7.57.29 PM.png (550KB, 1874x984px) Image search: [Google]

550KB, 1874x984px

>>1143813
>You mean input resistance?
Yes.

>between 100k and 10M
That becomes a problem when I'm measuring high resistances.

I know 6 figures is overkill for pretty much everything, I'll only go for it if it's not much more expensive to build. Jesus this thing is becoming enough of a complicated project i might as well run it with Nixies or VFDs just for the challenge factor. RIP power consumption but it would look better than any LCD.

>While current input ADCs with low input resistance exist
I guess that means the ammeter in a 4-wire circuit doesn't equate to a voltage ADC across a reference resistor. Either way it shouldn't make a difference. You've at least reminded me that I don't have to buy a bunch of ADCs because virtually none come in single channel anymore. I'll just have to find one with differential inputs. But it would be more accurate to be able to have 3 ADCs, one across each resistor and one across both, and then multiplex them to figure out the impact that each ADC has on the measurement, but this would be fucking overkill.

>replace the series resistors with current sources
I've already ruled out current sources because I want to measure resistances in the MΩ range. Either way I have to measure something, or two things with 4-wire sensing.

Now my only remaining problems appear to be how to calculate the correct resistance value taking into account the 4-wire sensing, (which I assume is done to find the effect of wire+voltage reference series resistance), and using voltage multipliers and dividers as to not overload the two resistors with current. As it stands I'll be able to measure 0.013Ω with a 0.47Ω reference resistor to six digits, and 130µΩ with the same reference to four digits, but that's assuming I can push 10A through them. I'm aiming for a 100th of this current and I hope I can get multipliers with this kind of accuracy and multiplication/division factor. Pic related is accuracy spreadsheet assuming 24 bit ADC without 4-wire sensing.

File: 20170311_164254.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

hey, got a simple and easy project for once.
my beloved logitech g930 headset's battery crapped out after a solid 5 years.

i found a cheap replacement on ebay from germany. but it died after one day, got a new one after complaining but that also died.
i suspect the NTC in the battery isn't the same and the headset refuses to charge it.

after seeing a cool mod on reddit i decided to use 18650 batteries from my old laptop to power it.
a single cell has over 4 times the capacity of the original battery, and I have 6 of them.

disregarding the chargecircuit completely, ill just charge them out of headset hand have a batteryholder on the side.
so i can charge and replace as needed with no downtime.

I'm waiting for single batterholders aswell as charge circuits, so for now i just charged a few with constant current on y bench supply and temporarily hotglued a double batteryholder in place.

when i get the single holder I'll mount it more nicly and permanently, butting a slot in the cover.
works great so far, after 2 days of testing.

File: 20170311_163954.jpg (3MB, 3264x2448px) Image search: [Google]

3MB, 3264x2448px

>>1143949
ad hoc constant current charging.

I also have a big RC charger I could use, come to think of it.

>bigclive not in the OP
Shame

Anyone have experience with these type of cheap chinese raspberry pi displays? Are the viewing angles ok? I want to make a smart picture frame/alarm clock with my raspberry pi zero w.

http://www.ebay.com/itm/7-inch-LCD-Screen-Display-Monitor-for-Raspberry-Pi-Driver-Board-HDMI-VGA-2AV-/171280831162?hash=item27e12212ba:g:M30AAOSwFV9Xy5VK

Its a bitch that small screen like this can cost $30 - $150 but you can get a 19 inch refurbished IPS screen from HP for $75.

File: 20170313_155825.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

>>1143949
got 2 more to burcher

File: 20170313_161937.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

>>1144067
one got a but banged up so i put it aside.

File: application.png (15KB, 405x140px) Image search: [Google]

15KB, 405x140px

Anybody ever use one of these CUI miniature power supplies?
http://www.cui.com/product/resource/pbo-3.pdf
I'm having a hell of a time getting mine to work properly. The output voltage is strangely sensitive to the value of the filter capacitors. Putting a small (<1.5mA) load on it causes large ripple even with the filter built as per the application circuit. A larger load (33mA) causes a much smoother output, but it's not at 3.3v as per spec, it's only 2.5v. No load, I get 2.0v with no significant ripple.
Nothing I do can get the output above 2.5v.
Very strange behavior. My circuit is almost exactly this circuit, with all "required components" (C1, C2, C3, L1) present.

Where can I find AvE level explanations for electric shit?

I notice him and bigclive are missing.

Quick question, does anyone here have a little knowledge of Nixie tubes? I'm just wondering what 0-9 tubes an IN-7, IN-15, or IN-19 would go with. I'm pretty sure they go with the IN-4, IN-12, and IN-18 respectively, but I'm not quite sure. Also 18s are too damn expensive.

File: raspi.png (918KB, 999x542px) Image search: [Google]

918KB, 999x542px

Should this actually power a pi?

I think I might have fucked mine by messing with the micro usb port back contacts and can't get it to power, dosen't want to power on when trying with the gpio pins. There was one hairy moment but no obvious damage on the board.

>>1144196
google around and you'll find one of the sites with data on them all. then compare sizes.

>>1144198

Looks fine...

>>1144198
The magic smoke is leaking out

>>1144198
you're just breaking it in, if your lights don't dim you're probably not hot enough to cure the silicon.

File: led.jpg (125KB, 1000x1000px) Image search: [Google]

125KB, 1000x1000px

>>1143802
>You can't see the other side of the board so they'll be there
They do show the back (or what is allegedly the back) of the board if you scroll down to the description. All it has is a model number of some sort.

>>1144321
Then I haven't a clue, the resistor(s) will be obscured by the metal shells. After running the maths, it's quite likely that they'll be in series, because the power dissipated by the resistor will be 3* less than if they were in parallel. This means there very likely won't be enough voltage drop from your 8V to light the LEDs in their current configuration, even without the resistor at all.

Having said that, it's not like buying them won't get you 30 SMD LEDs along with the metal shells. If you already have the capacity to do SMD soldering and PCB etching, it will only cost you a dollar or two in trials and materials to make a replacement board to re-solder the LEDs on to in parallel with the correct resistor.

Do speakers just receive pulses of electrics to make them go?

Could these pulses power a small motor?

I'm thinking a wireless bluetooth vibrator that can be controlled by playing certain sounds through a phone it's connected to.

Fuck paying £50< for actual smart sextoys.

>>1144467
I'll have to ask - are you a boy or a grill ?

File: 1468603572972.jpg (41KB, 600x450px) Image search: [Google]

41KB, 600x450px

>>1144490
I can pretend to be a girl if that's what you are into

>>1144499
Talk dirty to me )))

Hahah, jokes aside, you can use a PWM driven stepper motor work with pulses so to speak. But otherwise, non-PWM DC motors dont work like that.

Its too much hassle to do just to fuck yourself in the ass anyway.

File: SKU040342-003[1].jpg (25KB, 361x361px) Image search: [Google]

25KB, 361x361px

>>1144502
You don't fuck yourself with a vibrator. It vibrates.

>>1144467
You would need a transducer for that

>>1144467
You could try using a servo. By putting audio (alternating current) through a motor, you'd just make it rotate from side to side, probably ineffectively. Vibration motors have an offset weight on the end that makes the vibration. By putting AC into a servo it would go in and out instead of rotating, making a vibration. You'd need a dirty amplifier and batteries, because the power from your headphone jack wouldn't be enough. Ideally you'd be outputting an AC square wave to vibrate the servo with maximum efficiency, but I don't think frequency generator apps can do that. It should be doable in the hardware however.

Also servos make axial vibration rather than radial, so it won't feel the same.

File: rotary-servo-anatomy1.png (61KB, 700x332px) Image search: [Google]

61KB, 700x332px

>>1144734
A server is just a motor with feedback.
Do you mean solenoid?

>>1144737
I guess so. I assumed servos were always linear and that solenoid was just a term for a linear coil, which it is, but I guess it refers to a linear motor too.

>>1144734
The idea is to use the already existing vibrator motor sealed inside the bullet.

File: Screen Shot 2017-03-15 at 2.26.25 PM.png (1MB, 1028x772px) Image search: [Google]

1MB, 1028x772px

>>1144767
>bullet
What is this referring to? Also for using an existing vibration motor you should just put DC up to it with PWM for speed control. This looks vaguely respectable:
https://www.aliexpress.com/item/DC-New-1-8V-3V-5V-6V-12V-2A-Low-Voltage-Motor-Speed-Controller-PWM-1803B/32388277945.html

It takes up a little room so you could always try making your own PCB with a different (slimmer?) form-factor, and getting your own components, but to be fair it will run you ten times the price. If you like I could trace the circuit on that aliexpress pcb, though it may not be an ideal circuit in the first place. Also if your pooter-blender runs more than 2A then you might have to look for a more powerful speed controller, along with a new prostate.

File: motor.png (10KB, 1034x547px) Image search: [Google]

10KB, 1034x547px

So I'm making a little rc car that can go either forward or backwards, each direction powered by a switch. Where do I put diodes/whatever to prevent it from shorting itself?

File: DC-motor-reversing-switch-schematic-wiring-diagram-285x275.jpg (41KB, 285x275px) Image search: [Google]

41KB, 285x275px

>>1144862

you dont need no steenkin' diodes.
just a DPDT switch.

File: .jpg (10KB, 300x300px) Image search: [Google]

10KB, 300x300px

>Battery charging
I have 2 pic related and 1 charger.
Charger is 9v 300ma and instructions are to charge for 6 hours.

I have a box of 9v wall warts and spare tamiya connectors.
If I make another charger from one does it matter if it's not 300ma, probably more?

I don't think there's any electronics in the battery or charger safety wise, just straight DC as long as it's powered on.

>>1144879
Do they come in the form of two buttons? So I have two on states and an off state as opposed to just two on states?

File: schematic.jpg (250KB, 1232x768px) Image search: [Google]

250KB, 1232x768px

I want to install LEDs in this old receiver to replace the lights. The lights in question are on the bottom right of the schematic. Would it just be as simple as installing a rectifier with a smoothing capacitor before the lights?