/ohm/ - Electronics general: Heathkit Edition

What kind of processes do you guys use to make PCBs?

I want to make a guitar pedal, and think it would be great to make the PCB too.

I have a 3d printer. Should I mount a marking pen on it and plot it out? Then use acid to dissolve the traces?

Should I mount a dremel to it and mill the traces out?

Is there a better way?

File: Free-Delivery-font-b-Relay-b-font-SMH-112DM-dedicated-font-b-microwave-b-font-font.jpg (62KB, 600x400px) Image search: [Google]

62KB, 600x400px

>>1013204
Once a relay starts sticking, is it toast?
Maybe my problem isnt the relay at all

>microwave makes loud popping sound during cooking
>timer hits 0 and it beeps
>light and turntable do not stop
>open door, everything stops, close door it starts back up by itself again.

In the repair manual it says that problem is caused by relay 2 not closing and to replace the control board.
I cant find a board to replace it if I want so im gonna try and fix it.

I unsoldered both relays, both normally open relays
Relay 2 works as it should when I bench test it
Relay 1 was stuck closed when I initially tested it with just multimeter. It also seemed like there was something loose jiggling inside.
When I put power to relay 1, it made kind of a weird noise, I checked and the poles were closed.
After I took away the power away, the poles opened up. It works perfectly now, bench testing it. It also isnt jiggling inside

The control board isnt burnt smelling. There are no obvious signs of heating, no obvious signs of a failed component.

Does that sound like a failed relay?
If I put this relay back in, will it just stick again?
It was a huge pain in the ass to get the board out. And I cant find anywhere but Aliexpress to get the relay, suck to wait 30 days for it.

>>1013210
Using transfer paper and a laser printer sounds a whole lot easier than that.

>>1013210
Photoresist master race

>>1013223

If you have a scope, probe the low voltage coil and see if it is being triggered properly

Also relays aren't very specific. If you need to replace it, just get the closest one you can find.

>>1013210

Consider just doing it on perfboard.

File: IMG_1168.jpg (396KB, 960x1280px) Image search: [Google]

396KB, 960x1280px

This is my micromouse I'm working on. Could have done it as a team at school but no one had any clue what they were doing so I'm doing it myself at home. I'm not using steppers or servos like a lot of people do, I used normal DC motors. Problem with that however is that I don't have an easy way of telling the robot how far it's traveled. So there's two ways of measuring distance traveled I'm thinking of: First is the good old encoder method though I'd rather not add more sensors but I could just move one of the extra wall sensors as I have two on the back and two on the front. Second option, and I don't know that this would actually work in any sort of reliable way, would be to simply measure the distance the robot travels going in a straight line during a certain time interval. Problem with that is as the battery drains the robot slows down, at least the way my power supply is set up now. I have it running off two 9v batteries which are regulated to 5v for the microcontroller and go straight into the motor driver microcontroller. There's probably a more efficient way to power it but I'm sort of flying by the seat of my pants here and changing things as I test.

Anyone every build a micromouse before? I'm building mine from scratch without really copying any specific design.

>tfw have lots of components and no projects

>>1013548
>Have components and some ideas but my electronics and programing skills are severely lacking.

So I too sit on a mountain of components.

>>1013552
Sucks mayne

I scraped an ac brushless motor from an old tool, ive got a project in mind but im not sure how to get it to be variable speed or how i could check how many amps it draws or its other stats. could i just get a cheap motor control unit from ebay and only hook up 2 of its wires? it has 3.

>>1013522
Open loop control (calculating motion from the motor drive signal) isn't going to work. Even sensing wheel rotation is unreliable because wheels can slip.

You need to measure properties which are directly related to the mouse' position. Measuring any kind of derivative (motor signal, encoder pulses) and integrating is a fool's errand.

There's a reason that micromouse mazes are specified to have a black floor and walls with white sides and red tops.

>>1013557
> im not sure how to get it to be variable speed
If it's designed to run directly from a 50/60 Hz supply (either mains or low-voltage), you'll probably need a variable frequency drive (VFD). Most AC motors are synchronous, i.e. their speed is dictated by the AC frequency.

> or how i could check how many amps it draws or its other stats.
Look up its part number. If you can't find details, throw it away and buy something with a data sheet.

> could i just get a cheap motor control unit from ebay
"Motor control unit" doesn't tell you shit. There are a few dozen different types of motor and a controller will be designed for exactly one of them.

Unless you are a motor otaku, you need a known motor so that you can ask someone if a specific controller will worth for that specific motor.

Is there any special considerations that should be made when using a dc drive as a variable dc power supply?

I know that trying to use ac drives as an inverter causes alot of issues with fooling the input circuits to the dc bus that 3 phase is actually present when your charging with a single phase, and that unless your very careful with your RLC characteristics the pwm signal you get will look nothing like a pure sine wave etc.

But it seems like dc drives would be much more simple, the drive simply presents a voltage to the armature windings, and the current is a function of ohms law, and as long as the fault circuits of the drive dont fine the current too far out of whack for the 'hp' of the motor that isnt really there it should just poop out vdc and adc as you would expect. Am I wrong?

I am hoping to use some kbic drives, probably a kbic 200, to power some high power led arrays, and just use the 0-10v input on them to control intensity. I think by de linearizing and putting a resistive element into the circuit I can get over 6v of useful resolution out of it.

>>1013569
guess il not use it then, know of any household items or tools that i can scrap to use instead? want to make it cheap.

>>1013572
High-power DC motor drives are normally PWM and typically rely upon load inductance for filtering.

If you want to drive LEDs, you want a constant-current regulator, not constant-voltage.

>>1013592
Find something which actually operates at variable speed (e.g. a food mixer) and use both the motor and the speed controller.

>>1013572

0-180VDC @ 12 amps with 0-10v control with 3 trim pots for setting response curve characteristics and integral current min/max settings in a ul listed package...all for $40-70...seems too good to be true

>>1013598

These drives are scr and put out fixed voltage. I just need them to power a dc bus that I am power 5 sub boards for power distribution using a current mirror setup and fusing along with a small resistive element for each parallel string of leds

>>1013603
> These drives are scr and put out fixed voltage.

An SCR (thyristor) is an on-off device; it can't supply any voltage other than its input voltage ("on" state) or 0V ("off" state). The closest it can get to an intermediate voltage is PWM.

It might include low-pass filtering to smooth the PWM to (almost) DC, but given that it's sold specifically as a motor drive, there isn't much reason for it to do so (motors naturally have large inductance and thus don't care about PWM versus "real" DC), and at least one reason not to (unnecessary extra cost).

>>1013618

Thanks, I'll have to fire up the scope and test my assumptions in light of this. It seems to be clicking now. I think I can work this out with either some fuck huge capacitors or some inductance seeing as how I'm just trying to keep a dc bus charged that effectively feeds a form of power supplies anyway.

>>1013562
Well then what do people do?

>>1013631
Optical sensing. Hence the black/white/red.

Traditionally it was just LED/photodiode point sensing; nowadays I suspect that it's more likely to be CCD cameras and computer vision.

I assume that they also sense wheel speed (and MEMS accelerometers/gyros are dirt cheap nowadays), but you still need some form of direct position sensing to prevent integration drift.

Sup, /ohm/
Small question: I need an IC that acts as a decade counter (enables one of N pins in sequence), but with more output pins.
Does this exist? Or am I better off just using multiples of other ICs?

>>1013703

I think they make MCUs with small processors and lots of I/O

>>1013204

Is their any disadvantage to using a voltage regulator in an application where it will get switched constantly? e.g. in a stepper motor driver on the output of an Hbridge

how do i transmit a simple binary signal wirelessly about 5 meters in a house as cheaply and simply as possible?

File: wireless_krmwerll.jpg (151KB, 800x1272px) Image search: [Google]

151KB, 800x1272px

>>1013727
top image:
The simplest way is to use a plain RF send and receive board pair.
This just transmits plain binary pulses, there is no error checking involved. Your MCU code must do that, if you desire it.
This only transmits from one board (the bigger one) and receives on the other, it is not a 2-way setup.
There is also no choice of channels so interference can ruin your plans.
The 315mhz ones is for euroland. US people can legally use either, but the 433 should have slightly better range/less problems with interference.


The better way is to use a "NRF24L01 2.4GHz board". The cost is not much more and it works way better.
These are more complicated to use, but have a lot of advantages-
--2.4ghz, so more range/less interference than the plain ones above
--the chip does error-checking and automatically resending failed packets
--multiple channels, with some above the wifi range
--each module supports 2-way communication

middle image:
The NRF24L01 boards with a PCB antenna cost about a buck and will very likely do what you want.

bottom image:
There is ones with a stub antenna for around $3.00, if you want to make sure to get the best possible results.

File: LVT.jpg (59KB, 799x392px) Image search: [Google]

59KB, 799x392px

So i have this small transformer that i found while dumpster diving and i was curious about the voltage output on it. I found a service manual for a microwave that uses the transformer. It is not the microwave the transformer came out of as the control board that i found with it does not match up with the service manual. Here is the image from the service manual detailing the test procedure for the transformer.

File: Schematic.png (329KB, 834x1197px) Image search: [Google]

329KB, 834x1197px

>>1013795
The transformer is in the upper left hand corner, and im pretty sure it steps down from 120 volts, but im not sure to what exactly. The way it shows it hooked up it almost looks like a choke, but it has 3 leads on it.

>>1013703
You can construct a decade-type (1-of-N) counter from a binary counter and a decoder (or multiple decoders).

You can also use a ring counter (a shift register with the output connected to the input). You can chain multiple shift registers to get the desired number of bits. The main issue with a ring counter is the need to ensure that the initial state has exactly one bit set.

Decoder ICs are typically limited to 16 outputs, while shift register ICs are typically limited to 8 outputs.

If the number of outputs is large, you'd try to use a matrix arrangement where possible. That lets you switch MxN devices with only M+N outputs.

>>1013800
Hook it up to the mains and measure the secondary voltage(s).

A centre-tapped transformer is useful if you need a balanced supply (i.e. positive and negative rails as well as a ground rail). Or you can just ignore one of the taps it if you only need positive and ground.

File: This shit.jpg (122KB, 684x692px) Image search: [Google]

122KB, 684x692px

>>1013808
Is it center tapped? that what i wasn't sure about because i don't have 2 other leads for the mains voltage, in the schematic the Primary is hooked up and the secondary goes to nothing.Heres a picture of the actual part with scale.The part number does not give great results in google, mostly just people selling it.

What's up gurus. I'm about to build a guitar amp with my buddy using the tiny giant amp board from musicpcb.com. I'm going to build it into a more traditional amp head and cabinet.

Anyways, a couple questions. The power supply for the board needs to be between 15 and 20vdc capable of delivering 4 amps. I want to house it all in the head, so I'm looking at using a transformer, bridge rectifier or 4 diodes, and a capacitor for that, with the same style jack that desktop computers/amps use(used to know the name, I'll be able to figure that out). Does that sound like a solid idea? Also, any idea what I could pull the transformer out of to save a few bucks. If not I found a suitable one that delivers 18v at up to 10 amps, but it's $50. Money's not an issue overall, just nice not to spend it if I don't have to.

Second question, I'm thinking about putting a tube preamp in there if I can find a novice level schematic. The couple I've seen so far are in the 60 volt range, so would it work if I used 2 different transformers in there to power separate boards from the same mains hookup? I would imagine it's fine, but then should keep the grounds on the board seperate? Is this idea even feasible?

I'm not too experienced with electronics like this, I've built small circuits before and I'm good with an iron. Kind of using this as an introduction to amp construction, I've only done amp mods in the past.

>>1013800
I'm fairly sure that the 6010W2L003D isn't the transformer on the PCB, but the one labelled "Lighting LVT" in the middle of the schematic.

File: mousy.jpg (64KB, 960x420px) Image search: [Google]

64KB, 960x420px

>>1013645
i don't think most are using computer vision. Most of them seem to only be using infrared distance/obstacle sensors.

File: Bumperino.png (62KB, 858x893px) Image search: [Google]

62KB, 858x893px

>>1013831
Holy hell right you are. And there's the voltage labeled on it.Sorry for the stupid just read the schematic wrong. Thanks for the quick response,and in the spirit of contribution to the thread heres this nice ohms and watts law thing that was in the back of the service manual.

>>1013804
>you'd try to use a matrix arrangement
That's exactly what I was intending to use the decade counter for, have it act as the input to a matrix.
I appreciate the advice, but I doubt I'll ever incorporate it into anything. Sounds fun, but a bit over my head at the moment. I'll look up all the unknown terms in your post, though.
Thanks!

>>1013829
The transformer for a 15V 4A linear supply is going to be massive. And expensive. I would just scrounge a laptop adapter and filter the shit out of it if necessary.

File: Base_Diags.jpg (149KB, 1500x1048px) Image search: [Google]

149KB, 1500x1048px

>>1013879
Theres always these guys. They will wind whatever the hell you want and they are damn good transformers.

File: Polarity-rectangular-chips.jpg (272KB, 960x556px) Image search: [Google]

272KB, 960x556px

>>1013883
It cut out my link but look up mercury magnetics.
They will even rewind your transformers to specs of your choosing if you find something close.

so VCOs often use tuning diodes in order to change the resonance frequency of the tank circuit via a control voltage.

however since those diodes mostly have a capacity in the pF range (afaik) they cannot be used for oscillators with a frequency around lets say 1 MHz, right?

how would you change the frequency via a control voltage in that range (kHz, low MHz)?

>>1013879
That's perfect for this. Thank you.

>>1013884
I've modded a few Jcm 800s and 900s where the owners got transformers from mercury. Seemed really solid, better than the OEM ones. I wasn't aware they did customs, I assumed they did drop in replacements. I'll definitely look into them for my next(much more serious) project. Thank you.


Anyone interested, I decided that a tube pre wouldn't be cost or labor effective, and that my buddy wouldn't be able to appreciate it since it's his first amp. I'm throwing a fender style eq in the front, followed by a clean boost and a ts808 clipping and make up circuit separated by a dpdt, into the amp pcb. I'm going to utilize an 11v tap that's present in the amp. I'm going to set it up so that at any point one of the boards can be pulled and replaced with one that serves a similar purpose, say swapping the ts808 style board into one based off a big muff.

File: vco.jpg (54KB, 1284x496px) Image search: [Google]

54KB, 1284x496px

>>1013891
Varicaps meant for AM radios can go up to 500pF or so, which is suitable for 1MHz LC oscillators.
If you need to go lower, then many VCOs are built around the idea of converting the control voltage to current and using that current to charge/discharge a capacitor. Pic related is a well-known circuit, but there are many others. You can also use dividers after high frequency VCOs to get arbitrarily low frequencies.

>>1013891
For a given capacitance, you can lower the frequency by using a larger inductor.

That's probably still not feasible if you want kilohertz frequencies. In that case, you'd typically just use a different type of oscillator and (if necessary) use a low-pass filter to get a sine wave out of it.

A square-wave VCO is fairly straightforward at any frequency, and so long as you don't need more than a 3:1 frequency range, a low-pass filter will remove everything except the fundamental (a square wave only has odd harmonics).

Another option is an L-C type oscillator with L replaced by a gyrator using a FET as a voltage-controlled resistor. This was fairly typical for VCOs in audio synthesisers.

why do ferrite core manufacturers not want me to purchase their products? epcos is the only one to sell through digikey or mouser, which is great if they stock what you want.

>>1013891
For what it's worth, the varactors that are available today are a lot smaller capacitance-wise than the varactors they used to make.

>>1013948
I suspect that as the alternatives have become practical at higher frequencies, they've tended to supplant varactor-based oscillators at those frequencies.

I'm ordering a bunch of stuff and I feel like throwing in some logic level MOSFETs.
Any recomendations with a broad usability?

>>1013223
Replace the relay, it's sometimes possible to open and clean them but usually not worththe effort. Relays are a nothing part, a couple quid. Just buy the correct type and pinout and you'll be fine.

File: 1462104961151.jpg (8KB, 300x289px) Image search: [Google]

8KB, 300x289px

hey,

I don't know much about electronics.
I have a led tube and a cable that connects to it on one side. when i plug th cable into an outlet , the tube lightens up.

I would love build a lamp with this setup and recreate this cable, because it looks really shitty.

whats inside the cable. how do I do this ?

>>1014125
wow that parrot brings me back used to love the lego pirates

>>1013295
>Swifts
you'll never be master race

If I am designing a digital circuit, is it wrong to just think of a transistor as a switch, and ignore things like capacitance?

Im planning on building a board with the TB6560AHQ

The problem is my motors are 3.3v 2A

Can I reduce the voltage on the output of the chip to be ideal for my motors while still retaining the signal?

>>1013548
>tfw have lots of ideas but no tools and no money for them.

>>1014168

better to ask the datasheet than to ask random strangers, coz the answer is gonna be way more reliable. from what i can see, the max motor voltage is around 40V but there's no minimum, so you can use 3.3V

>>1014183

Might as well just test different voltage levels, then. I didn't realize that the minimum not being documented could mean "it has no minimum"

Another problem I'm having is I don't understand "CLK" and "OSC"

What is the point of OSC? does CLK have to be a fixed frequency or does it just trigger the next step on a rising edge?

>>1014158
Depends. If you need any speed, then you usually can't ignore it. Also, if you're driving big mosfets and particularly if you're using them to switch inductive loads, then you might need to take the capacitances into account even at low speeds.

>>1014202
OSC seems to be an oscillator for running the internal PWM. CLK is for your use, to control stepping.

>>1014208

I don't understand the part of the datasheet about OSC. how does the external capacitor effect the actual output?

>>1014168
> Im planning on building a board with the TB6560AHQ
> The problem is my motors are 3.3v 2A
The maximum output current is set by the resistors on the N{FA] and N[FB] pins. The output voltage is dictated by VM[A] and VM{B}, and the minimum is 4.5V. But it doesn't matter if this is above the motor's rated voltage because the IC performs current limiting. The rated voltage assumes an unregulated drive.

>>1014202
> Another problem I'm having is I don't understand "CLK" and "OSC"
CLK determines the stepping rate. OSC is the PWM oscillator.

>>1014212
See pages 16-25. f[chop] is the PWM (chopper) oscillator, whose frequency is determined by the capacitor connected to the OSC pin.

Basically, the IC determines the desired current at any given point in time. Because a motor winding is an inductive load, the current cannot change instantly. The IC applies either forward voltage (charge mode), zero voltage (shorting the windings, slow decay mode) or reverse voltage (fast decay mode) to increase, maintain or decrease the winding current. f[chop] controls how fast it can switch between modes. A higher frequency will reduce ripple (as it can more accurately match the actual current to the desired current) but increase EMI and reduce efficiency (due to switching losses and having to drive any capacitative component of the load).

>>1013903
>>1013906
>>1013948
thx for the input

Moving on from simple circuits and need some help in clarifying stuff:

So a Microcontroller is a small processing unit containing a processor, memory for storing variables etc.

Is the major difference between that and other chips that the microcontroller has a lot more flexibility?

Also, what's the difference between a micro-controller and an arduino board? Given that the arduino has a micro-controller on it, is it used for prototyping, with the microcontroller being used in the final step?

>>1014241

> Is the major difference between that and other chips that the microcontroller has a lot more flexibility?

No, not really. The difference is that a microcontroller is a type of computer while other chips are whatever they are.

> Also, what's the difference between a micro-controller and an arduino board?

an arduino is a board that uses a microcontroller.

Really all a microcontroller is is a CPU like what your computer uses with more of what it needs to run on 1 single chip. This makes them cheaper, smaller, use less power, etc.

>>1014241
>Also, what's the difference between a micro-controller and an arduino board?

Same difference as the engine and a car. Sure you can use the engine on its own if you know what you're doing, but you might want to have all the conveniences the rest of the car provides.

>with the microcontroller being used in the final step?

Nobody prevents you from sticking a full arduino in a product you're shipping, but if you have capability to use a raw chip instead, you should probably do that.

>>1014241
> So a Microcontroller is a small processing unit containing a processor, memory for storing variables etc.
A microcontroller (uC) is a CPU with RAM and basic I/O devices on a single chip. Compared to a "system-on-a-chip" (SoC), a uC has less memory and more basic I/O. E.g. a SoC may include PC-level I/O such as ethernet, WiFi and/or video, whereas a uC's I/O capabilities tend to be lower-level: GPIO, serial, I2C/SPI, CANbus; if it includes USB, it's more likely to be as a USB device rather than a USB host.

> Also, what's the difference between a micro-controller and an arduino board?
An arduino board contains the microcontroller plus some basic support chips (mainly to support programming directly from a PC via serial port or USB), crystal oscillator, etc. And it's a board rather than a chip. Also, the microcontroller is pre-programmed with a boot loader.

File: Screen Shot 2016-06-26 at 5.54.21 PM.png (545KB, 943x617px) Image search: [Google]

545KB, 943x617px

If I take out S2, VR 5K, and S3 and hook this top circuit up to my own 48VDC power supply will the circuit run?

There is very little info on the component im trying to get working.

File: Screen Shot 2016-06-26 at 5.53.23 PM.png (898KB, 935x705px) Image search: [Google]

898KB, 935x705px

>>1014293
This is said component.

Only thing I forsee going wrong is the crystal may require and 'excitation voltage' (voltage spike) before it begins running. It's a piezo.

I gave up on trying this a while ago (it's why i got into electronics) and this schematic came up and looks solid. Want some more experienced anons comments because im still in noob territory.

File: moswut_44ui2h5iu3.jpg (85KB, 1190x486px) Image search: [Google]

85KB, 1190x486px

>>1014089
>I'm ordering a bunch of stuff and I feel like throwing in some logic level MOSFETs.
>Any recomendations with a broad usability?
I usually keep some IRLZ34 mosfets around, tho I have not yet done anything where I had to be really picky about the details.
www.vishay.com/docs/91327/91327.pdf

Pic related: they sell these little "arduino MOSFET modules" in China-land dirt cheap,,,, but they put a crap mosfet on them(?).
They always use an IRF520 or IRF540, and both need ten volts to turn on fully. At 5v you get voltage, but only tiny amounts of current....
Yea I know about gate driver chips. I don't want none unless I gotta.

I buy those boards and cut the IRF520/540 off and put an IRLZ34 on instead.

>>1014294
I dunno very much about piezos in general,,,, but piezo humidifiers are usually run so hot that they must be submerged in water to avoid burning the piezo element.
This is the reason that they always have a water level switch to shut them off when the water runs low.

>>1014587
Yes they do, I touched one running and got a burning sensation within a second

Doesn't scorch or burn your skin I find, but I heard the physical vibrations more kill the skin upon contact

What exactly is verilog?

When chips are designed, do they just write them like a program and generate a chip from that?

>>1014636
> What exactly is verilog?
A hardware description language.

> When chips are designed, do they just write them like a program and generate a chip from that?
Sort of. A "first draft" version can be generated automatically, but there will be a fair amount of manual intervention.

Hello,

I am looking for a conductive ink whose resistance does not change over time once it has dried. Has anyone heard of anything alike ? It would be perfect if the ink is carbon based.

Sup /ohm/. I am having some trouble etching my PCB's. When using the toner transfer method, how do you guys prevent the paper from moving while ironing it? I always get smudged out transfers. Also, other tips/techniques on ironing the paper are very welcome. Thanks

>>1014665
http://www.instructables.com/id/Conductive-Paint/

>>1014714
Could you just superglue two edges?

Masking tape and electrical tape is also a thing.

> tfw the IC you want only comes in a tiny package for ants

https://www.youtube.com/watch?v=FKqiEw9MbQo

>>1013204

So I took apart a 2.4 ghz wifi USB thing

it had a plastic antennae with no wire going through it.

Is it possible it's just fake? why would they do that?

File: fake antenna.png (222KB, 624x249px) Image search: [Google]

222KB, 624x249px

>>1015308
Like this?
The antenna is fake.
The golden trace is the atual antenna.

>>1015310

That's pretty funny. The weird part is mine isn't a knock off, it's name brand Linksys

>>1015311
Well it will still work for it's intended purpose.
Sometimes they do things like that in product design.
They think people are more likely to buy it if it has an antenna.

File: SKU124662a.jpg (113KB, 600x600px) Image search: [Google]

113KB, 600x600px

Hi again, /ohm/. Want some more dumb questions?
I want to drive a 12x4 matrix of white LEDs. All I have is a tiny Arduino Pro Micro with not enough pins to do the job.
So I found this LED driver, MAX7221 (picrelated). Arduino Pro Micro will be controlling it and it seems to be able to do what I want in theory, but I'm not sure if it will have enough power to drive all LEDs close to max brightness (even though it's apparently lights only one row at a time and switches between rows very fast).
The LEDs I'll be using have 3.0-3.4 forward voltage and 20 mA max current. Frankly, I have no idea what these numbers mean.
But I do know that I'll be using power from a USB cable from my computer, which is 5V and 300-400 mA.

So, my question: is that LED controller going to have problems having all LEDs on at the same time?
If yes, how can I fix that?

>>1015342

do the math, nigger. 48 LEDs at 20mA is 960mA. so your power source cant even handle that much. but if you use a diff source, like a 1A phone charger, then the chip itself wont be able to provide 20mA coz the max rated source current is 40mA, but it has a 1/8 duty cycle so the avg current can only go as high as 5mA.
so, dont expect max brightness unless you add a bunch of transistors. still 5mA is okay if the LEDs are efficient.

File: 1325254629725.png (7KB, 299x276px) Image search: [Google]

7KB, 299x276px

>>1015342
>I want to drive a 12x4 matrix of white LEDs. All I have is a tiny Arduino Pro Micro with not enough pins to do the job.
>So I found this LED driver, MAX7221 (picrelated). Arduino Pro Micro will be controlling it and it seems to be able to do what I want in theory, but I'm not sure if it will have enough power to drive all LEDs close to max brightness
The MAX7221 chip is made to drive an 8 x 8 matrix of LEDs. You can use less than 8 rows and 8 LEDs per row, but you can't use more rows or LEDs per row than that.
You only want to use 12 x 4 LEDs, which is 48 LEDs.
48 ÷ 8 = 6 ,,,, So to space the LEDs out over 8 rows equally, you only need to drive 6 LEDs per row.
That may not be the easiest way to arrange them logically, but it's the best way to spread them out for power distribution and you can't have 12 LEDs per row anyway.

>So, my question: is that LED controller going to have problems having all LEDs on at the same time?
Probably not, since you will have only 6 LEDs per row.
>If yes, how can I fix that?
You can draw power from multiple powered USB ports (the ones on the mobo, or ones on a powered hub) but you can't combine them again,,,, and the MAX chip is what needs the power, because it is what operates the LEDs.

The easiest solution here is that you can also write your code so that only one LED lights up at once.
Overall the display appears only a tiny bit dimmer. Camera pictures can catch it flickering but your eyes won't see it at all.
If you are worried about this issue, you should just write it to run one LED at once.

>>1015367
>The MAX7221 chip is made to drive an 8 x 8 matrix of LEDs. You can use less than 8 rows and 8 LEDs per row, but you can't use more rows or LEDs per row than that.
Welp,,, you kinda CAN do this, if you used more than one MAX chip....
You can daisy-chain multiple MAX chips together to make the whole thing function as one logical array, even tho it spans multiple chips. So you still get 8 rows, but the rows can be way longer than just 8 LEDS.
It would be a bit wasteful to add a second chip tho when you haven't used all the outputs on the first one.

>>1015367
>8 rows equally, you only need to drive 6 LEDs per row.
Yeah, I know. It blew my mind when I found out you can arrange matrix however you want, if you are willing to do some ugly stuff with the wires.

>write your code so that only one LED lights up at once.
I re-read some forums and I'm pretty sure that's how it actually happens in the chip. I think that's why it needs only one resistor for the whole matrix.

>You can draw power from multiple powered USB ports
Yeah I can do a lot of things, but then you look at gaming keyboards that have 100+ LEDs running full blast off of one normal USB cable. That means it's possible and might be achievable for a noob like me.

File: maxresdefault.jpg (81KB, 1280x768px) Image search: [Google]

81KB, 1280x768px

I learned basic circuit analysis from a standard EE textbook, then bought these to learn more about working with electronics in practice, only to realize that proper lab tools are going to cost me a fucking fortune.

>$400 for a decent scope
>$400 function generator
>$400 power supply

and the components used for the analog labs total $250. Can't a student be an electronics hobbyist without taking a loan?

>>1015371
>Yeah I can do a lot of things, but then you look at gaming keyboards that have 100+ LEDs running full blast off of one normal USB cable. That means it's possible and might be achievable for a noob like me.
Yea, but just because you think they are all on at the same time,,,, does not mean that they are.
TV IR remote LEDs are nothing special and they operate normally at 38 kHz.... Most common dirt-cheap LEDs still work pretty well at 1 megahertz.
It would be no big deal for the keyboard to have a chip to cycle them individually at 100 kHz. Persistence Of Vision, mang.....

If you wanna drive them all at once, then there's only two ways to do that:
--use them in series with enough voltage, or-
-use them in parallel with enough amps.

>>1015416
>Yea, but just because you think they are all on at the same time,,,, does not mean that they are.
I understand that. I already came to terms with the fact that multiplexing is the only way to go and it's actually a good thing.

I meant it as "those keyboards get enough power from a USB cord to drive 100 LEDs, I should be able to drive 50".

>>1015412
>>$400 for a decent scope
Depends entirely what you want to do. Much cheaper (used) scopes are totally sufficient for many purposes.
>$400 function generator
Again, depends. I need a function generator only rarely and my first one was self-made and quite bad.
>$400 power supply
China cheap. Alternatively, make your own. Decent lab supplies aren't that difficult to make.
>components used for the analog labs total $250
Dunno what you're going to do exactly, but there's generally no need to buy a huge component arsenal before starting. Buy more as you go.

All that said, it's easy to sink lots of money on tools and materials.

File: s_o_d_revealed_02a.jpg (308KB, 1400x1800px) Image search: [Google]

308KB, 1400x1800px

Update: the screwdriver of death.......

In the multimeter thread (now gone) I ran across a place selling these "tester screwdrivers".
I know about the little neon-bulb testers and how they work, but this didn't look like that.
It looked like total insanity and it was only $1.50, so I ordered one. It actually arrived pretty fast, for aliexpress stuff.

Pic related: the original picture of the aliexpress page I posted, but now with the teardown photo overlaid at the bottom.

Part #1 is not metal, it is some kind of resistor that measures around 1 megΩ end-to-end. Feels like carbon.

Part #2 is not a fuse: it is a little plastic cylinder with end caps that make it look like a fuse (it measures about 6mm x 25mm).
Inside is a tiny glass lamp bulb with the two leads looped back out each end, so they touch the 'fuse' metal end caps.
I think it's a mercury vapor because inside is just two electrodes and what appears to be some liquid mercury. The mercury moves around if you knock the lamp a bit.
...I can't find any mercury vapor bulbs this small on Google tho.... ? Is it mercury or something else? "miniature neon bulb" shows lots of pics, "miniature mercury vapor bulb" gets nothing related

There's no other hidden parts. The metal screwdriver shaft that is just plain steel.

I haven't applied power to it yet. I'm still kinda scared. Plus I live alone so my wall-current experiments have to be fail-safe. And it may be a 230-volt thing anyway (I am USA, plugs=115vac)

>>1015412

>$400 for a decent scope

For a novice, pretty much any cheap digital scope in the ~$250 range will do. If you're doing mostly small, practical projects (especially digital ones) you can actually get by without one for quite a while. I did. It kind of sucked at points, especially in a few cases where I really needed to see what kind of noise was on some data lines, but I managed.

Still, of all the things you could want for an EE lab, a scope is at the top of the list.

>$400 function generator

Again, not necessary to spend that much for entry-level equipment. And, honestly, you don't really need one most the time.

>$400 power supply

wat r u casul?

A power supply is one of the classic babby's first electronic's projects. Yeah, a transformer-based linear power supply will be heavy, inefficient, and not capable of all that much power, but it will cost peanuts in comparison and be sufficient for the vast majority of things you'd want to do with it.

Plus, power supply design is a big deal. No good reason to not at least dabble there a little early on.

>>1015431
>I know about the little neon-bulb testers and how they work, but this didn't look like that.
To me, it looks very much like a normal neon bulb tester.

>>1013210
>Is there a better way?
Perfboard/point-to-point for small projects, letting PCBWay or OSHPark just print the damned board for you for large projects.
If you want to have the fun of etching your own PCBs, that's fine of course... but you'll learn quickly it isn't fun at all and the first time you spill ferric chloride all over a workbench, I promise you it will happen, it will also be your last.
The hobby etching firms can do it cheaper, and they even do it fast.

>>1015432
A tiny little DS203 isn't a replacement for a good bench scope, but I promise you, you will use it a hell of a lot more often than a bench scope. It gets you a not-half-bad scope and a not-half-bad function generator in a very compelling package.

>>1015429
>All that said, it's easy to sink lots of money on tools and materials.
Components are my drug and DigiKey is my pusher.

>>1015434
>it will also be your last
i-is that a threat..

>>1015437
>i-is that a threat..
No. It's a promise. ;-)

>>1015440
MODS

>>1015441
ha! :-D

>>1015434

Honestly, the only issue I really have etching my own boards (photoresist + copper chloride) is that I can't do plated through-holes. Having to manually fill in vias is a massive pain in the ass that greatly increases assembly time, nevermind the fact that you outright can't use a through hole as a conductor for something that you can't physically access both sides of (like a screw terminal).

The second I figure out a good workaround for this, the only time I'll ever be ordering boards is if I need a lot of them.

>>1015436

DigiKey? That's some low-tier addict shit.

I once needed 6 crystal oscillators for a thing. But some eBay perusal snagged me a bag of over 800 of them for only 10x the price!

...I wish I could say that didn't happen often.

File: ltaoeanalogcomponents.png (800KB, 1837x5064px) Image search: [Google]

800KB, 1837x5064px

>>1015432
>>1015429

alright, i'll go for a DIY power supply and see how far I can get without a function generator (or use an audio jack as a shitty substitute).

But what do you think of this list of components? Do they seem overpriced?

>>1015451
I don't know about their prices, but the list in itself is rather weird for beginner's first kit. Motor driven pots? Incandescent lamps? Constant current diodes?
Unless you have some project book which uses all those, many of those components will be collecting dust for a long time, possibly forever.

>>1015464
those the listed components for the AoE lab book

http://learningtheartofelectronics.com/parts-lists/parts-lists/analog-parts-list/

>>1015342
> I want to drive a 12x4 matrix of white LEDs.
Is the matrix layout fixed, or can you choose to make it e.g. 6x8? The physical layout doesn't matter, only the connections.

> All I have is a tiny Arduino Pro Micro with not enough pins to do the job.
Arduino Pro has 14 GPIO pins. If you add an explicit decoder for the row, 12x4 needs 8 pins (4 decoded to 16 for the row, 4 for the column), 8x6 needs 9 pins, 6x8 needs 11 pins, 4x12 needs 14 pins.

> I'm not sure if it will have enough power to drive all LEDs close to max brightness
48 LEDs @ 20mA = 960mA. You can only get 500mA from a USB port.

> (even though it's apparently lights only one row at a time and switches between rows very fast).
For maximum brightness, you'd be driving the LEDs with more than 20mA. With a reduced duty cycle, the 20mA limit represents something between the average current and the RMS current; the peak current can be higher.

> If yes, how can I fix that?
Forget about the MAX7221; you don't need it.

What you need is a decoder, some high-current buffers, and a sufficient power supply. If you're limited to USB, you'll have to settle for driving the LEDs at the equivalent of 10mA constantly, or about 1/2 maximum brightness. E.g. for 8 rows and 6 columns, you'll be using 80mA with a 1/8 duty cycle for an average current of 10mA and a RMS current of 28mA.

The buffers for the rows will need to source (or sink) the current for all 6 LEDs with a 1/8 duty cycle. The buffers for the columns will need to sink (or source) the current for one LED constantly. These can be either IC buffers or discrete transistors (typically in common-emitter mode).

>>1015471
>or can you choose to make it e.g. 6x8?
yes

>Arduino Pro has 14 GPIO pins
Did I say how many were not taken? Because all of them are taken, while MAX7221 exposes a comfy 3 pin SPI.
Also, it has 21 GPIO pins.

>MAX7221; you don't need it.
No, I do need it.

>decoder, some high-current buffers, and a sufficient power supply.
See? I told you I need it.

Anyway, this whole thing is beside the point.
All I need is my 12x4 too look pretty all the way. No LED can be brighter than the other, but they don't need to be always on.
And also It would be a second matrix in a very tight space. Not knowing half of the words in your post is a way lesser problem than actually fitting your master plan there.

That's why MAX7221 is so intriguing. It's just one chip (+ another wire mess) and you have a working* matrix.

The original question would be how working would it be. Would there be a decrease in brightness across rows? Would it just be very dim? Would it burn my house/arduino down?
That's what I'm interested in.
But thanks for your reply nonetheless.

>>1015451

A lot of the common stuff (resistors, caps, diodes, pots, switches, etc.) can be had much cheaper direct from China. eBay or AliExpress, that is to say.

Honestly, for anything I don't need to be particularly exacting (or, in the case of chips, outright fake), I default to eBay. It's invariably cheaper and (being the default source of electronics in the first place) just as good for the common stuff. I resort to an "actual" supplier only for stuff like microcontrollers or odd stuff that I can't seem to find elsewhere for one reason or another.

>>1015483
In that case, the 7221 is a reasonable fit. The specified drive current is 30-45mA per segment, 320mA per digit. That equates to an average current of 3.75-5.6mA at 1/8 duty cycle (8x6) or 5-6.66mA at 1/6 duty cycle (6x8). Which isn't maximum brightness, but even with external buffers you could only increase that by 1.5x before hitting the 500mA limit for USB.

And it seems that this part is actually available (Maxim is notorious for "releasing" products which you can't actually buy anywhere). Compared to discretes, it's not exactly cheap, but that's not really an issue for a one-off.

What is the difference between general purpose caps and ones labeled as audio in terms of sound? Is it just audiophile crap? Or am I missing the point completely?

>>1015617
> What is the difference between general purpose caps and ones labeled as audio
Labelled by whom?

Capacitors are normally classified by dielectric type (e.g. electrolytic, tantalum, ceramic, polyester, etc), with more detailed parameters (e.g. ESR, ESL, tolerance) in the data sheet.

If you're talking about large electrolytics (i.e. power filtering capacitors), "audio" normally means that they have low ESR, meaning that they will be better at filtering out mains hum than other capacitors with the same ratings.

If you're talking about smaller capacitors, for signal processing (e.g. filters), you tend to avoid dielectric types with significant non-linearities; particularly ceramic capacitors, whose capacitance varies with voltage and which also exhibit microphony (the capacitance varies with external force, i.e. they behave like a microphone).

>>1015466
So... You going to build everything in that book and get all the components on one go?
If not, you could consider a bit more general and smaller set of components and buy more as you go, based on what you're building.

>>1015483
Mouser and Digi-Key want $8-$10 each for these chips....

If you are in the USA and can wait a bit for shipping, then there is MAX7219 modules on aliexpress for under $1.50 for the whole thing.
some use a SMD chip right on the PCB, but most of them use a DIP chip in a socket.
,,,,
Or you can buy ten of either [DIP] chip for about $1 each. The 7219 is way more popular in China-land, it seems.
The China chips might be copies or obsolete versions, but they still usually work like they're supposed to.

The 7221 or 7219, what is the difference? This page tells you about that-
http://playground.arduino.cc/Main/MAX72XXHardware#SelectProduct
(either will work)

File: byoc-figure-4-5.png (10KB, 753x498px) Image search: [Google]

10KB, 753x498px

Any electrical engineers with TTL & 8-bit MPU experience?
I want to build an 8-bit computer from scratch. Chose the Z80 since it’s simple and there’s a fuckton of info about it.
I’m trying to design a single-step circuit for troubleshooting. I’ve read Steve Ciarcia’s “Build your own Z80” and it had a circuit for stepping the MPU. >pic related
Ciarcia’s circuit used a 7474 flip-flop and a 74121 monostable multivibrator, with a 7400 to de-bounce the step button (I’ll be using a 74LS74 & 74LS00). The only issue is, he was using an original Z80 running at 2.5MHz, and I’m using a Z80A running at 4MHz.
What I’d like to know is should this circuit still work? I’m still a bit of a noob in electrical engineering. I looked up the specs on the ICs, and the pulse width of the 74121 can be adjusted by altering the values of the resistors and caps in the circuit. But does this even matter?
Am I worrying over nothing; can this circuit still be able to function at the faster clock speed?

>>1015758
Never used Z80, but based on a cursory look, the chips (74/74LS) are more than fast enough for 4MHz Z80.
It seems that the monostable pulse length needs to be shorter than the minimum instruction cycle and with the current component values it might be bit iffy. If it doesn't work, try halving the capacitor value.

File: z80_timing.jpg (66KB, 877x476px) Image search: [Google]

66KB, 877x476px

>>1015771
Thanks! I'll keep that in mind.

BTW, Z80 executes three machine cycles per instruction cycle. Each machine cycle needs 3-4 clock cycles depending on the op code (most are 3 clock ticks).

>>1015758
The original (nMOS process) Z80s had a number of dynamic latches, meaning that they can't be single-stepped (or clocked slower than a few hundred Hz).

If you want simplicity, go for a 6502; much simpler architecture and similar information availability.

>>1016047
That thing stops execution by forcing the processor to insert wait states. Clock runs normally.

>>1016047
I know about the clock limitations on the z80, that's why I'm using the wait pin like
>>1016056 said.

I may do a 6502 later on. It was dirt simple with only 3 registers, accumulator and two address, but it accesses memory so well that could create 256 virtual registers in RAM.
Later, I chose the z80 for the shear volume of documentation.

>>1016047
>>1016062
Also, I didn't want to putz around with a 2 phase clock signal. The z80 only needs a single clock signal. Remember I'm a electronics noob.

>>1016069
6502 only /needs/ a single clock signal (phi0); it generates the two-phase clock internally and makes this available on the phi1/phi2 outputs.

>>1016047
The original nMOS 6502s have the same issue (minimum clock speed due to DRAM registers)..You can use the RDY input to suspend execution or add wait states.

>>1016119
Didn't know that. I thought only the CMOS version (65C02) could generate the second clock.
But then again if I do, do a 6502 based system somewhere down the line I'll be using the 65C02 since it's easier to get a hold of them.

Thanks.

File: elegtroniggs.png (8KB, 1354x638px) Image search: [Google]

8KB, 1354x638px

Can someone explain to me the following in pic related:
1)Why do people switch (cut) the GND line in a DC current where they switch a motor for example?
2)There is this thing i cant name in English where you allow the current from the GND pole to the VCC pole of the motor via a diode.
The problem seems to be that, after switching off, there is still current generated on the GND pole of the motor that would destroy the switch on the GND line, therefore it is allowed to flow back through the diode to the VCC pole (why does it travel back to the VCC pole more easily than it would destroy the switch to get to GND?).
3) Wouldnt the problem be non existant if i switched the VCC line instead of the GND line?
4) How would i adress this problem in an AC circuit?

>>1016133
> Why do people switch (cut) the GND line in a DC current where they switch a motor for example?
Because it's simpler. If the switch is a transistor (BJT or FET) and you put it on the high side (between V+ and the load), then the control voltage is referenced relative to V+ (NPN or nMOS needs base/gate above V+ to turn on, PNP or pMOS needs base/gate close to V+ to turn off). If you put it on the low side, the control voltage is relative to ground (so with NPN or nMOS, 0V = off, a few volts is on, regardless of V+).

> 2)There is this thing i cant name in English where you allow the current from the GND pole to the VCC pole of the motor via a diode.
Flyback diode.

> The problem seems to be that, after switching off, there is still current generated on the GND pole of the motor
The current through an inductor cannot change instantly. V=L.dI/dt. When you turn the switch off, the inductor current continues to flow; you need to provide a path for it (which isn't through the switch).

> why does it travel back to the VCC pole more easily than it would destroy the switch to get to GND?).
It goes wherever it can. If you analyse the circuit at an instant, an inductor is essentially a current source. Kirchhoff's laws will give you the currents and voltages.

> 3) Wouldnt the problem be non existant if i switched the VCC line instead of the GND line?
No.

> 4) How would i adress this problem in an AC circuit?
You'll need to clarify what you mean.

>>1016137
>Because it's simpler. If the switch is a transistor (BJT or FET) and you put it on the high side (between V+ and the load), then the control voltage is referenced relative to V+ (NPN or nMOS needs base/gate above V+ to turn on, PNP or pMOS needs base/gate close to V+ to turn off). If you put it on the low side, the control voltage is relative to ground (so with NPN or nMOS, 0V = off, a few volts is on, regardless of V+).

And if i use a relay (magnetic coil) all that stuff doesnt apply, right?

>You'll need to clarify what you mean.

My AC situation:
I have a relay which i use to switch the "hot" wire of the 230V AC house network to power a transformator which gives out 12V DC. The relay switches the trafo on and off 3 times a day.

230V------relay------------Trafo+

Neutral----------------------Trafo-

If the transformator is an inductive load as well, i ll have to introduce a flyback diode as well to protect the contacts of the relay, right?

>>1016150
>And if i use a relay (magnetic coil) all that stuff doesnt apply, right?
Because its a physical separation of contacts i mean.

>>1016150

a relay contact can go on either the hot or the neutral, however common sense puts it on the hot, in case you happen to touch the wires on the load, you wont get a shock when the relay is off.

you cant use a flyback diode across the transformer coz it would short half of the AC cycle and burn up.

>>1016150
> And if i use a relay (magnetic coil) all that stuff doesnt apply, right?
Right.

> If the transformator is an inductive load as well, i ll have to introduce a flyback diode as well to protect the contacts of the relay, right?
If the load connected to the transformer is inductive, you'll need a snubber, varistor or similar on the contacts (a plain diode won't work with AC). The transformer itself isn't an inductive load (or any kind of load; it's what's connected to the transformer which matters).

If you're switching the relay with a transistor, you'll need a flyback diode on the relay coil (which is itself inductive).

Could I build a crystal radio, but for like microwaves that produces audio without any amplification? Particularly the X-band?

Not intending to get signals understandable by humans, just wanting to fuck around here.

>>1016181
> Particularly the X-band?
Nope. E.g. for a 7GHz tuned circuit, you're talking about an inductance of 100pH and a capacitance of 5pF.

File: Microwave_Coaxial_AM_Diode_Detector.gif (19KB, 325x361px) Image search: [Google]

19KB, 325x361px

>>1016161
>The transformer itself isn't an inductive load
Without load it is nothing but a big inductor. In addition to the inductive kick on turn-off, you'll also get a massive current spike on turn-on due to core magnetization.

>>1016181
While traditional discrete coils and capacitors are quite shitty at 10GHz, their transmission line equivalents work very well. You can build a crystal set like receiver from a transmission line resonator and a diode. Such simplistic receivers can be made to work at extremely high frequencies.

>>1016257
So how can I calculate how big a dish I need to receive enough microwaves to get something barely audible?

>>1015491
>Honestly, for anything I don't need to be particularly exacting (or, in the case of chips, outright fake), I default to eBay.
I recently got a giant bag of fake resistors from China on eBay. No shit.
If you buy from eBay (and really, it's usually a good idea) test what you buy before using it to build shit....

>>1015617
>>1015634
There are in fact capacitors out there labelled as audio/audiophile, and it's bullshit.
Specs are specs.

>>1016119
>The original nMOS 6502s have the same issue (minimum clock speed due to DRAM registers)
Not just DRAM regs. A bunch of gates in the 6502 use line precharge and inherent gate capacitance to hold a gate in an on state. Clock too slow and that doesn't work.

>>1014208
>Depends. If you need any speed, then you usually can't ignore it.
It's also possible to burn out mosfets by not driving the gates hard enough; the longer they spend in an intermediate switching state, the more power they're dissipating/wasting.

File: stuff.png (8KB, 1354x638px) Image search: [Google]

8KB, 1354x638px

>>1016257
>Without load it is nothing but a big inductor. In addition to the inductive kick on turn-off, you'll also get a massive current spike on turn-on due to core magnetization.

Thats what i thought.
While i can understand where the current is generated and goes in a DC context (from coil to GND) i dont get where the inductive surplus in an AC circuit, after turnoff, would go as AC is alternating poles.

Will it be generated at all?
Will it be generated on two poles (Hot and neutral)?

>>1016282
Can't say I know the topic too well, but since you know even less, maybe you should consider building something you understand bit better first? It's not fun when your design does not work and you don't even know why.

Well, whatever. Apparently something like 0.1nW is enough for a crystal set, when it's designed and built well and you have sensitive earphones.
You'd need to know the transmitter's effective radiated power (transmit power and antenna gain) and the free space path losses (assuming line of sight) between you and the transmitter. The ratio between the power after losses and that 0.1nW would be the required minimum antenna gain, which can then used to calculate the required size. Then add some extra gain to give some margin for fuck-ups. Or something like that. You might want a second opinion.

What are you trying to receive? Some radar?

>>1016304
>i dont get where the inductive surplus in an AC circuit, after turnoff,

it doesnt go anywhere, coz the voltage rises and falls comparatively slowly, so does the magnetic field. you only get a large angry voltage spike from a coil if you shut off current instantaneously. like women, coils dont like to be suddenly rejected, but if you ease out of it, that's fine.

>>1015433
>To me, it looks very much like a normal neon bulb tester.
Hory sheet. the wiki page for 'voltage tester' even shows one.... I'd never seen one like this. :o

I had an uncle who was an electrician, he showed me what most of the stuff he had was used for when I was pretty young, and yet I never saw one of these things.
When I asked him about it, he said that these weren't allowed at the company he worked for. Anything that allowed electricity to touch you was not permitted, so he didn't have one.

>>1016282
>So how can I calculate how big a dish I need to receive enough microwaves to get something barely audible?
I dunno much about microwaves, but there are "arduino"-style microwave emitters and detectors on aliexpress. They are meant for intrusion systems, but they are cheap and might be entertaining to play with.

Also note: in the low-RF spectrum most cheap discrete components (resistors, capacitors, inductors) still work, but when you get into the higher-RF and beyond, many 'normal' electronic parts don't act like they're supposed to.

>>1016399
so that sinus curve i see on the oscilloscope is enough of a slow ease rhytm to negate the danger of frying my relay by after-shutoff inductive current?
Did i get that right?

>>1016304
In the context of flyback transients, the fact that it's AC doesn't matter. You can treat it as DC where the voltage and current are varying (slowly) over time. Only the values at the turn-off point matter.

For a transformer, the total flux cannot change instantaneously. Provided that there's a load connected to the secondary, cutting the primary current will dump the stored energy into the secondary. And if there's no load connected, you won't have much primary current to start with.

If you're switching the current with a relay, flyback transients are far less of an issue than for a transistor (which can easily be destroyed by a single low-energy transient).

>>1016432

yes, you got it right. however, there was no need to even worry about this in the first place. these flyback voltages are only a danger to semiconductor devices that are sensitive to over-voltage and reverse-voltage conditions. when dealing with switches, relays, transformers, motors, lamps, etc, you dont have to worry about this stuff coz they're not fragile like semi-conductors.

>>1016529
That's not entirely accurate. If you switch a large current through an inductive load, the resulting arcing can have a significant impact upon the lifespan of the switch.

For this reason, such loads (e.g. large motors) should always have a snubber across either the load or the switch (preferably the load, as that also handles the case where you cut the power by means other than the switch).

However:

1. A transformer connected to a resistive load behaves like a small inductor (representing the transformer's leakage inductance) in series with a resistive load. The flyback current is the transformer's leakage (no-load) current, not the current used to actually power the load.

2. For physical contacts, small currents (or short spikes) don't matter much even if the generated voltage is high. Damage to contacts from arcing is determined by the energy of the arc and how often it occurs. Whereas a single low-energy transient can destroy a transistor.

>>1013894
Hammond and Classic Tone also make good transformers.
Classic tone seem to be high voltage for tube amps mostly, however. I'll probably get one from them if I build a tube amp.

>>1016285
>I recently got a giant bag of fake resistors from China on eBay. No shit.
details?

How are they fake?
carbons advertised as metal films?
are they just zero-ohms value?
are they mebbe inductors?

China sellers offering carbon resistors as metal-films has been found to be a very common problem; even wholesalers try to do it to big resellers.

File: 3_6.jpg (33KB, 710x533px) Image search: [Google]

33KB, 710x533px

Wien bridge oscillator is GOAT

>>1016583
>How are they fake?
They were made to look like 1%, 100ppm metal film resistors. I can't tell what they are, but they're 10% at best at room temp and have the worst temperature coefficient I've ever seen. They change almost 100% just from the heat of holding them in my hand. Hit them with a hot air wand and woah nelly.
So yeah, probably some total crap carbons.

>>1016547
>That's not entirely accurate. If you switch a large current through an inductive load, the resulting arcing can have a significant impact upon the lifespan of the switch.
...or weld it closed the first time you try to open.

>>1016285
>I recently got a giant bag of fake resistors from China on eBay
>mfw I just bought 1460 resistors for $5.89 w/ free shipping on aliexpress
>no face

>>1016399
right. When you put a voltage across an inductor and current starts to flow, the energy of the current*voltage drop is being stored in the magnetic field of the coil. Open the circuit and the current instantly drops to zero. But all that volts*amps energy is still there and in response voltage goes to infinity. You've just created a singularity with Ohm's law.
Except it isn't a real singularity; in the real world, the coil *will* dump that energy, usually at the point where the instantaneous voltage climbs high enough to ionie the air (and any insufficient insulation in the way) and arc.

File: wiencircuit.jpg (25KB, 399x372px) Image search: [Google]

25KB, 399x372px

>>1016625
At least add amplitude regulation before claiming so.

File: IMG_20160702_122828.jpg (4MB, 2992x4000px) Image search: [Google]

4MB, 2992x4000px

How does my soldering look bros? I've only had proper experience soldering wire etc.

It's just a bunch of transistors and resistors, I'll be using it to interface between my old car and an rs232 port.

>>1016723
you're doing all right, but could stand to use a little more flux and flux when you reheat joints. Also clean your tip a bit more often :-)
Those irregular joint shapes are from iffy wetting and oxide buildup causing you to use more heat then necessary and holding it there longer than needed, yet still having inconsistent heat transfer. But it's not actually that bad.

File: IMG-20160618-WA0002.jpg (73KB, 636x626px) Image search: [Google]

73KB, 636x626px

>>1016730

Hey thanks for the advice, I appreciate it.

>>1016739

In addendum to this, I'll post a follow-up after I touch it all up, I'd like to see how I did.

>>1016702
Whats the purpose of that lamp?

And how does that regulate the amplitude?

File: DSC_0010.jpg (3MB, 3840x2160px) Image search: [Google]

3MB, 3840x2160px

Bit of a noob question: Can I replace the fan in this power supply (the one on the right) with a generic pc fan like the one on the left without causing problems to the power supply or the fan? The problem with the current fan is that it won't start at the lowest voltage and it whistles until the voltage creeps up.

>>1016763
I'm not an expert, but it would most likely be fine.
Be careful connecting it, though. The one on the left has an additional pin, so you'll have to figure out which one of them is not needed and not connect it. The third pin usually pulses every revelation or something like that, used to monitor RPM.

File: bldcc.png (75KB, 456x825px) Image search: [Google]

75KB, 456x825px

I'm trying to make a 3 phase motor controller but I'm stuck on trying to design the PCB. The first attempt I tried failed, I had the MOSFET drivers close to the MCU but it had had issues with weak drivers and long traces between the driver and FET gate.

This time I put the drivers as close to the FETs as possible but that makes the MCU trace longer and it's not feasible to run a ground plane under the PWM signal traces the entire way. Anyone know if the signal is likely to be ok? The board is pretty small 100mmx50mm.

>>1016755
Tungsten wire has a positive temperature coefficient of resistance. Higher output amplitude -> higher voltage across the lamp -> hotter filament -> higher resistance -> lower gain -> lower output amplitude.
While regulated output amplitude is in itself a nice feature, a more important benefit is that when you regulate the gain of the amplifier so that it is just sufficient to maintain oscillation and not more, the distortion will be much lower. Lamp regulated Wien bridge oscillator was a patent worthy invention and HP's first product.
Interestingly, it is quite difficult to beat the lamp in performance with more modern solutions.

https://en.wikipedia.org/wiki/HP200A

whats your favorite simulator for circuits in general? always been using OrCAD pSpice but i really hate it

mostly for oscillators btw

>>1016809
Well that's interesting.

Yknow, I have no idea how that wien bridge oscillator works and I'm a senior EE major kek

>>1016785
Your PCB image is too hard to read. Typical motor PWM frequencies are not that picky about routing.

>>1016785

Signal should be fine. It's going to be fairly low frequency and I can only assume there's at least some cursory filtering/buffering on the driver's end. As long as it's of suitable magnitude, the PWM signal can get pretty jank and still work fine.

>>1016876
Since the amplifier part has a phase shift of zero (it's a basic non-inverting amp), it means that the phase shift caused by the RC filters (the Wien bridge) must be zero as well in order for that circuit to oscillate.
So you have a voltage divider made of 2 resistors and 2 capacitors. If you assume that the both resistors and the both capacitors have the same value, it is pretty easy to show that the phase shift will be zero when f=1/(2*pi*RC).

>>1016908
I'll try to do the analysis by hand next week. Thanks man

Also, if I'm not mistaken, does the Wien bridge use the op amp power to create the oscillations? I dont see where the Vin is at.

Working with logics for the first time.
To tie a signal high (1) do you ground it? Is a resistor needed for 5v TTL?

Using 74LS logic.

>>1016971
>To tie a signal high (1) do you ground it?
High implies not ground, but it's a convention. You decide.
>Is a resistor needed for 5v TTL?
All TTL circuits require the sacrifice of one 471 ohm resistor to the TTL god. Place it in your circuit but don't connect it to anything.

>>1016971
>>1016989
In our digital logic class my freshman year of EE we never used resistors lel

>>1013645
>but you still need some form of direct position sensing to prevent integration drift.
Maybe use a Kalman filter

>>1016989
Jokes aside, what you're saying is tie the pin to +5v?

>>1016998
If you're pulling it high, yes.

You can put like a 1k pull up resistor in there if it makes you feel safer.

>>1017007
Thanks.

>>1016913
Yes, oscillators need something which provides power gain to counter losses, which in this case is done with an opamp. If there's not enough power gain to sustain oscillation, the oscillating energy will be dissipated sooner or later.

https://en.wikipedia.org/wiki/Barkhausen_stability_criterion

>>1017024
Got damn i love electricity.

My only regret in life is not taking up the electronics hobby in high school or earlier. Now I'm an EE senior trying to figure out how to apply this shit I have learned, and having my mind blown in the process.

>>1016913
> I dont see where the Vin is at.
It's an oscillator; it doesn't have any inputs.

The fundamental requirement of an oscillator is that the complete loop has unity gain and zero phase shift.

A typical oscillator consists of an amplifier with its output connected to its input via a passive filter. The amplifier's gain counteracts the filter's attenuation to provide unity gain overall. Either the amplifier is non-inverting and the filter has zero phase shift, or the amplifier is inverting and the filter has a 180-degree phase shift.

The trick is in maintaining stability. Maintaining zero phase shift basically means that phase shift needs to decrease as frequency increases (constant delay) so that any drift is opposed rather than reinforced. The oscillation frequency is the reciprocal of the delay.

Maintaining unity gain requires some kind of automatic gain control. For square wave, saturation provides this. For a sine wave, using a filament lamp in a voltage divider is a simple solution, although using a FET as a voltage-controlled resistor is more typical.

File: $T2eC16R,!zUE9s38+GdTBR8IoBZ3pQ--60_1.jpg (15KB, 400x400px) Image search: [Google]

15KB, 400x400px

Can anyone recommend a decent potentiometer for a steering wheel? Preferably single turn.

>>1016785
The main consideration for high-current PWM is to make the area of the high-current loops as small as possible. These are basically single-turn solenoids which can generate substantial magnetic fields which will be inductively coupled to everything else on the PCB.

>>1017048
> potentiometer for a steering wheel?
The potentiometer itself is less relevant than how you couple it to either the wheel or the case.

Unless the mounting of the wheel is particularly rigid, you're going to want some flexibility so that any undesired forces aren't coupled directly to the wiper.

Basically, you have four "joints": case->wheel, wheel->pot-wiper, pot-wiper->pot-track, pot-track->case. The pot-wiper->pot-track joint needs to be fixed, so any motion in the wheel->case joint needs to be taken up by either the wheel->pot-wiper or pot-track->case joints (otherwise, it will end up as motion in the pot-wiper->pot-track joint).

Rather than trying to make the pot-wiper->pot-track joint more rigid than everything else, you should be thinking about making one or both of the others less rigid while still transmitting torque about the potentiometer's axis.

File: relay.png (202KB, 640x269px) Image search: [Google]

202KB, 640x269px

OK, so I have this relay
what exactly do I have to pass to the S pin to get it to switch?
it seems like no matter how I wire it up, the common and NC outputs are the ones that are connected

I'm not using an arduino, I'm just learning this stuff using a breadboard

>>1017190
given there's a transistor on the board with the relay, i'd say the S pin is connected to the base acting to switch the larger 5v rail. try connecting the 5V and S pin with a 1k resistor and see what you get,

>>1017207
nothen'
I suppose it's broke?

>>1017209

relays are tough to kill. more than likely, something else died. just stick 5V across that diode: positive to the banded side. betcha that relay is gonna click.

>>1017217
nope, nothing happens
tried up to 9v just for the hell of it
nothen.

sup /ohm/

i'm looking for a digital circuit simulator
what i have already tried was logisim but i want a sim with vhdl or verilog support
yes, i know google exists but i want your opinion on which is recommended for hobby uses (i'm mainly building microprocessor circuits)

>>1016766
Nice one, thanks.

>>1016763
Yes and no. The extra pin in the other fan could be RPM monitor as mentioned by >>1016766, OR or speed control, if the extra pin is meant for speed control when you connect negative and positive you'll get really low RPM, maybe the fan will start as you wanted but wont be able to reach the proper RPM to cool your PSU. Sorry for my English, im not a native speaker.

>>1017426
>Sorry for my English, im not a native speaker.
lol it okei

I have a Tektronix TPS2024 Oscilloscope and a Wavetek 278 function generator.

I set up the function generator to output 1 kHz at 10 V peak to peak and connected it straight to the oscilloscope. My oscilloscope read ~11 V peak to peak and ~980 Hz.

How do I know which one needs to be calibrated? Alternatively, how do I calibrate them?

>>1017661
you calibrate one then use it to calibrate the other
to calibrate it you send it to a calibration company and then bend over

>>1017661
You can check the scope y scale with a power supply and a trustworthy multimeter. The timebase can be verified by using a trustworthy crystal oscillator and an optional divider.
Apparently the both instruments are supposed to have crystal controlled timebases, so I suspect that the frequency error might be caused simply by the rather short memory of the scope. Check if the sweep speed setting affects the reading.

Well, and of course there's that option the other anon suggested, if you hate money.

File: IMG_0035.jpg (477KB, 975x556px) Image search: [Google]

477KB, 975x556px

Trying to generate a 4MHz clock signal. I'm using a 74LS04 as a series resonant oscillator.
With a 4MHz crystal it's OK, the wave is not total square and the duty cycle is a little off, so I decided the use a 16MHz crystal and use a 74LS74 flip-flop counter as a frequency divider.
Unfortunately the output signal at 16MHz is nowhere near square.
Am I doing something wrong or is the 74LS04 simply not up to the task at the higher frequency?

Any suggestions?

>>1018138
Did you really take a picture of the screen?

>>1018143
What do you mean? How else would someone see the results I'm getting.
Have I committed a /diy/ sin?

>>1018138
Why do you care whether it's square?

What matters is that the signal can drive the flip-flop, i.e. the maximum is above the minimum V[hi] and the minimum is above the maximum V[lo], and the minimum pulse width is satisfied.

If a 74LS isn't fast enough, use a faster family, e.g. 74ALS, 74S, 74ACT or 74AS.

>>1018138

that looks more or less like what one would expect at 16Mhz. nothing wrong with that pic.

as for screenshots, in the 90's we used polaroid cameras with a special square shroud. in 2016 we kinda expect a screen dump from scope memory.

>>1018151
Insert a flash drive (had to be formatted FAT…) into the USB port and press the green button with a printer on it.

File: MnT7WDQ.png (41KB, 800x480px) Image search: [Google]

41KB, 800x480px

>>1018173
To add, I know because I'm 99% sure we own the same scope (Rigol DS1054Z) but any of Rigol's scope can do screen dumps.

>>1018162
>>1018169
Thanks a lot! I'm kind of new at digital logic.

>>1018173
I just got the thing and haven't played with all the features yet. I'm used to an analog O-scope. I'll do that next time, thanks.

>>1018138
Your picture makes me wonder if you have a 1 x probe (instead of 10x) and a shitty ground.

In any case, like the other anon said, the shape isn't really critical as long as the levels are suitable and the transitions are fast enough and monotonic.

>>1017394
Doesn't Altium have FPGA sim support?

>>1016285

Were there reviews on the product or the supplier? I tend to buy stuff only when there's at least one review confirming it works, I haven't had a bad time yet.

>>1017048
>potentiometer for a steering wheel
Oh please for the love of god don't. Potentiometers degrade. Have you never used a knob on a stereo where the knob makes the volume go blaring to max and shit? Use the same kinda thing they use in mice with a calibration button, or have some other deal, like a bunch of photodiodes and LEDs to estimate the position of a spot on a rotor.

Why do we even care about the beta (hfe) of a BJT? Like why even learn about it? I'm told time and time again that beta is an unreliable parameter and it is subject to vary with current, temperature, and you looking at it funny. Every book I've ever read mentions to not base your biasing on the beta yet every example problem has you calculating resistors based on the beta or you're given a beta.

So what the fuck is the deal with the beta parameter and why even use it at all if it's so fucking unreliable?

>>1018435
>Why do we even care about the beta (hfe) of a BJT?

are you kidding, dude? beta is a measure of gain. gain is the single most important thing transistors have to offer. how you gonna make a radio, for example, if you cant boost that microvolt-level signal into volts so you can hear it?

to get around the unreliability of the gain parameter, you just design with more gain than you need, and use feedback to limit it. so, if you need a gain of 50, you use a transistor with a min. gain of 100 and adjust the resistor ratios so it only amplifies 50 times. that gives you a ton of overhead for unforeseen circumstances.

>>1018303
>Were there reviews on the product or the supplier? I tend to buy stuff only when there's at least one review confirming it works, I haven't had a bad time yet.
(not that guy)
the reviews of the sellers I bought from has nothing about the matter. most people don't conduct an in-depth check I guess. I know I didn't.

I'd heard people say that there was fake parts from China but they never specifically said which parts.
the usual example was some fake atmega chips got sold, but that was only for a brief time and those were 100% non-functional.

the blue 1/4-watt resistors are usually advertised as 1% metal-film.
When you remove the paint most of them don't appear to be metal film, and they don't show the correct response when heated.
they hit their values at room temperature so they can still work for some things. it's not a total loss, just a lesson.

Both packs I bought are fakes. At this point I'd just assume that ALL of the blue ones sold as 1% metal film are fakes.

I got some of that type from Newark and they are real metal-films.
If I bought any from a US-based reseller like Newark, Digi-Key, Sparkfun or whatever and they were fake, I'd definitely raise the issue.
It's not worth it for the China sellers. It wasn't much money and doing anything takes too long to bother.

>>1018306
Nothing lasts for ever. A potentiometer is much cheaper than an optical encoder.

The main thing about using one for a steering wheel is that you need something which will survive continuous (rather than occasional) use.

IIRC, servos (which have the same issue) tend to use pots with conductive plastic tracks rather than carbon.

>>1018435
> So what the fuck is the deal with the beta parameter and why even use it at all if it's so fucking unreliable?
You can't avoid using it. However, any robust design won't rely upon a specific value of beta, only that it's between some minimum and maximum value.

Most of the time, you only care about the minimum value, i.e. you design a circuit which works with any beta above the specified minimum value. For linear circuits, this typically means using negative feedback so that the closed-loop gain is only marginally dependent on beta.

>>1018594
Are there circuit designs more robust than voltage divider bias or is that pretty much the best there is?

File: image.jpg (69KB, 800x600px) Image search: [Google]

69KB, 800x600px

I was tinkering around with 2 old signal amplifiers and was successfully able to get over the air signal from about 5 inches. I want to try to expand that range to something ALOT bigger, but know nothing about electronics. I was thinking about mabye repurposing an old microwave power supply, but i have no idea how these things work.
What I want to happen:
-not die from electric shock
-have a broadcast range of 100-500ft
-as little power draw as possible
-needs to be NTSC

File: active-bias.png (5KB, 459x424px) Image search: [Google]

5KB, 459x424px

>>1018639
> Are there circuit designs more robust than voltage divider bias or is that pretty much the best there is?
Voltage divider bias only depends upon beta to the extent that beta will affect the base current which will tend to drag the bias voltage toward the emitter voltage. If the current through the voltage divider (V/(R1+R2)) is large compared to the base current, beta won't have much effect.

For an AC-coupled stage, it's fairly common to use an emitter follower (which significantly reduces the DC base current) then bypass it with a capacitor (which restores the AC gain without affecting the DC operating point).

If you want better than that, you can always use an active bias circuit (e.g. pic related).

>>1018584
Well yea but there's optical encoders on aliexpress for ten bucks... You just need a microswitch to make a self-centering point.

>>1018643
What frequency? You can get AM freq chips for stupidly cheap.

>>1017394
Probably the free Xilinx or Altera tools will be your best bet. Ignore the FPGA part, just use the sim.

>>1018643
>mabye repurposing an old microwave power supply
Yeah dude, this doesn't make any sense.

I'm looking for some sort of needle dial that I can control from a computer/arduino/DAC. Searching for needle dial on eBay shows up lots of digital depth gauges or watches or bimetal thermometers, which aren't what I want.

Anyone have an idea?

>>1016723
Ugh, that's not how you solder those perf boards. You should avoid making solder bridges between pads. It's a waste of solder, those joints are unreliable and it's all rather clumsy. You cannot make a proper solder joint like that. Solder is not supposed to act as conductor by itself, it's supposed to make a joint between two conductors.
What you do is after populating the perf board with components you either bend their legs to make connections, use pieces of wire to make traces or just use jumpers between pads. And when you solder try to connect elements point to point instead of making those gooey snots.
Some examples in links bellow should give you an idea what it should look like:
http://www.primrosebank.net/computers/mtx/projects/mtxplus/video/build8.jpg
http://www.runoffgroove.com/mreq.html
https://itp.nyu.edu/archive/physcomp-spring2014/Tutorials/SolderingAPerfBoard

It's even possible to solder on perf boards if they have only holes without those copper pads, although they help to keep the components in place.

>>1018667
Analog meter, panel meter, voltmeter or galvanometer.

File: diy.png (618KB, 537x540px) Image search: [Google]

618KB, 537x540px

DIY oscilloscope kit, thoughts?
http://www.aliexpress.com/item/r/32371281977.html

>>1018741
Looks like a fun little kit but I doubt it can compete with any cheap second-hand scope.

>>1018741
Basically one step above using your PC's sound card with an external attenuator: it can measure DC and sample rate is somewhat higher. There's much less memory, though.
Might be a fun little kit like the other anon said, but as a scope it is awful.

>>1018741
These are complete and utter garbage can't be used for anything.

Except if you want a really really expensive soldering practice board.

>>1018674

Shit okay, I didn't know any better, I'll redo the board

File: FT231XQ.jpg (5KB, 110x110px) Image search: [Google]

5KB, 110x110px

Anyone know what is the difference between "Basic" (FT230x) and "Full handshake" (FT231x) RS232 UART? Trying to figure out which FTDI chip I need for my project

Any tips on level shifters is welcome too...

>>1018803
speaking of that
have any recommendations for soldering practice kits?

>>1018828
serial cables have 9 pins (or 25)
most of the time you only need 3
if you don't know this then you are out of your depth with this.

Whats a good source to learn applications and uses of transistors in circuits?

My EE electronics class did fuck all for me.

>we didnt even talk about transistor amplifiers

>>1018693
Thank you. Found what I was looking for.

>>1018741
Go watch a EEBlog's video about those.
TL;DW - they are badly shielded so your trace will have a huge amount of noise mixed in.

>>1018828
> Anyone know what is the difference between "Basic" (FT230x) and "Full handshake" (FT231x) RS232 UART?
230x only supports the Tx, Rx, CTS, RTS lines, which is sufficient for a fixed serial link (e.g. PC-to-PC or PC-to-TTY). 231x also supports DTR, DSR, DCD and RI, which are needed for modem connections.

> Any tips on level shifters is welcome too...
MAX232 family. They only go to +/-9V, but that's sufficient for everything you're likely to encounter unless you're working in a computer museum. Until serial ports vanished altogether, laptops often only had TTL-level serial ports (i.e. 0V/+5V), and they usually worked.

>>1019039
http://www.allaboutcircuits.com/textbook/semiconductors/

This is pretty good for general application of a number of semiconductors. If you want to learn how to build a certain type of device, you're best finding a book on that specific application (ie RF books for radio projects). Often times, certain circuits are able to be dropped into circuits as larger blocks to achieve desired effects.

>>1019110
>laptops often only had TTL-level serial ports (i.e. 0V/+5V), and they usually worked

RS232 standard requires that the high signal (1) be a negative voltage and the low (0) was a positive voltage. That's why many UARTS required both a -V and +V from the power supply.
TTL levels are +5v for high (1) and 0v for low (0).

I agree the MAX232 is a good solution since it only needs +5v. It uses carge pumps to push the levels to +10v & -10v, it also does the level conversion so low is pos and high is neg.

Can I just hook an ir photodiode directly up to my PC's power on pins to use a remote to turn it on?

>>1019123
> RS232 standard requires that
Most people don't care about what standards say, they care whether stuff works in practice.

Laptops often lack negative and 12V rails, so they just use 0V for 1 and 5V for 0, and if you want your device to work with a laptop it has to handle this. Many serial chips will auto-detect any bias voltage to allow for this.

>>1019126
Not reliably. You'd need to ensure that the current at ambient illumination was below the threshold and the current when illuminated was above.

In practice, it would either never turn on or turn on at random, depending upon the sensitivity.

>>1016730
>>1016754
Here.

I'm not satisfied at all with that shitty board I made so I want to do a different route.

Is it a non-retarded idea to make my project with copper plated board?
My other question; could I use board that has copper on both sides and when I need a trace to go to the other side of the board could I use small wire to go through the board and solder it on both sides?

Thanks.

>>1019126
Maybe a bit too complicated but wouldn't it be more reliable to use a microprocessor to trigger a transistor to act as a button?
You could then hook an IR photodiode to it and have the microprocessor trigger the transistor only when it reads your remote's signal.

>>1019160

Fuck remove that first post number.

File: bussboards_dd2897.jpg (238KB, 938x862px) Image search: [Google]

238KB, 938x862px

>>1019160
>I'm not satisfied at all with that shitty board I made so I want to do a different route.
>Is it a non-retarded idea to make my project with copper plated board?
As I've seen it: phenolic board with solid copper on both sides is usually called "copper clad board".
I wouldn't recommend using that unless you etch it, which is what it's made for.
I dunno what you are doing but cutting runs out on all four sides would be rather a drag. Not to mention drilling the thru-holes.

What some people like using a lot is bus board: it has the holes and copper plating in strips that connect rows or groups of holes. To limit the length of a group of connections in a long row, you just slice the strip with a engraving tool or whatever.
Pic related: some bus board examples. This is a more-expensive brand, but they do have some nice variations.

Also note: making your own "traces" is easier when you have bus wire--bare tinned wire.
The normal stuff is 22 gauge but I prefer to use thinner 28 gauge for digital stuff since there's never much power involved and the thinner stuff is easier to work with when you have to make runs along side each other (without touching!) on .1" spaced perf boards.
I got my 28g bus wire at Newark; most places only sell 22g.
,,,,,,
Or if you are wanting to do it cheap(er) or fast(er)--you can also strip and unravel some regular stranded wire and just use that.

File: noise.jpg (82KB, 1018x460px) Image search: [Google]

82KB, 1018x460px

>>1019160
You can etch the copper clad board to have suitable routing. It isn't even particularly difficult or expensive.
If you don't want to do that, then there's the ugly construction (pic) and its relative Manhattan construction.
Nothing really prevents you from dremeling the routing, but that's bit tedious and the results are generally quite ugly.

And yes, you can certainly use a soldered piece of wire to go from the bottom side to the top side. I've personally used thick copper wire (0.8mm diameter) for that purposed and then riveted it in place before soldering.

>>1019126
>Can I just hook an ir photodiode directly up to my PC's power on pins to use a remote to turn it on?
That depends on how much current the power button has to conduct. Have you measured that?

Even if the current limit is within what a photodiode can take,,,, I'd think it would be safer to use some kind of intelligent filtering.
Like say, set it up so that it only turns the PC on if the photodiode reads high for at least 250 milliseconds but not more than 260 milliseconds.
The timings don't need to be accurate to milliseconds, but you get the idea here: establish a set "high" time, and ignore any "high" state that is shorter or longer.

You could do this with analog circuits or with another microprocessor.
The analog way is much simpler but I am more familiar with programming and neither way is very expensive to implement.

File: banana_probes_fre09ref.jpg (61KB, 1070x486px) Image search: [Google]

61KB, 1070x486px

Cheap Crap to Maybe Buy (or not)

I found these kinda by accident...

In one of the EEVblog videos, Mr Jones reviews a rather expensive Gossen multimeter.
It has probes where the tips are pointy, but they are also made to fit into female banana jacks.
I'd never seen that before, tho I had not looked I suppose.

Hantek makes probes like that too for their o-scopes, and if you wanted a pair you can get them on aliexpress.
Most places want around $10 for them, but there is one place that only wants half that--pic related.
If you search on aliexpress for "multimeter probes HT19" you can find all the listings.

I don't know but I am almost certain that the back end of these probes is a regular female 4mm banana jack, so you 'd need at least two of those cables also.
I don't got any of these probes so I don't know, but the cable kits that come with the o-scopes all show a pair of probe cables that have male banana plugs on both ends.

I decided I didn't need any. Somebody else might tho.

>>1019202
Looks pretty handy for the oscilloscope variety, but a bit pointless for a Multimeter since you don't need any special cables. You could probably make your own for very little money.

>>1019197
This hurt my eyes.
Reminds me of this:
https://www.youtube.com/watch?v=2o8MDCIlOEk

>>1019194
>>1019197

Thank you both.

To *194: I'm perfectly willing to get the supplies needed to etch my own boards, that was my intention when I mentioned the idea, I was just worried about the through hole connection.

I am trying to make a system with two motors, one that runs when the reverse signal is given, and one that runs when the forward signal is given using a set up that was previously connected to a stepper motor.
I have an H bridge, power and ground, a few tip120's and an input wire that sends the signal for forward or reverse. How would I add in the Hbridge so that it can redirect the signal to different motors?

>>1018674
>Ugh, that's not how you solder those perf boards.
You mean, that's not how you do it. Dragging solder bridges is a long time-honored way of building on perfboard. Solder makes a perfectly good conductor for all but high-power applications. The connections are reliable.

>>1019240
>I am trying to make a system with two motors, one that runs when the reverse signal is given, and one that runs when the forward signal is given using a set up that was previously connected to a stepper motor.
>I have an H bridge, power and ground, a few tip120's and an input wire that sends the signal for forward or reverse. How would I add in the Hbridge so that it can redirect the signal to different motors?
More info would be helpful.

An H-bridge has two outputs that can swap polarities.
If you just got two motors and you intend that each only spins one direction, then all you need to do is connect a power diode to each, and hook their wires up oppositely to the H-bridge output.... So when the [+] is on one side, only one motor will spin, and when the [+] is on the other side, only the other motor will spin.

If you are replacing a stepper motor because it can't spin fast enough, then you would probably need to get bigger DC brushed motors since ones the same size as the stepper would not have nearly the same torque.

BLDC motors are nice but they usually cost a bit more than an equivalent-size stepper setup--using a "CNC-type" driver. They spin much faster than a stepper can and they still can put out a LOT of torque at lower speeds. I bought a setup (drive + motor) that is the same frame size as a NEMA-17 motor a while back, and it cost around $80 (about $10 more than a stepper motor setup the same size).

RC motors are BLDC, but RC controllers don't use separate Hall sensors... and RC controllers are built pretty small. I dunno how they'd hold up to a lot of high-torque use.

>>1016785
Are you sure you're using the right MOSFET/voltage? Most need 10V to turn on for >1 amp

>>1019253
So here is all info I have:
Project goal; Makerbot Replicator 2 3D printer extruder using pressure in a syringe instead of the normal NEMA17 stepper motor.

Progress so far: I made a mount for everything and am using two solenoids (US Solids normally closed 1/4" 12V DC Brass) and have them connected through tubing to an air source and to just the open room for the relief. One is currently wired to two (green/black) of the four stepper wires through the circuit I found here (http://www.bc-robotics.com/tutorials/controlling-a-solenoid-valve-with-arduino/) with the stepper motor having one of its input wires being used in place of the one that connects to the arduino. The other solenoid is hooked up the same way and uses the other two wires (yellow/black) from the original stepper set up.
It currently will activate the first solenoid when the Gcode would normally say to extrude, so that bit is working, though it clicks a lot and I don't entirely know why.

Where I am stuck; when I tell the stepper to reverse instead of activating the relief solenoid, the pressure one gets activated again as if I told it to extrude, hence why I believe the H bridge is needed. I also can't regulate the pressure low enough so that it extrudes slowly but that would require like a psi gauge/valve kind of thing which I haven't been able to find at all.

So the reason I am at /ohm/ instead of /3dpg/ is because I am trying to fix up the circuit and h bridge specifically for now. If any more info is needed let me know, thanks for the help so far, I am relatively new to this, so this was a much bigger project than I originally expected it to be and I am a bit out of my depth.

>>1019240
An H-bridge is used to run a single motor in either forward or reverse. If you have two motors, you don't need an H-bridge, just a single transistor for each motor.

Other than that, you need to provide more specifics.

File: osc-div.jpg (80KB, 1618x470px) Image search: [Google]

80KB, 1618x470px

>>1018138
Still working on an 4 MHz oscillator.
I tried to use a 16 MHz crystal with a 74LS04 and run it through 74LS74 flip-flop, but could not get anything out of the divider.
I see from the specs the clock on the 74LS74 can handle up to 20-25 MHz.

To troubleshoot I substituted a slower crystal into the oscillator and I get a good signal, but still nothing out of the 74LS74, just a flat 3v. I should be getting 2 MHz out of the first flip-flop and 1 MHz out of the second.

What am I doing wrong?

File: xtal-osc.png (6KB, 628x701px) Image search: [Google]

6KB, 628x701px

>>1019514
Scope pin 3. Does the signal meet the level and timing requirements in the data sheet for the specific part that you're using?

FWIW, pic related is what I use.

>>1019514
Why C2? Unnecessary.

>>1019514

that 47nF cap is so completely wrong. it destroys the output characteristics of the logic gate. maybe with CMOS, it would be fine, but not TTL.

File: 1382873043014.gif (158KB, 500x282px) Image search: [Google]

158KB, 500x282px

>Computer stops working
>Check everything except MOBO
>Nope
>Check MOBO
>Eight (8) blown capacitors and no standoffs
>Get quoted $0.35-$0.89/ea. from a store 50 miles away that may or may not have them
>Another store 60 miles away has them for $2.25 each
>No way to buy them online
>All of $45 to my name and no gas in my car
JUST

>>1019286
Following step-by-step recipes and designing your own circuits require entirely different skill sets. Experience with the former is of no use for the latter.

If you know enough electronics to design your own circuits, you don't need the recipes. If you don't, you need to follow the recipe to the letter; don't try to tweak it.

File: DS1Z_QuickPrint1.jpg (199KB, 800x480px) Image search: [Google]

199KB, 800x480px

>>1019578
>>1019580
>>1019587
Thanks all! Getting rid of C2 did the trick.
I'm not sure why it was there either, it was in the diagram I used for the oscillator. Might have been bad design or a misprint.

I'm still such a noob not to notice that made no sense. it's perfectly obvious now that I look at it. Stupid me!
That should be good enough to run a z80.

>>1019061
I looked for quite a bit but couldn't find his video on it
got a link?

File: measuringwheel.jpg (267KB, 1200x1200px) Image search: [Google]

267KB, 1200x1200px

>>1013522
Could you use the principle of a measuring wheel to do it?
Maybe some small magnet on both wheels (using just on one would differ from the actual distance when you turn)
Or a third small wheel with this as its purpose (located under the center of the mouse, so it won't be affected by the above problem).

>>1013204
Total newbie here. If I'm working with chips with a fuckton of pins (64) do I really have to connect all of them?

>>1019770
It depends entirely on the chip in question.

File: confused_hf701d4.jpg (57KB, 427x480px) Image search: [Google]

57KB, 427x480px

>>1019770
>Total newbie here. If I'm working with chips with a fuckton of pins (64) do I really have to connect all of them?
The spec sheet for the chip will usually say what to do with unused pins. Often different pins have different rules.
Often it will say to ground unused inputs.
Sheets for some devices will specifically tell you to leave unused outputs floating, or NOT to just ground them directly.

If I can't find a spec sheet, then my general assumption is that you ground all unused inputs and outputs through a ~10k to 20k resistor (each separate pin gets its own resistor).
...I've never had a spec sheet TELL me this, but it usually rules out any weird stuff happening.
Unconnected input pins can generate noise input (random values) and unconnected output pins can build up voltage that causes other odd problems.
Input resistors drain the input to zero but still block noise from entering.
Output resistors drain the outputs but still protect from an active output being shorted to ground.
Might be kinda aspie, but anyway.
If the [whatever] works like it's supposed to, I don't really care that it took a few more 2-cent resistors.

>>1019770
I was where you are at just a week ago.
Asked the same question.
Google "floating inputs".
Like >>1019827 said, depends on the chip.
For TTL logics inputs need to be tied high (logic 1). Some CMOS logics need to be tied low (logic 0).

This can be done by connecting the input pin to +V for high, or ground for low. Make sure to use a resistor suitable for the chip your using or you could overload the chip. Google pull-up/pull-down resistor (for TTL chips I find 5K-10K is a good value).

Floating inputs are bad in high frequency applications or when the logic needs to change states very quickly. It causes noise or erratic behavior.

I need a connector (or two, whatever) that can support a total of 7 pins, 4 of which must carry high current (30A+). Connector is on the backside of an instrument panel and can stick out quite a ways, but can't be too wide. Any recommendations?

File: Free-Shipping-5sets-New-font-b-30A-b-font-600V-Power-font-b-Connector-b-font.jpg (106KB, 450x452px) Image search: [Google]

106KB, 450x452px

>>1019939
30A is going to require a beefy connector, something like this is probably the smallest you're going to be able to go.

30A for mains is often a twist lock connector.

>>1019827
>unconnected output pins can build up voltage
...wtf man
Leave unused output pins alone. Inputs should be driven mainly to stop excessive current draw but also floating inputs can drive outputs randomly and cause noise where you weren't expecting it.
>>1019903
For TTL the current drawn by an OUTPUT is higher when at logic 1 than logic 0 so you should tie inputs accordingly, i.e. and gate= both high, xor gate= one high one low etc..
The only real problem 99.99% of people will have trouble with floating inputs is power consumption. If you have glitching from a floating input on a different gate then you need better smoothing on you power rail, that's the point of the capacitors everyone wanks themselves off about on every IC.

>>1019955
Good to know. I had to pull the input high on a 74LS04 (hex inverter). That would make the outputs low.

>>1017190
The ones I've used just want a connection to ground. Try that first.

I've got something of a solar system that floats between 11.7 and 14V on the battery, and I wanna power something a solid 12v no matter what, until a low point then shutdown.
I'd usually just cheat out and boost it to 24v then drop it to 12v, but we're talking something like 500-1000w, so power efficiency is critical here.
Just wondering what ideas come to mind.
(I do have a inverter, but I'd rather not.)

how hard is it to control 5v 20a with a <3.3v low current output of a pic32?

File: attiny85_unused_43n4jknjk4.jpg (274KB, 1200x1098px) Image search: [Google]

274KB, 1200x1098px

>>1019955
>...wtf man
>Leave unused output pins alone. Inputs should be driven mainly to stop excessive current draw but also floating inputs can drive outputs randomly and cause noise where you weren't expecting it.
witness

File: arduino.jpg (42KB, 620x400px)

42KB, 620x400px

I am trying to make a kill switch for my 1/8 RC car and i have everything all sorted but no way to kill the car yet. What is the easiest and safest way to do this with arduino. I was thinking of relay in-line with the throttle signal wire, is it safe to 'break/open' this connection and then reconnect while the car is moving. I would think it would just act like the transmitter lost signal... other ways would be to put the arduino before the receiver and have it pass signals to the ESC but i am not sure if throttle response would be affected. any help? its for brushless motor and car ESC.

>>1020158
You just need a power transistor (either a Darlington type, or a discrete Darlington pair). 3.3V isn't enough for a FET.

>>1020158
At that current, you're probably better off with a logic-level FET and a level shifter. You can get 20A+ BJTs, but both beta and Vce[sat] tend to be fairly crappy at max current (like, beta=10, meaning you'd need 2A base current).

>>1020171

I've got PNP darlingtons on hand. I can try to use them to make a power amplifier.

also I think i can do with limiting the current to 3 amps. Is their an easy way to do this without dissipating 3 amps?

>>1020190

*dissipating 15 watts, not 3 amps

>>1020190
> I've got PNP darlingtons on hand.
Not directly useful for switching 5V from a 3.3V output. Even if you set the outputs to high-impedance, there's usually static-protection diodes between the output and Vcc, so you'll need a level shifter.

> also I think i can do with limiting the current to 3 amps. Is their an easy way to do this without dissipating 3 amps?
Not unless you can use PWM. Either you saturate the transistor or you dissipate Vce*Ic as heat.

>>1020195

instead of a level shifter, could I just one of them tiny black transistors?

I think they are supposed to be easier to drive

>>1020067
no effect

I can get the LED to turn on if I wire the + pin to ground, the S pin to power (the - pin seems to have no effect)
which leads me to think that the pinouts are backwards
but plugging the - pin into power or ground still doesn't switch the relay
so I don't know.

File: driver.png (4KB, 482x484px) Image search: [Google]

4KB, 482x484px

>>1020196
You can have the logic output drive an NPN transistor whose collector is connected to the base of the PNP (Sziklai pair).

If the PNP is a Darlington, the required base current should be reasonably modest, so the NPN can be a low-power transistor which will typically have much higher beta (like, a couple of hundred).

If switching speed matters, connect a pull-up resistor from the PNP base to Vcc (BJTs turn off faster if you pull the base toward the emitter rather than just turning off the base current).

Be sure to limit the PNP's base current one way or another (it's best not to rely upon the NPN's beta for this).

File: sn74ls00.jpg (349KB, 1577x479px) Image search: [Google]

349KB, 1577x479px

>>1020169
That thing is a complex microcontroler that probably has internal pulldowns and pullups. It can also be programmed to put unconnected I/O pins into high impedance mode.

With simple logics (TTL or CMOS) you need to either ground or pull up (using a pull down or pull up resistor to prevent excessive current draw) to prevent floating inputs.

It has been said, "it depends on the chip".
If the chip in question hasn't been specified, we can't give a precise answer.

So I'm working on an audio project thats currently on breadboard with jumper wires, everytime I adjust the wires the ADC will read different value ranges making it unstable for my data acquisition.

Will it be better for now if I replace the jumper wires with solid core wire?

>>1020269
Breadboards aren't really the best option for anything requiring precision.
And if this ADC of yours is something used for audio (16b resolution or higher), layout in general becomes quite critical.

>>1020273
its an Arduino, so nothing precise. I just want a stable baseline signal but bumping the wires around changes it for some reason.

File: dmmtest_01_0459.jpg (123KB, 1200x548px)

123KB, 1200x548px

So recently I had two DMMs, but decided to order a couple more, and ended up with four more....
And then wondered how accurate they were, as they all give slightly-different values.
So then I ordered the voltage and LCR calibration things, and tested the meters.

Pic related, meters left to right:
#1 is the All-Sun DT830B cheapo POS everybody hates. I bought this recently as a loaner. It cost me $2.50.
#2 is a Velleman DVM850BL I've had for about 20 years now. I paid $60 for it, but they cost $15-$20 and there's China copies now for $7
#3 is a Extech EX420 I bought 8 yrs ago when I needed to count Hz. It's nicer than the Velleman but I rarely used it. I kept it packed in its case and usually grabbed the Velleman to use...
#4 is a Mega328 ESR meter bought recently. Not perfect but it costs very little and works pretty well.
#5 is a Mastech meter I bought recently because I wanted a meter that would test inductors. After ordering this I heard about #4, and ordered the ESR meter too.
#6 is a HoldPeak HP770D I bought because it seemed to have a lot of nifty features that other meters didn't have. And it does... but I didn't really "need" it, I guess.
#7 is a PeakMeter pen DMM I bought just for using continuity and dc-volts in digital circuits. I thought it might be handier than using a regular meter and it only cost $11.

Above the meters, the red (left) box is the LCR reference ($25) and the clear small box is the voltage reference ($15).

File: dmmtest_02_0460.jpg (44KB, 800x316px) Image search: [Google]

44KB, 800x316px

>>1020310
(continued)
Pic related: the LCR and voltage reference modules I got both had test stickers on them, showing how they read on an expensive bench meter.

The voltage reference chip has a max limit of only 10 mA.
It would be possible to run the 10 volts setting through the ~8.44 kΩ resistor, and that would give a DC current reference of around 1.18 mA. I didn't bother with that tho.

File: dmmtest_03_0461.jpg (108KB, 1423x303px) Image search: [Google]

108KB, 1423x303px

>>1020312
The end result: even the cheapest meter does pretty OK at low voltages.
None were far enough from the reference values to worry about the difference.

The HoldPeak is a 40K-count meter and has 5 digits, so it will usually show 1 more decimal point than most other meters, and sometimes 2 more decimal points.
My inner autist loves the idea but the difference of the extra digits often isn't that significant.

The ESR meter is probably the best way to test inductors and capacitors, since it tells you the capacitor ESR and the leakage and it is "accurate enough"...
My only complaints is that there's no documentation provided.
It's only got one button so you don't need "instructions" as such, but there was no info given on what the min and max testing limits are for the different component types and this is a common practice with these things.

>>1020169
Yes in a microcontroller pins default to inputs on reset so its a bit of a corner case. Ideally it would allow you to set a fuse for an internal pull-up that was enabled during reset.
>>1019965
I said that wrong, it draws slightly less current when the output is high not more. But then do the inputs draw more than the output does? I guess you should just check it yourself really.

>>1015412
400 dollars for a decent used scope. Get 100 mHz dual channels ar least. Go analog. I like 80s era Tektronix.

Power supply... get a 20 dollar 2 amp from china. Pretty stable and does what you need. Get two you will find yourself needing least two different voltages at times.

As far as function generator just get a cheap dds board from china, some that does a sine and square to maybe 1 ghz.

What you missed is a decent multimeter. Get a used fluke on ebay. 87V will do most of what you need. 77 and 27s will do too. Stay away from the radio shack special.

I repair test equipment for a living, I am telling you a multimeter is the first thing you want to buy. Spend 200 on a used fluke on ebay.

>>1013631
What do you mean that mess up but cool

I need to fix my software

File: res.png (602KB, 702x522px) Image search: [Google]

602KB, 702x522px

>>1018525
>>1016683
>At this point I'd just assume that ALL of the blue ones sold as 1% metal film are fakes.

just got these in the mail:
>>1016686
same ones; blue 1/4watt "1%" resistors
measured a bunch with my fluke and they're more like 5%
they don't seem to be effected by just body temperature. haven't tried hot air.
I'm pretty happy with them

out of stock now but here's the link anyway
http://www.aliexpress.com/item/r/32576054669.html

>>1020207
Imma try on one I have, just a sec...
Wiring backwards has no effect, I could swear I remember hooking up two leads once and a LED lighting without flicking the relay, but bleh. If connecting the S pin to ground and 5v doesn't work, and you've tried both while swapping the positive and negative, I dunno what to tell you.
It could be a dead transistor, if you upload a picture of the bottom of YOUR board I'll probably be able to work it out.

>>1019591
what caps?

File: half.png (22KB, 1568x660px) Image search: [Google]

22KB, 1568x660px

trying to simulate a half wave rectifier but not waveforms

the blue curve is the filter cap but there is quite a drop when the input voltage goes zero. the line should just be going down slowly until the next cycle but that isnt happening

I think it may be a problem with the diode model?

>>1020643
but the waveforms arent right***

english problems

>>1020643
Does it still happen if you halve the sine amplitude?

>>1020612
6.3v 1800μF

The closest things anyone in town had were caps for repairing TVs and monitors.

Although come to think of it, I might be able to dig through the dumpster of a computer repair place for a MOBO or three.

File: halfinput.png (23KB, 1591x628px)

23KB, 1591x628px

>>1020649
halving the input, the voltage now looks normal

but i think the problem is still in the diode models but idk where to find the right model for 1n4001. I tried 3 different models and each give different results

>>1020654
1n4001 is rated for 50V. In other words, the maximum peak input voltage for a half-wave rectifier using it would be 25V.
Apparently your simulator is modeling its reverse breakdown.

>>1020536

> go analog

Don't go analog unless you have to. For 400 bucks you can get a decent semi-modern digital scope for sure. Go as modern as possible.

Honestly 400 bucks is enough to get a modern-ish scope and power supply. You probably don't need a function generator because it is a more niche tool.

File: IMG_1422.jpg (139KB, 640x480px) Image search: [Google]

139KB, 640x480px

>Hey I posted on /g/, but this seems to be a better thread. So I'll just copy pasta my OP.

Hello, I'm a bit of a moron doing a fun summer project. In a nutshell I want to get these lights to sync with music, the trouble is I cant figure out how to turn them on.

In the picture to the left you can see my black power source, which can deliver up to 5 volts

On the bread board I just have some jumpers that attach to the lights positive and negative voltages. As well as three MOSFETs.

My problem is whenever I plug in the voltages from my power source to the lights voltages, my power source cuts off telling me I am drawing too much current. I understand I should have a resistor of some sort then, but how much?

Im using the following tutorial here:

>http://popoklopsi.github.io/RaspberryPi-LedStrip/#!/

Technical information for the lights here:

>https://www.adafruit.com/products/1460?gclid=CJLLoLe2oskCFQH3HwodrIUFHw

any comments or advice appreciated, thanks.

just wanted to comment that my power supply is probably not enough (in the tutorial it states I need 12V) is there any chance I could use an old computer adapter for this and some other part?

>>1020695
I'm no technical genius but according to the link, the led strip requires roughly 2 amps of current and from your story I assume that your power supply cannot deliver those amps.
So you going to need a more powerful power supply.

The 5 volts is fine but it won't work if the current is lacking.

>>1020695
The LED strip you bought are very different from the one in the tutorial. The tutorial uses a simple 12V RGB strip (Just resistors and RGB LEDs). What you bought is a 5V addressable LED strip. It's got digital electronics for controlling the LEDs built into it. Connect the LED strip's digital in pin to your RPI's SPI data out. Send a set of bytes representing the RGB values of each LED.

>>1020690
You can get a used analog much cheaper than a new digital. Calling a function generator niche is like calling a circular saw a niche tool for a carpenter. While a novice will find little use to start the faster they learn to use it the more they will be able to do. Get the multimeter first though.

>>1018741
You would be better off getting a used one on ebay. 20khz bandwidth will get you thought most audio. Soon at you start to look at digital circuits you are going to wish you had not wasted money on a toy. Get at least a 100 mHz dual channel. They go for about 100 dollars used on ebay.

>>1017661
Metrologist fag here. I work for a calibration lab. The Wavetek is out of calibration. Based on what is more likely to drifft base on age, manufacturer, equipment type and models. You can most likely find the procedure to cal the Wavetek online and just tweak in the frequency. Do not change anything else unless it is out of tolerance. No matter how much you think you can make it run better. Always takes everything out of tolerance. Never tweak the power supply in the signal generator unless you absolutely have to. That will also take everything else out of whack.

>>1017711
I agree with doing this before you take my advice too.

>>1020861
Yes, but one should be careful buying a used analog scope: make 100% sure it works. 20+ year old scopes will drift out of spec, caps will be going bad, etc. You can fix them, but it's not what someone just getting into serious electronics is going to be ready for. If you want to use it as a tool (as opposed to collecting), don't pay more than 100USD, and I would always try to buy local (vibration in shipping can be bad) and have them show you it operation before you buy.

Still, the features that even a entry-level, budget, DSO offer are immensely advantageous. If you can at all spring 400$+ get a DSO. CRO are antiques these days, they will get you by on a budget, but you're eventually going to want to invest in something better.

Function generators are also a must, I'm currently looking for something myself. Trying to find something better than this 30$ bare board that goes up to 5MHz but isn't 400USD+.

For power supplies, check ebay/craigslist. I got a HP6214C last Friday for 25$! A little patience and you can find some good deals.

I will second getting a DMM first, in fact get TWO. Don't need to spend a ton, don't need a bench meter to start either. Budget 50-100 per DMM. Two meters lets you check measurements against each other to see if they are drifting and you often will want to be able to measure current and voltage: you need two meters.

>>1020588
>same ones; blue 1/4watt "1%" resistors
>measured a bunch with my fluke and they're more like 5%
get a wire brush wheel on a dremel tool and sand the paint off a couple, then compare them to pics online

when I searched the matter with Google, I didn't see anyone claiming that they bought the cheap Chinese ones and actually got 1% metal films

>>1020888
This is good advice. Many sellers will show it with a signal on the screen to show it is fucntional. As far as buying an analog scope.

One bit of advice as far as DMM if you are going to get a low cost dmm (extech, radio shack, reed instrument, etc) make sure to check the specs as far as tolerances. Some will be as much as 3 to 5 % of reading in AC ranges. At 5% 100 volts that reading could be between 95 or 105 volts and the reading is still in manufacturers specs. Also get one with a separate low and high current connectors. This means that you are using a separate circuit in each range and get the a good reading.

Are there any good places to find junked PCBs these days? I've been trying to find some to scavenge components from but everywhere just fucking recycles them now and even the recycling places just bundle them and ship them off.

>>1020903
Craigslist, yardsales, and ye old dumpster diving. I found a gold mine of equipment out of a cellphone tower once.