the /ohm/ thread

Can I use a solder blob on a protoboard as a bus for 300mA 1/8 duty cycle safely?

File: indian pcb prototype.jpg (78KB, 451x338px) Image search: [Google]

78KB, 451x338px

>>1038445
>blob
If we are talking "lengthy" bus connections above 2-3 holes, just lay down a stripped wire first. Blobs are a bitch to work with on a distance.

>>1038484
What do you mean "first?" I like the stripped wire idea but I'm not totally clear on how to implement it.

File: clock_solder_400x507.jpg (61KB, 400x507px) Image search: [Google]

61KB, 400x507px

>>1038485
Just solder down a solid core wire sans isolation at one point, then gradually solder it down point for point. That way the solder will flow nicely between the gaps without looking too blobby. See pic related for a sleek approach where only every other point is held down by solder.

>>1038489
Cool, thanks much.

This is a bit more /g/, but they're too busy circlejerking over phone themes to answer. You know how apartments use those little smart cards to operate laundry machines now? What if I just bought a reader/writer for one of those things, loaded it up to 10 bucks, read it and saved the data, then just erased and re-wrote it every time I need to clean my khakis? Infinite laundry?

>order crap off dx.com
>says 24 hour shipping, dont believe them but chose the free communist subsidized shipping anyway
>a month later, still aint here
Anyone hear of a cargo ship getting sunk by pirates lately, or am I getting fucked?

>>1038499
Sure, laundry's free in pound-me-in-the-ass federal prison.

>>1038502
Unless the machines are hooked up to the internet or a phone line, and the card is just an account number, they'd never catch me.

File: stock-photo-angry-convict-prisoner-jailbird-behind-bars-137722457.jpg (520KB, 1500x1128px) Image search: [Google]

520KB, 1500x1128px

>>1038503
>they'd never catch me!

>>1038501
>says 24 hour shipping,
They ship in 24 hours but it may take six weeks for you to get it.
Also, could be stuck in customs somewhere - especially if your order included lithium batteries.

>>1038511
Last thing the shipping tracker said was "Left Shenzhen Center". A month ago.

File: 58455444[1].jpg (223KB, 500x750px) Image search: [Google]

223KB, 500x750px

>>1038503
>they'd never catch me

Be sure you include AdBlocker in the data...

>>1038512
I've bought quite a bit of stuff from DX and never had a problem.
They have 'contact us' links on the site so you can report a missing/lost shipment.
I'm sure they'll make it right once they find out what the status is.

>>1038503

Most of them ARE networked and just use the card as an account, specifically because it's trivially easy to do exactly what you want to do.

>>1038519

>so i'll just copy one of my neighbors account info and let them get billed for it

Are resistors into the gate or from gate to ground on an n-channel MOSFET necessary? It's pretty low power (5V gate, 12Vds 300mA 1/7 duty cycle.)

>>1038535
> Are resistors into the gate or from gate to ground on an n-channel MOSFET necessary?
It depends.

Resistors into the gate may sometimes be necessary to damp ringing from the parasitic L-C circuit. This is more likely to be an issue if the package is a through-hole type and/or the trace connecting the drive signal to the gate is long (i.e. significant parasitic inductance). It's also more of an issue if the gate drive voltage is close to the rated maximum (i.e. if overshoot is likely to cause Vgs[max] to be exceeded).

A resistor between gate and ground is necessary if an "off" signal is (or may be) high impedance (open-circuit) rather than sinking current to ground. Gate capacitance won't discharge itself.

>>1038532
Thats actually a pretty fucking good idea mate

>>1038499

You know magnetic strips are really not that hard to read

I saw a defcon talk where I dude claimed you could read the binary off of one with iron filings.

And if that doesn't work, I bet you could hack a reader together with an inductor and an oscilloscope

>>1038589
Writing them so that the official reader is happy is the more tedious part.
Not impossible of course, but how much time and money you're willing to spend to fuck with a laundromat?

>>1038535
Other anon mentioned ringing reduction. This is generally a problem with big mosfets and big, fast gate drivers.
With smaller drivers (processor IO pins etc.) one reason to use series resistors is to reduce the peak gate current. Another is to limit the current spikes coming back from the fet due to gate-drain capacitance.
Ground resistor is usually added to keep the gate at defined state when the driver is in undefined state or floating. Processor IO pins typically float during reset, for example.

File: file.png (50KB, 1956x680px) Image search: [Google]

50KB, 1956x680px

I'm having a hard time figuring out the switching for the LEDs in my audio spectrum analyzer design. I have a chip that does the FFT for each band and multiplexes the output, with each band being one column on the display.

The hard part is that the chip which takes the signal from the FFT chip has ten open collectors as the output, with each one controlling one row of LEDs. The chip that decides which column is being controlled by an LM3914 at a time has a current output that's way too low to power the LEDs. I can't use that signal to switch a FET before the load because the source voltage is higher than the signal, and if I did after the load I'd need twenty times as many.

If it'll help, I'll upload the full LTSpice schematic, and I'll be around to answer questions. Please, help me. I've come to far with this to fail now.

>>1038602
Corrections:

too* far

The two sets of LEDs and resistors labelled "1 band" should have the supply voltage connected. Same for the other two.

>>1038589
They use those new chips they put on credit cards, no magnetic strips. So I'd have to order a reader from the chinks.

Just checked my mail. The month and a half long wait for stuff from DX is over.

>>1038607

what a high tech laundromat

>>1038602
Jesus that must be one massive LTspice schematic. Please post it.

>>1038630

file dropper dot com slash spectrumanalyzer

Haven't used that site before but after a few tries at the CAPTCHA it worked. I'll rehost on google drive or something if it doesn't work for you.

The ETC means there are 10 more bands, 600 LEDs worth, that I didn't feel the need to put in the schematic.

I got up to the point where I realized there's a serious problem, then stopped adding stuff. The two FETs' drains were supposed to be attached to a 12V supply.

>>1038602
>>1038637
So basically you want the LM3914 to do your bargraph effect but it doesn't have enough current capability? Since the 3914 outputs are open collector, use them with load resistors to get levels for driving a P MOSFET gate, which you now put "above" the LEDs. Then you move the N MOSFETs to "below" the LEDs.

>>1038609
>Just checked my mail. The month and a half long wait for stuff from DX is over.

all that gettin' antsy over nothing...

File: tmp_31380-hqdefault(2)-1047758680.jpg (11KB, 480x360px) Image search: [Google]

11KB, 480x360px

How could I make an efficient lab bench power supply controlled by an atmega328 (arduino)?
I was thinking about buying a chunky transformer or fixed output switching PSU and use a linear regulator like a LM338 or 317 and transistors controled by the uC to set the voltage and the current limit, this however, seems quite inefficient, and I was wondering if I could build a switching psu based on the microcontroller?
Any tips or documentation/tutorials which can help me build the thing?
Thank you.

>>1038810
>this however, seems quite inefficient
Still status-quo at the chain end, since the trade off is stability over efficiency with linear regulators.

If this may not be a concern, you could easily implement a switching PSU easily using almost any µC, since they all come with ADC and PWM peripherals nowadays.

Dave Jones has you covered with a good introduction on PSU design
→https://www.youtube.com/watch?v=CIGjActDeoM

>>1038817
Thanks, I'll look into that, I remember having seen a couple of videos, but not the whole series. If I can't make a switching PSU I'll probably end getting my hands on two or three chunky transformers and use relays so the uC can switch between them at different voltages so the regulators run cool and efficiently.

>>1038810
Well, the "efficient" requirement rules out linear regulators.

You'll probably want to use an off-the-shelf fixed-voltage supply followed by a uC-controlled buck regulator. Isolated topologies and (well-regulated) variable-voltage don't really mix; you can do it, but it's significantly more complex than either a fixed-voltage isolated or variable-voltage non-isolated supply.

For a uC-controlled switcher, use the uC's PWM unit with the duty cycle controlled by software feedback. This avoids the need to run the feedback loop at the switching frequency.

For the fixed-voltage supply, you want something that will run unloaded, but it doesn't need to have good regulation when unloaded

Does anyone know of relatively inexpensive, high-performance, yet simple microcontrollers on the market now?

I know for low performance, there's a broad range of very cheap AVRs that vendors are lining up to sell you in any quantity. For somewhat higher performance, you can go to the Cortex M-series, and get stuff in the neighborhood of 100 Mhz, and they're still cheap and easy to get.

For gigahertz performance, it seems that people go to boards, which have operating systems on them.

Are there chips as simple to use in hardware as AVRs, but simply run a hundred times faster?

>>1038885
Depending on what you consider "efficient", linear regulators can be really nice. I can use 16-18V input for example and get 15V out easily for a single regulator, including a high amp capacity. more than 80% efficient across nearly all the current range, even if it dropped three volts. A simpler 18->15 switch mode supply would need far more specific parts, have an explicitly designed current capacity, and be largely unadjustable. You can get them 80-90% efficient, but that requires a lot of tuning for specific conditions.

File: segmentTest.jpg (370KB, 1331x998px) Image search: [Google]

370KB, 1331x998px

I see most people setting up 7-segment displays with a resistor on each pin. I only had one suitable resistor, so I put it on the common pin.

Is there any problem with multiplexing the segments at 1ms on/ 1ms off? That's the current setup, and aside from being a little work to set up, it seems to be working fine.

Does anybody feel a certain attraction towards electrolytic capacitors? Its almost sexual, as if you just wanted to lick it or eat it in one bite.

What did I mean by this?

>>1039106
You probably shouldn't post about things you don't understand in the slightest. Like switching regulators.

> I can use 16-18V input for example and get 15V out easily for a single regulator, including a high amp capacity. more than 80% efficient across nearly all the current range, even if it dropped three volts.
IOW, linear isn't inefficient if the regulator drops hardly any voltage. But the question was about a variable-voltage supply, so that's irrelevant here.

The efficiency of a linear regulator is almost exactly Vout/Vin. E.g. 5V from 12V will be a fraction under 41.666...% efficient. For a variable-voltage supply, the input voltage has to be higher than the highest desired output, which translates to shitty efficiency at lower voltages.

Whereas the efficiency of a switching regulator is ... whatever you're willing to pay for in terms of parts cost and design complexity. You can get better than 99% if you really need it. There's really no reason to get less than 90% unless it's incredibly price-sensitive (i.e. no corner left uncut).

> A simpler 18->15 switch mode supply would need far more specific parts, have an explicitly designed current capacity, and be largely unadjustable.
This is nonsense. All of the common non-isolated topologies are trivially adjustable; you just change the division ratio in the feedback loop.

> You can get them 80-90% efficient, but that requires a lot of tuning for specific conditions.
This is also nonsense.

To get less than 90% efficiency from a switching regulator. you have to either screw up badly or simply not care about efficiency. And they don't require "tuning".

>>1039139
It's a current-limiting resistor.

If you put one on each segment's pin, the current will be limited for each segment individually, and you'll get consistent brightness no matter how many segments you activate.

If you put one on the common pin, the total current will be limited for all the segments combined, and the brightness will vary depending on how many segments are activated, and might not be equal between activated segments.

>Is there any problem with multiplexing the segments at 1ms on/ 1ms off?
If you mean "1 ms on / 6 ms off", then no. That should work fine, although it will be at most one-seventh as bright as the display can handle.

>>1039139
> Is there any problem with multiplexing the segments at 1ms on/ 1ms off?
Well, a 50% duty cycle only lets you have two segments.

>>1039146
>>1039147
Sorry, I guess I described it wrong. It cycles through each segment by lighting for 1ms then delaying for 1ms before lighting the next for 1ms, etc

>>1039146
Yeah I've got them all lit at the same brightness this way. Wasn't sure if it will kill the display/resistor very quickly or if it's safe

>>1039145
I'm going by a switched-mode supply based on the MC34063 http://www.onsemi.com/pub_link/Collateral/MC34063A-D.PDF

What kind of designs have you seen where a SMPS doesn't require having +/-10% limits or better, or uses more affordable components compared to a 25c regulator, 10c pot, and a heatsink from anything? And I don't mean theoretical designs, I want to see an actual SMPS for cheap that can go between 3-24V with at least an amp capacity.

>>1039149
500 pulses per second = ~70 pulses per number. Works fine, but 70 pulses will be on the edge of flickering if you turn your head quickly. You'd need at least 2khz to do four numbers, and the device which would appropriately know whether a segment was on or off with the same synchronicity.

>>1039149
>Wasn't sure if it will kill the display/resistor very quickly or if it's safe
As long as it's enough resistance for one segment to be continuously lit without harm, it's definitely safe. There's no way for more current to flow through in total.

>It cycles through each segment by lighting for 1ms then delaying for 1ms before lighting the next for 1ms, etc
There's no point in having that idle millisecond in there, unless you prefer for it to be dimmer. It'll cut the brightness in half.

>>1039151
> What kind of designs have you seen where a SMPS doesn't require having +/-10% limits or better,
"+/-10% limits"?

> or uses more affordable components compared to a 25c regulator,
A switching regulator will be more expensive until either the current is at the point where the linear regulator needs a substantial heatsink while the switcher doesn't, or you need a discrete transistor for current boost (which will need to be larger for a linear regulator, due either to increased power dissipation or the fact that a step-down switcher has Iin<Iout, or both).

Note that this is for a DC-DC converter. For a SMPS (i.e. mains-to-DC converter using an isolated topology), the reduction in transformer size for a switcher outweighs everything else once you exceed ~5W.

> an actual SMPS for cheap that can go between 3-24V with at least an amp capacity.
The 34063 will do that with an external transistor (using the internal transistor on the high side will dissipate too much power at 1A due to non-saturation). At low voltage, the diode drop starts to matter (at 1A, that's only ~200mW; at >10A you'd consider using synchronous rectification).

Whereas a linear regulator doing 3-24V @1A needs >24V input, so at 3V you're dropping 21V@1A=21W, so thermal management is going to be an issue. A 34063 with an external transistor could do that with less than 1W dissipated.

>>1039212
Look at the damn calculations you fuck. If you change any of the expected values (like Vout) all the subsequent circuit values change (C, L, and Rsc).

This paper has some real-world not-manufacturer-magic results:
http://www.onsemi.com/pub_link/Collateral/AN920-D.PDF

Most of the designs are around 80% efficient, and to get higher amps, they have to add more components. A linear regulator is the chip and a resistor, and something like the LT3081 can be put in parallel easily for many amp outputs, and can even be assigned voltage output and current limits (they can be set by an analog signal on specific pins).

Yes it's less efficient, and sometimes catastrophically so, but that poster probably isn't going to be drawing ten amps. What even is an application where you'd want 3V 10A from a lab power supply.

>>1039245
Also, because you didn't read what I had put, I went and found myself an adjustable switcher:

http://www.onsemi.com/pub_link/Collateral/LM2575-D.PDF

At 12->5V 1A it's 77% efficient, but they don't go into efficiency or limits/current after adjusting. It is cheap.

>>1039151
>And I don't mean theoretical designs, I want to see an actual SMPS for cheap that can go between 3-24V with at least an amp capacity.

Not going to jump between you to, but, seriously...

http://www.ebay.com/itm/DC-DC-boost-converter-Constant-Current-Mobile-Power-supply-250W-10A-LED-Driver-/181940997254?hash=item2a5c874c86:g:ZlsAAOSwxN5WVAXJ

$5. Voltage AND current controlled. I actually bought a few to run some "100W" LEDs. Never actually checked the efficiency and I doubt the components are exactly top quality, but the output is fairly clean (at least with the load I was using) and it's made on a clad-aluminum PCB.

To more closely approximate the stated requirements:

http://www.ebay.com/itm/Boost-Buck-DC-Adjustable-Step-Up-Down-Converter-XL6009-Module-Voltage-LI-/182122452625?hash=item2a67581691:g:zm4AAOSwvt1WRCJb

Or...

http://www.ebay.com/itm/LM2596-DC-DC-Adjustable-Buck-Converter-Step-Down-Module-Power-Supply-1-23V-30V-/262356694546?hash=item3d15ad7612:g:o70AAOSwv0tVFO8c

A dollar. A CANADIAN dollar, no less. I probably wouldn't try it at 3A (they can get plenty hot at just 2), but it certainly works.

Also LMAO @ getting efficiency numbers from a datasheet for a single chip.

>>1039029
If people need 100 times more performance in general, it tends to mean that they also need peripherals which are suitable for that much power - which in turn makes the processors more difficult to use.
On the other hand, if you need that 100 times performance only for one single simple task, then there's a good chance that dedicated hardware or CPLD/FPGA is better way to handle it.

>>1038861
I don't know about your requirements, but personally I'd never take a switcher for my primary lab supply.

Are there any DIP ARM microcontrollers? I know about the NXP LPC1114, but all the major distributors are out of stock now, and it's probably never coming back.

>>1039256
I appreciate the thought, but I was asking for solid information on what high-performance microcontrollers are available, not vague thoughts and opinions on why what some situation that might exist should exist.

Does anyone know of any simple microcontrollers that are easy to include on boards and program without an OS at gigahertz-level performance, or feel they have a good understanding of what is and isn't available and can say what the highest performance options are?

>>1039258
Digikey is great for doing this kind of search.
The only DIP ARM they have in stock is LPC810M021FN8FP

>>1039029
You can firmware program any microcontroller board. You don't have to install an OS and write applications on top of that if you don't want to.

Is a BLDC motor just an AC motor with integrated controller/inverter to be driven from DC, or are there other differences too?

>>1039245
Okay, you don't understand the basic theory of how a switching regulator works, so I'll explain. it's a feedback loop. You vary the duty cycle (either linearly or simply skipping pulses) according to whether the output voltage is too low or too high. Any given selection of component values gives you a maximum voltage and a maximum current at any given voltage, but you can obtain any voltage between zero and the maximum by simply reducing the duty cycle.

You can make a functioning buck converter using just a comparator to control the transistor voltage too low -> turn on, voltage too high -> turn off. The output voltage will match the reference voltage on the comparator's non-inverting input. The only requirement is that the comparator has enough hysteresis that when the output voltage is ideal, you get oscillation rather than the transistor being operated in the linear region. The main reason this design isn't used much (although it /is/ used) is that the switching frequency varies with load; this makes it harder to filter switching noise compared to a fixed-frequency design.

Component selection is largely economics: you don't use larger values than you need to because they cost more and occupy more space. The only hard constraint is that the ratio of C/L needs to be large enough that dumping all of the inductor's stored energy into the capacitor (i.e. if the load current goes from maximum to zero instantaneously) doesn't result in too much overshoot.

>>1039276
> Is a BLDC motor just an AC motor with integrated controller/inverter to be driven from DC
Sort of. Except the controller isn't necessarily integrated.

Typically, it means a polyphase AC motor designed to be driven from an electronic speed controller (ESC) rather than with a 50/60 Hz sine wave.

> or are there other differences too?
Some of them have feedback outputs so that the controller can monitor the rotation. But for 3-phase motors, this isn't necessary; you can just monitor the back-EMF on the phase which isn't being driven.

I recently taught myself how to solder because I wanted to do some things with the Raspberry Pi and a broken NES. I purchased those assembly kits to practice and I was surprised by how much I enjoyed the process.

Question is, is it worth it to learn the theory behind all this? Should I read the books? I have no interest in working in the field, only making fun projects here and there.

What do you guys think?

>>1039288
If you need to put out this question - probably no.

>have no interest in working in the field
There, you answered it yourself pretty clearly.
Developing something new is work, even as a hobby.
What you consider fun is only the last step of assembly after finally getting your project to work.

>>1039279
>Some of them have feedback outputs so that the controller can monitor the rotation. But for 3-phase motors, this isn't necessary; you can just monitor the back-EMF on the phase which isn't being driven.
The RC ones all seem to use feedback, but all the CNC ones I've seen had Hall sensors in them.

I dunno about non-RC little ones, which is what I'm asking about....

I need a small motor to spin a ~1" thin metal disk in air, at 240 rpms. The metal disk has holes that serve as an optical shutter, so there's not really much of any mechanical "load" on the motor.

The motor needs to turn as smoothly as possible, so I am supposing that a brushless motor would work better than a brushed-type.

I've found these little coreless motors used for the tiny quadcopters, how do you drive these exactly? They appear to have only two wires.
Is it like a PC fan, where there is a circuit built-in and you just apply variable-DC to control the speed you want?

I know for the little appliance-grade stepper gear-motors, the driver boards almost always use a ULN2003 chip. Is there a chip like that for driving small 3-phase brushless motors? Or do little brushless motors just always come with their own built-in controller?

I search for 'mini brushless motor' and caint find shit, all I get is quadcopter parts results. :P
Kinda like how if you search ebay for hot-wire/resistive wire cutting parts, all you get is vaping shit now

>>1039288
>I enjoyed the process.
>is it worth it to learn the theory behind all this?
Yes, to a limited extent.

Don't jump straight in on the heavy mathematical side, like an engineering student is made to. Learn useful bits of theory you can play with right away. Browse the books (start with the beginner ones and work your way up to the advanced ones, don't start with the advanced ones just because you want to eventually have advanced knowledge). Study the stuff you find interesting, and the stuff that seems to be in the way of things you want to understand.

You might want to tackle the hard math at some point, but it really doesn't matter much for most people, other than that someone, somewhere is doing it to optimize the components and figure out the formulas, calculators, and charts you can plug numbers into. Most practical work can be done with arithmetic and a little algebra and trig.

>>1039288
>Question is, is it worth it to learn the theory behind all this? Should I read the books? I have no interest in working in the field, only making fun projects here and there.
Trying to learn all the theory for just a hobby is a rather boring way to go about it, IMO.
This is a hobby; you can skip to the interesting stuff. You don't gotta pass a test first.
>What do you guys think?
1. Skim through a basic book on electronics. Memorizing the whole thing isn't necessary.
2. Pay particular attention to Ohm's law, you use it a lot.
3. Understand how to use a multimeter to measure volts / ohms / amps properly. (read the instructions!)

If you come to something you don't understand you can always look it up online.

>>1039511
> The motor needs to turn as smoothly as possible, so I am supposing that a brushless motor would work better than a brushed-type.
There probably isn't much difference in "smoothness", but it's easier to drive a BLDC motor at a specific speed than a (mechanically-commutated) DC motor.

With a lightly-loaded BLDC motor, you can just feed it fixed-frequency 3-phase, and so long as the frequency and voltage are vaguely in the right ballpark, it will spin exactly in sync.

Feedback only matters if you're pushing it to the limit or you need to vary the speed quickly.

> I know for the little appliance-grade stepper gear-motors, the driver boards almost always use a ULN2003 chip. Is there a chip like that for driving small 3-phase brushless motors?
There are a lot of such ICs; google "3-Phase Motor Driver IC". Or you can use a microcontroller.

> Or do little brushless motors just always come with their own built-in controller?
Bare motors are probably more common. Ones with integrated controllers are less useful if you want to do anything other than just make it spin at a (roughly) constant speed.

One point: 240 RPM (4 Hz) is rather slow. You could probably get a stepper to spin that fast. If you needed any torque, you'd find something much faster and gear it down. Most (non-stepper) motors will be designed for much higher speeds; which doesn't really matter except that you'll get negligible back-EMF from it, so you'll probably need some form of current limiting.

File: NixieBoard.jpg (112KB, 1250x857px) Image search: [Google]

112KB, 1250x857px

This is most likely what I'm going with. Any suggestions? I'm a newbie at electronics so I'm looking for best practice type of advice and what not. Thanks guys.

Well, maybe is the wrong thread, tell me if so, I'm VERY new at electronics, but I always wanted to go on it.

I see lot of projects with arduino and that kind of boards. They have something related to this kind of electronics?

I can see some components and stuff, but I don't really know anything beside the fact arduino is very flexible at hardware, or at least more than other little boards.

File: 18vh1a.gif (2MB, 360x202px) Image search: [Google]

2MB, 360x202px

>>1039692

Check out my post above yours. The microcontroller is what is at the heart of the Arduino. There is much you can control with it from physical motors to LEDs and high voltage drivers.

Made this with an Arduino and 4 shift registers.

>>1039700
So, it is still mostly software?

File: road-gravel_0.jpg (186KB, 504x378px) Image search: [Google]

186KB, 504x378px

Does anyone know of a method of detecting clumps of gravel in shallow sand electronically? I was thinking maybe there's a temperature difference that a thermal camera could pick up but I don't know.

>>1039692
>I can see some components and stuff, but I don't really know anything beside the fact arduino is very flexible at hardware, or at least more than other little boards.
An Arduino is a microprocessor development board. It has a microprocessor chip and some other stuff to allow connecting power and a couple ways to program it, but it doesn't really "do" anything on its own. You have to add parts to get it to do stuff.

Arduinos aren't "more flexible" than other microprocessor development boards; Arduinos just became very popular very fast because (in my opinion--the MAIN reason) Arduinos are made to be especially easy to use.
(Most) Arduinos connect to a PC by a USB cable, and you only need to download & install one program (the Arduino IDE software) to write programs for arduino boards.
Most other dev boards are aimed at professional users and aren't nearly as easy to use.

Another reason for Arduino popularity is that--depending on where you live--you may be able to mail-order China-made copies of arduino boards for very cheap prices.

Arduino processors are really pretty slow and don't have a lot of memory; there is other boards that run way faster and/or have more memory,,, but then again, most of the things that 'normal' people think up aren't really technically demanding tasks.

>>1039692

arduino = little computer built into a chip on a board

you can program it in a dumbed down version of C

the outputs are digital

you can also sample analog signals with the ADC

>>1039745
>>1039747
Then, Arduino is not more oriented to hardware than RPi or even Odroid? Curious, I was very wrong. I though arduino was used because it let some flexibility to the time of doing electrical things (that's why I'm here)

>>1039770

Arduino is really for people who want to automate things, or output digital signals the easiest way possible. The board is really slow and really big but really easy to work with.

Another option is to get a cheap usb programmer and build your own board. You really only have to make a couple connections and use less than 10 external components. the benefit of this is you can write C instead of arduino shit and it is cheaper. This can be a pain in the ass.

Also, Atmel (microchip now?) sells this "Xplained mini" board with the same chip as the Arduino. It is really cheap (<15 bucks) and it is supported by the Atmel Studio IDE, which is really useful. It has full debugging support which makes it easier to program C. It lets you view the control registers which are very confusing for someone who hasn't ever programmed a microcontroller before, but with this you will be able to figure it out if you are persistent enough.

>>1039770
arduino is much more hardware oriented than rpi. it's more fundamentally targeted towards hobby projects where you need to read sensors or basic inputs such as buttons, perform some logical operation, and then apply an output to some actuator or relay or LED. that's simplifying of course.

the rpi on the other hand is what you'd use if you need to drive a display or manage internet connectivity. it's for higher level programming. it can do GPIO but you'd be wasting its potential if you used it like an arduino.

generally speaking if you need to make an electromechanical system and don't really need desktop style processing you want an arduino/

>>1039770
The original Arduino boards only have an 8-bit CPU with 1K of RAM, which means you can't exactly use them as a personal computer.

Also, the essence of the "Arduino" project is less about a specific CPU and more about the interfacing, allowing you to connect "shields" (peripheral circuits controlled by the CPU).

While the Pi does have GPIO pins, it's capable of being used as a general-purpose computer.

tl;dr: RPis are intended to be programmed, Arduinos are intended to be connected to things.

>>1039778
>>1039780
This explains better what I tried to say, sorry for my bad english.

It is recommended for a very starter in electronics but with kinda knowledge on computers to start with arduino and peripherals?

>>1039770
I'd say Arduino's much more suitable for getting started in basic electronics. It (mostly) uses 8-bit AVR microcontrollers with no operating system, so it's really not doing anything but exactly what you tell it to.

If you're not super price-sensitive, and have no experience with electronics, you might want to look at Ruggeduino. That's an Arduino clone which is made to be tolerant of bad circuits hooked up to the pins.

When you get a little farther along, ATtiny85 is neat. That's a little 8-pin DIP you can program with Arduino software (as you can with many other AVR chips), so it can run a complex control program, but it this tiny thing like a 555.

This is largely unrelated, but another thing that's helpful for the beginner is to pair up some electrolytic capacitors of the same size and connect them negative-to-negative (or positive-to-positive, it doesn't matter). You end up with only a quarter of the capacitance you'd have by hooking them up in parallel (two 1 uFs together make one 0.5 uF this way), but now it's non-polar (can take a charge either way) and acts much more like an ideal capacitor.

This isn't something you should be able to guess from learning about capacitors, but it works due to some weirdness about the internal chemistry and physics of electrolytic capacitors specifically, and they're much more appropriate for beginning students.

Of course, experienced circuit designers only use this in special circumstances, because it normally costs at least twice as much.

>>1039782

it doesn't use an OS but it has a bootloader which really makes it trivial to program IMO

>>1039782
I'm learning a lot, very thanks guys. I always wanted to stick my head into this but I always start to search and stops inmediatly because I feel like everything is overwhelming and that I'm searching wrong. I hope OP has enough for me, or I have enough knowledge to start from here.

>>1039781
others here might disagree but arduino is an excellent way to start with electronics. the arduino IDE abstracts more confusing concepts like registers with comfy functions.

if you use the arduino ide, you write
>pinMode(3, OUTPUT);
instead of
>DDRC |= _BV(DDC3);

the hardware is still fragile though so don't connect things backwards or draw too much current from pins or from the regulator. also don't fuck around with inductive components like relays or motors or solenoids without reading up on them first. everyone fries something once that way.

>>1039781
It's not necessary to have Arduino at all to get started with electronics. You just need a solderless breadboard, power supply, and basic components (and preferably, a soldering kit, a stack of universal pcbs, and a good supply of wire): jumpers, LEDs (just get a hundred assorted colors -- order lots, they're cheap and you'll burn them out, RGB leds are compellingly pretty, and segmented or pixel displays are also interesting), resistors (1/4w assortment, more count of more values is better), capacitors (assorted values ceramic and assorted values electrolytic, again more is better), diodes (you'll probably want both signal diodes and rectifier diodes, maybe a zener assortment, maybe schottky and germanium), transistors (3904 and 3906 are basic BJTs you'll want a good supply of, there are lots of other interesting kinds you might want), and switches at least.

It's worthwhile to stock up when you're ordering, because this stuff is so cheap. A dollar for a hundred of this or a few dollars for a hundred of that. Saves you a lot of trouble over getting a stingy little kit that supplies you with like five NPN transistors and five PNP ones.

There are lots of interesting circuits to learn about with just a few basic components, like voltage dividers, long-tailed pairs, circuit mirrors, and multivibrators.

The book "Make: Electronics" has a good guide to the supplies for a beginner.

You definitely also need a multimeter, and you should get an oscilloscope. DSO138 (around $30 assembled) is enough for earliest playing around (the kit isn't a good first soldering project, but it's okay as a third or fourth), but if you have time for bargain-hunting, a used old analog scope is so much better.

A cheap Arduino can be used as a software oscilloscope if you're really pressed for cash, but it's not good.

File: Picture1.png (40KB, 812x489px) Image search: [Google]

40KB, 812x489px

my first comment here. my oven stopped turning on. and i think the problem lies with my transformer, but i am not sure.
my suspicion lies in my picture. a 120v ac is supplied from the left, connected to the transformer. but the secondary winding has no ac output (o volt). what i also checked is the following:
-i know transformers are known to be stable, also i don't see burn marks on it.
-the primary winding has continuity with ~90 ohm resistance between wire 1 and 3. and wire 2 is completely isolated.
-there are 4 secondary winding wires and all of them has continuity with 1~2 ohm resistance.
-there are no continuity between primary and second wire.
so how likely is the transformer in this case to be broken? since it checked out everything other than output, maybe the problem lies in another board?

>>1039723

Well no. You still should have a decent understanding of C++, but there is still some circuit theory in there. The information is all free out there on the internet.

>>1039833

there should be resistance between 1 and 2
from your description the wire in the primary is open
the tap at #3 is for use on a different input voltage

>>1039833

i disagree with the previous poster, i dont think it matters that the third wire shows an open circuit. first thing i'd check is the presence of DC voltage across the diodes, and then the caps to the right of the transformer. if there's no appreciable voltages, then i'd cut the wires from the secondary to isolate the transformer from the circuit. if there's no AC output voltage on the wires, then the transformer is dead. otherwise there's a short-circuited component somewhere on the PCB. if so, feel around to see what's hot.
(make sure your meter is set appropriately to DC or AC volts depending on what you're measuring)

>>1039924
>i disagree with the previous poster, i dont think it matters that the third wire shows an open circuit.
reread his post - the open is between the two connected to the ac input
>work on hvac - furnace xformer uses similar wiring

File: 1471415769633.png (52KB, 812x489px) Image search: [Google]

52KB, 812x489px

>>1039924
>i dont think it matters that the third wire shows an open circuit.
The SECOND wire is the open

>>1039930
so the second wire broke?
>>1039924
why would i check what comes after the transformer when i already checked the transformer itself? how would a bad capacitor impact the connection before it?

File: Picture1.png (22KB, 812x489px) Image search: [Google]

22KB, 812x489px

>>1040082
>>1039930
>>1039928
well another thing, the wire 2 is the hot wire. so maybe it was designed to not have contact with wire 1/3? or else it would of shorted (exploded).
maybe the transformer look more like this? idk.
now i am leaning towards the idea of parts failure after the transformer :/

File: b1141643-900x900[1].jpg (102KB, 900x900px) Image search: [Google]

102KB, 900x900px

>>1040082
>so the second wire broke?
The wire you labeled #2 should connect to wire inside the transformer.
It was bad phrasing because I was responding to bad phrasing.
>>1039924
>i dont think it matters that the THIRD WIRE shows an open circuit

He obviously meant the wire you labeled #3 But that was wrong because you said #2 shows an open circuit.

The transformer has a tapped primary so it can run off more than one voltage supply value.
Without seeing the device or the transformer or a schematic I can't tell you what those voltages are.

The tap for the lower voltage (#2) has burned out inside the transformer.
(the red mark on the tap on the winding I drew on your pic here >>1039930

This is the tap the manufacturer is using to operate your oven.
.
If you connect a voltage to #1 and #3 you should get a voltage on the secondary winding.

It won't be the correct voltage because you're not connecting to the correct tap on the primary.

If you want to test this I suggest you remove the transformer before doing so.

The pic related shows a 208/240 selectable.
Yours could be 120/240 or anything else - I don't know.
(you didn't post a model number or your location in the world)

tldr: Your transformer is fucked - get a new one.

>>1039833
>-the primary winding has continuity with ~90 ohm resistance between wire 1 and 3. and wire 2 is completely isolated.

For the transformer to work you must have continuity between the wire that connect to 120v

>>1040096
>between the wire that connect to 120v
between the wires that connect

>fucked up s key...

hi, just another update >>1039833
i can't find any transformer with my model#. so i try to fix it myself. i peeled off some of the plastic/tape covering the copper wire at the transformer's bottom, and saw an orange thermal fuse connecting to either wire 1&3 or 1&2. i suspect that the fuse broke.
also the transformer is connected with the pcb on 12 points, and i never soldered anything in my life. so i am scared i will mess something up. i will ask someone else to try prying the transformer out tomorrow.
>>1040096
ok thanks, i like this confirmation. i suppose the reason for the transformer not to blow up >>1040090 here
is because of the fuse acting as a switch?
i drew another picture with the fuse, it is supposed to be at the bottom, and the copper winding at top.

File: Picture1.png (60KB, 896x424px) Image search: [Google]

60KB, 896x424px

>>1040190
oops forgot to attach picture

>>1040200
>i suspect that the fuse broke

it's odd that you suspect the thermal fuse of being dead, coz the first thing a sane person would have done is to check it for continuity, instead of posting their suspicions on 4chin.

so after you check that, you can choose to live dangerously and just jumper across it or find a replacement. the temp should be marked. i've taken apart a few laser printers, so i have a bunch at diff temps, usu. around 100-130-deg C.

this wont necessarily fix the problem, tho, coz the reason the transformer overheated could be that some component on the PCB is shorted, or otherwise defective. i dont think you appreciate that problems can flow both ways: a downstream component can die, and that can be reflected back into the power supply, causing it to die as well.

>>1040277
> a downstream component can die, and that can be reflected back into the power supply, causing it to die as well
fuck! i am very afraid of this scenario. there are 3 main pcb in my oven and any one of those board cost 250~300ish to replace. i already spent a lot of time examining other parts and only to leave with inconclusive result due to how complicated those other boards look. i sure feel like slapping myself if the fuse didn't solve the problem.
of course i checked for continuity, i didn't even know that a fuse existed inside the transformer until after i find out that primary windings needs to have a connection >>1040096.
how else would they not have a connection if it wasn't the fuse?

>>1039254
>ebay SMPS
/diy/ you faggot, not pre-built

>>1040295

Literally just copy that design, then; it's just the reference schematic for the LM2596. 9 components in total. 11 if you want the oh-so-fancy LED on there. Beef up the filtering to clean the output a little.

Now, while I personally agree with the preference for a large-ish transformer and linear regulator, to say you can't get a cheap, capable SMPS is foolish. Getting one well-designed enough to be ultra-efficient and have a nearly ripple/noise-free output is another matter, but, for just about everything a hobbyist will do with it, those really aren't concerns in the first place.

>>1040321

As an aside, however, I do acknowledge that this is only a DC/DC converter, and you'll still have to find some way to actually get it (safely) plugged into a wall. But, by that same token, the transformer you'd need for a linear supply isn't free, either.

>>1040322
A bench supply is probably going to have two parts: a PSU which converts mains to a (roughly) fixed DC voltage, and a regulator which converts the fixed DC voltage to a variable DC voltage.

The first part can be a classic transformer + rectifier + capacitor design or a SMPS. The second can be a linear regulator or a switching regulator.

You could do it all in one go with a variable-voltage switching regulator, but that's a) complex (a variable-voltage isolated topology is somewhat more complex than either a fixed-voltage isolated topology or a variable-voltage non-isolated topology) and b) potentially risky (particularly if it's 240V rather than 120V).

For the first stage, there's really no reason not to use a SMPS; they're lighter and cheaper, and regulation doesn't matter if you're following it with a separate regulator.

For the second stage, the inherent inefficiency of a linear regulator will be a major issue if you want a wide voltage range and reasonable current. You'll have to reduce the maximum current at lower voltages to keep the power dissipation within limits, whereas a switching regulator can supply more current at lower output voltage as the input current scales with the duty cycle and you're not dissipating (Vout-Vin)*Iout as heat.

Also, the original requirement (>>1038810) was "controlled by an arduino". For a linear regulator, you're basically talking about a pure analog design with uC-controlled DACs instead of pots. Whereas a switching regulator is a reasonable uC project (I've stopped using the 34063 etc in favour of PICs, the latter being infinitely more flexible and behaving much better in continuous-conduction mode).

I posted this on /g/, but I figured I might have a little more luck on this thread specifically.

Hoping to get some help. I've been working on putting a Raspberry pi 3 inside an NES and been using the attached as a guide.

The 5v blue wire and the white ground wire - what are those and where can I get them? The guide I'm using doesn't mention this nor include this in a parts list.

Any info would be appreciated!

File: 1470056245495.jpg (340KB, 1583x2048px) Image search: [Google]

340KB, 1583x2048px

>>1040351
Damn, forgot to post the pic.

>>1040351
>what are those and where can I get them? The guide I'm using doesn't mention this nor include this in a parts list.

Hookup wire - if you don't know what that is or where to get it you're probably out of your league on this one.

Hello /ohm/, I still can't get solder to flow through vias. I've been experimenting with flux and temperatures and nothing seems to work. I'm afraid of thermally damaging the components if i hold it in too long. I used flux and held it there for 10 seconds and all I got was a hot board. What do? I'm using 1mm solder, getting 0.5 in a few days, will it matter?

>>1040407
Somehow I get the impression that you're trying to do something really silly...

Well, if you're talking about normal plated through vias: no need to solder them.
If you're talking about plated through holes for components: this depends on PCB design and component type too. Sometimes it simply isn't practical with soldering iron alone and it's better just accept that solder isn't getting pulled to the other side of the board (instead of continuing to grill the pcb and components).

File: USBwarmercooler1.jpg (27KB, 580x343px) Image search: [Google]

27KB, 580x343px

Hi /ohm/ I have a 40W peltier plate I pulled out of a portable fridge and I want to make a glass/mug cooler/heater, I already have a heatsink, fan...
I was thinking of adding a heating element on top of it, but I've seen some videos of people reversing polarity of the peltier plate to use it as a heater without flipping the plate.
Can this damage the plate? Also, is there any way to maximize the contact with the glass/mug, thus heating it faster and more efficiently? I thought about using steel beads on top of the mandatory aluminum plate, but that will probable be too unstable. Ideas?

>>1040367

Kind of agree with this guy. Check sparkfun or most other consumer hobby stores will stock it.

File: file.png (6KB, 370x475px) Image search: [Google]

6KB, 370x475px

Can I use the attached to switch power to my load? I want it on when the open collector is on.

The gate will be at ground when the collector is on so Vgs is negative. I'm not sure if that's allowed with a PMOS, or if it's supposed to be between 0 and Vgs(th). If not I'll have a higher supply and use a voltage divider to get the gate to the voltage of the load.

>>1040619
By the way, I can't put the load after the FET. It would make things easier but it's off the table.

>>1040627
I mean, can't put the FET after the load, sorry.

>>1040619
The FET is upside-down (that's the default orientation for LTSpice' pMOS). The source should be connected to the positive rail so that Vgs=0 when the collector is pulled high and Vgs<0 (i.e. gate negative wrt source) when the collector is driven low. The drain should be connected to the load.

> I want it on when the open collector is on.
By "on", I assume that you mean when the transistor is conducting (collector driven to ground)? Note that's considered a logic zero/false/low output (the relationship between voltage and low-high/false-true/0-1 is the same for open-collector with a pull-up as for a push-pull output).

If you want the load powered when the output is logic 1/true/high, you need an extra transistor (or move the load to the high side and use nMOS).

File: file.png (6KB, 342x426px) Image search: [Google]

6KB, 342x426px

>>1040633
>The FET is upside-down
Oops, fixed.

This'll work, then? I didn't realize current the drain->source flow of an n-channel is reversed in a pMOS, but that makes sense with Vgs<0.

You assumed correctly about the "on" vs conducting. Thanks for the info.

File: proposal.jpg (43KB, 1056x500px) Image search: [Google]

43KB, 1056x500px

Hey guys, I'm having trouble figuring out how an fx bypass would work. Take a look at pic related: will this work, and if now what do I need to change?

>>1040641
> This'll work, then?
Yes.

That's fairly typical for driving a low-side (one-side grounded) load where the supply voltage is above the logic voltage (if the supply voltage is the same as the logic voltage, you can just drive pMOS directly from a push-pull logic output).

> I didn't realize current the drain->source flow of an n-channel is reversed in a pMOS
Current can flow in either direction; the channel itself isn't polarised. However, many power MOSFETs have a "body diode", which will allow current to flow in the reverse direction even if the FET is off.

>>1040661
> Sorry about the crayon but we're not allowed to have sharp objects

File: 1471563457096.jpg (44KB, 1056x500px) Image search: [Google]

44KB, 1056x500px

>>1040661

It's usually done with the switch inside the fx box.

DPDT switch

File: gc_diagram1.jpg (238KB, 947x799px) Image search: [Google]

238KB, 947x799px

So I asked /g/ but it was fruitless as I'm sure is the case with many others BUT here I am.

I am trying to calibrate a gamecube controller's joystick so that I can manually set the resting and max/min values of the sticks x/y and assume I need to program the EEPROM chip somehow but honestly have no idea where to start. I've tried google but it only turns up with emulator settings which are not permanent and are a virtualization of the inputs anyways. I need to be able to map analog inputs and have it stay that way even after being unplugged. Any advice would be greatly appreciated

>>1040677
so which is positive and negative?

>>1040681
This is the coax signal cable.
Black is the outside/braid/shield
Orange is the inner/signal

plus and minus are for the power
the box has a separate switch for power.
power is normally left on while using the box

You don't switch the power to use or not use the effect.
You switch the signal through or around the box.

>>1040683
Why does the shield look like two wires that meet before the fx circuitry, and why is the shield only going through the fx circuitry (as opposed to an additional shield going directly to the output)?

I should mention, I'm using 1/4" tip-sleeve ports and this is'nt a pedal, it's a circuit bent toy.

>>1040690
You posted this >>1040661
>proposal.jpg

And you're questioning my drawing?

I'm through

>>1040351
A really stupid thing with the Pi3, it needs closer to 5.1V to work properly. Most USBs don't output a high enough voltage, so although they might output enough current at 5V, the Pi3 complains with the rainbow square under load. They make supplies specifically for the Pi3.

>>1040352
As for your image, it's very likely you'll damage your Pi by trying to power it from the GPIO. There's no smoothing, no regulation, no nothing on the 5V GPIO pin. Same with the test points. Just use a damn usb cable

>>1040601
There is some max junction temp for the Peltier that you don't want to exceed. Depending on how hot you want to run, it may be an issue.

>>1040679
How do you know there's even an EEPROM in there? Mass production stuff is going to be built as cheap as possible.

>>1040794
I guess that's true since the neutral position of the stick is determined when the controller is powered and can be changed by holding the stick in a different position... I didnt think of that. But there are different 3rd party retailers that modify 1st party controllers in such a way that allows for certain inputs to be performed easier and i have no idea how

>>1040797
They might be replacing the microcontroller with their own custom-programmed one.

File: arm.jpg (47KB, 620x278px) Image search: [Google]

47KB, 620x278px

>>1039029
Cortex M0s like NXP's LPC line is probably in the hundred times faster than AVR range and is quite simple (in terms of hardware, not necessarily babby mode software support.) Remember performance isn't measured in megahertz, so even if it's not that much higher clock it will run faster because it can do more each cycle. The reason you don't see high clocks on microcontrollers is those fast chips usually need quality power supplies, external crystals and memory all with impedance and length matched traces.

>>1040947
The difference in clock speed between "microcontroller" ARMs and the ones you get in phones/tablets/etc is partly about cost and partly about the application domain (i.e .you're not running dozens of processes, half of them written in Java).

External crystals aren't an issue; the cost is minuscule, and you usually need an external crystal for USB. Nor is RAM (which is built-into the chip; if you do use external RAM, it's for sequential-access buffers, not as "system" RAM).

>>1040807
how hard would this be? At this point I think it would be more efficient to just make my own controller using parts from other controllers with my own microcontroller. I'll also try to get ahold of one of the 3rd party "perfect" controllers that my friend purchased not too long ago and see what's going on inside of it.

File: myretardedsupply.jpg (39KB, 634x428px) Image search: [Google]

39KB, 634x428px

I'm powering a little robot using this setup. I need 5v for the microcontroller and an unregulated supply from the batteries for the drive motors. I am using two 9v batteries. This setup works but the regulator massively overheats even with a pretty sizeable heatsink. What is the more correct way to do this?

>>1041046
how hot is hot? did you measure it? it can be uncomfortable to touch without necessarily being too hot to work. just.

linear regulator dumps extra voltage into heat. calculate the power dumped as heat by doing 18v - 5v = 13v * current = power.
if you want to not have a hot regulator you need to reduce the source voltage or draw current which is usually not convenient or you can use a switching regulator. e.g. http://uk.rs-online.com/web/p/switching-regulators/6664379/

there is some stuff about shit capacitors and your regulator ringing and burning up too but its probably just inefficient.

>>1041046
Are the batteries in series or parallel? If they're in series you can reduce the heat significantly by having the supply to the regulator be only 9V.

Otherwise, a voltage divider could help.

>>1041051
They're in series because I need more than 9v for the motors.

>>1041049
If it works I don't care about heat but it has blown the regulator out with a smaller heatsink before after running for a while.

I have some buck converters, would it be better to use that instead of a linear regulator?

File: tegaki.png (3KB, 400x400px) Image search: [Google]

3KB, 400x400px

>>1041057
You'll have to get a ganged switch but this is an easy fix. (9-5)V^2 is much much less than (18-5)V^2

>>1041058
Oh yes that's it, I just realized I need to take only 9v from between the two batteries to the regulator instead of sending all 18v.

For a potentiometer like this, what does the 25R "resistance" refer to? Maximum? Center?

Is 0 Ohms the nominal minimum bound?

>>1041122
Never mind, read the datasheet. Didn't link it anyway.

Is there any reason to use a voltage divider rather than a straight rheostat in series with LEDs to control their brightness?

How do I represent a sum of product A'BC+AD+AC in product of sum form represented by product m notation?

>>1041057
> I have some buck converters, would it be better to use that instead of a linear regulator?
Yes. Linear regulators are inherently inefficient (e.g. 9V to 5V is at most 55.55% efficient), switching regulators can manage 99% if you're willing to "gold plate" the specs and 90% is feasible even for a price-sensitive design.

>>1041062
> I just realized I need to take only 9v from between the two batteries to the regulator instead of sending all 18v.
That will certainly improve matters (it will reduce the power dissipation by a factor of around 3). The downside is that one battery will drain faster than the other, which may be an issue for charging.

>>1041222
(A'+C+D)(A+B)(A+C)(C+D)

Can anyone recommend some nixie tubs which are easily available from sources like aliexpress/ebay?

I decided to make a nixie clock and I want a tube for it. After I got the tubes I can see from the datasheet what I actually need to build.

>>1041238
Well the robot only has to run for less than 10 minutes at a time so it should do.

File: 465661-koss_m80plus_speakers.jpg (54KB, 600x450px) Image search: [Google]

54KB, 600x450px

Bought some old speakers at a garage sale, thinking about making an amp and using them with my computer. Does this schematic look good, or would I be better off being a buyfag?
http://electronics-diy.com/100w-lm3886-power-amplifier.php

They're 4 ohm, so I can't just use an old stereo receiver.

I should test them and see if they sound any good before thinking about this too.

File: -2106169131712756037.jpg (120KB, 800x800px) Image search: [Google]

120KB, 800x800px

Does anyone have experience with these cheap soldering iron controller kits? Apparently, you hook these things up to a power supply like a laptop charger. And they make decent hakko T-12 series compatible stations.

http://s.aliexpress.com/qAFJvEfI

>>1041284
>They're 4 ohm, so I can't just use an old stereo receiver.
Sure you can. Many old, and new, receivers can handle 4 ohm. That standard ranges were 4-16 ohms. 8 is just the most common.

>>1041046
What are running from the 5V line that draws that much current?

File: 123_AVX.jpg (16KB, 248x346px) Image search: [Google]

16KB, 248x346px

What is the story behind stacking ceramic SMD caps? Everyone always use to say "no don't do this!" now I'm seeing them being stacked in highend aerosapce applications and various prototypes and companies are selling stacks.

>>1041316
The stacks have flexible leads and IIRC the capacitors themselves are attached to the end leads in a flexible manner.

>>1041322
Did they use to crack if you tried to stack conventional ones without the flexible ending?

>>1040679
May I ask what the end goal is?

>>1041324
Afaik yes, at least that's the reason behind pretty much all the mounting and handling restrictions of ceramic capacitors.

>>1041341
>Afaik yes, at least that's the reason behind pretty much all the mounting and handling restrictions of ceramic capacitors.
Couldn't you make a portable railgun with these? To store 2500 joules of energy using them I worked out the volume needed to be something like 30cm x 9cm x 9cm. And they discharge much faster than the aluminum caps home made railguns always use.

>>1041325
Have a way to program stick positions to make certain techniques that require precise inputs easier in super smash bros melee.

>>1041303
Mine only goes to 6.

>>1041401
You cheater

File: images (15).jpg (19KB, 448x299px) Image search: [Google]

19KB, 448x299px

Hey /ohm/. I have been thinking about a soldering and desoldering station and unfortunately nobody imports anything other than some chinsese stuff so pic related(zd 917) is my only option. Shpuld I do it am I about to make a mistake that I'm going to regret?

>>1041442
most of the top players have custom controllers like this. they're perfectly legal right now among other controller modifications that are embraced by a large portion of the community

>>1041444
What's the point of being a "top player", when the controller does your work?
Or conversely, what's the point of playing games which are so shitty that even the best players need special controllers to perform the moves for them?

>>1041455
the thing is, controllers like I am trying to make exist from the 1st party you just have to spend obscene amounts of money on what is basically a blind box situation. They buy these controllers (at almost double the cost of a 1st party) for the guarantee that they will work, bypassing the extremely expensive lottery that is acquiring a decently performing controller. Also there are many more levels than the physical input that are required to be considered a top player but i wont get into them since thats an entirely different topic. So, any ideas on how to do this?? or any idea how difficult it would be to replace a microcontroller in one? or how hard it would be to make my own controller?

File: 14857.jpg (35KB, 1000x666px) Image search: [Google]

35KB, 1000x666px

so, recently i've tried building an oscilloscope out of a portable TV set, a roadstar tv 400n specifically (pic related), i've wired everything in place like it should be :
>cut off the speaker
>cut off horizontal coil, connected it to the vertical coil wire

>connected horizontal coil to audio cable (runs into the microphone jack on a mixer and outputs onto a radio)

problem is, i've got a steady line horizontal line displayed on the screen but it doesn't react to the audio input, i've tried messing with some pots and rewiring it so the vertical coil is displayed but nothing seems to make it work, could it be that the TV signal recieving circuits are overriding the input or something, i identified a circuit labelled as TUMUB3EC-991 on the board and i guess it's the TV tuner, should i desolder it?

>>1041483
>>1041485

are you cheating?

File: 6.jpg (519KB, 1600x1200px) Image search: [Google]

519KB, 1600x1200px

>>1041483

> runs into the microphone jack on a mixer and outputs onto a radio

I've done that trick myself a few times. You need to poke around the circuit board looking for the input to the speaker amplifier. The "microphone jack" is a headphone jack. You can't just run audio into it and expect it to be amplified to the TV speaker you wired to the deflection coil.

Use a multimeter to identify any high voltage points on the PCB you don't want to hook up to your audio source. Connect your audio source to random safe points on the board until you get results.

Am I safe using a MOSFET rated 200 mA continuous when there's a worst case scenario of 550mA 1/7 duty cycle?

>>1041517
The datasheet says "Maximum Drain Current - Pulsed = 500mA" but doesn't explain the pulsing. Should I assume that's a 50% duty cycle?

http://www.st.com/content/ccc/resource/technical/document/datasheet/03/f9/c4/3d/7f/eb/4c/5e/CD00001232.pdf/files/CD00001232.pdf/jcr:content/translations/en.CD00001232.pdf

On page 11 there's a diagram of a typical step-down converter, from 25V to 5V. I want to go from 12V to 5V. Will I need to change anything? What does the timing capacitor do?

This is my first time using a buck converter. Please excuse my ignorance.

>>1041527
how about you tell us what fucking mosfet you are talking about

>>1041574
https://www.fairchildsemi.com/datasheets/2N/2N7000.pdf

>>1041566
Still curious about this. Also, is the inductor going to be an issue if IOut is changing? I've got a few ICs and a microphone running off 5V and I don't know much the current will fluctuate, particularly with the mic.

>>1041566
>>1041690
>What does the the timing capacitor do?
It determines the ton - toff durations. You don't need to mess with it. Refer to figure 4 on page 8 and Ct calculation on page 13.

>Will I need to change anything?
Check out the diagram on page 3. You can see there is a comparator that controls one of the inputs of an and gate. This means if the comparator gives zero output, the clock signal won't propagate. So no charge transfer will occur, meaning the output voltage will go down. So what you need to do is, make sure that the output voltage you want gets divided by a resistor divider and fed as 1.25V input to CII pin. In this case, VOUT*(R1/(R1+R2)) will be equal to 1.25 and comparator will keep the output voltage close to VOUT you desire by allowing or preventing clock signal to propagate.

I don't understand your other questions.

>>1041517
> Am I safe using a MOSFET rated 200 mA continuous when there's a worst case scenario of 550mA 1/7 duty cycle?
Not quite. 550mA with a 1/7 duty cycle is 208mA RMS.

But I^2*R isn't the only issue. There will also be a maximum peak current:

>>1041527
> The datasheet says "Maximum Drain Current - Pulsed = 500mA" but doesn't explain the pulsing. Should I assume that's a 50% duty cycle?
No, you should assume it's the maximum instantaneous current, i.e. that it applies even to a 1% duty cycle.

So 550mA @ 1/7 duty cycle is exceeding both limits, although neither by very much (assuming that the 1/7 duty cycle is at a frequency significantly higher than the thermal time constant).

>>1041698
Thanks for the response. Figure 15 gives a Vout of 5V which is what I want anyway, so I'm going to use the given values for R1 and R2. C2 is decoupling so I'll use that value, too.

Rsc and L are really what I'm worried about. I don't know what Rsc does from looking at the diagrams. Should it be different in a 25V to 5V versus a 12V to 5V converter?

The inductor seems like it'll prevent changes in current, and I'd rather allow them, as the electronics running off 5V should experience current fluctuations. Can I get rid of the inductor without messing anything else up?

>>1041701
Ah, thanks. I was looking at a IRF630 and 2N7000BU, 9A continuous current vs 200mA, at the same price, so I figured there must have been some drawback. But I suppose I'll just take the beefier one. I don't see any specs on it that are problematic.

>>1041566
> On page 11 there's a diagram of a typical step-down converter, from 25V to 5V. I want to go from 12V to 5V. Will I need to change anything?
Maybe; changing the input voltage from 25V to 12V will change the inductor voltage from 19V to 6V, which is a bit over 3:1. dI/dt will change by the same factor, so you would need to reduce the inductor value or reduce the oscillator frequency to obtain "similar" behaviour (personally I'd keep them the same; the current ripple is already quite high with those values).

The main thing to bear in mind with fig. 15 is that you're dropping up to 1.3V in the transistor, which limits you to 1A (and the case is going to get pretty hot at that). If you want more than that (or cooler operation), use an external PNP or pMOS transistor (fig.21).

> What does the timing capacitor do?
It controls the oscillator frequency. See fig.4 on p8.

>>1041690
> Also, is the inductor going to be an issue if IOut is changing?
No. The current draw dictates how quickly Vout will drop between cycles. C1 is sized based upon peak current, oscillator frequency and maximum ripple. A 470pF timing capacitor corresponds to a cycle time of ~25us. That's 25uC/A; with a 470uF filter capacitor, that's 53mV of ripple per ampere of output current (ignoring ESR).

>>1041710
> I don't know what Rsc does from looking at the diagrams. Should it be different in a 25V to 5V versus a 12V to 5V converter?
Rsc is the current-sense resistor. If the voltage between pins 6 and 7 exceeds Vipk (~300mV), switching is suspended. A value of 0.33 ohms limits the input current to ~1A.

> The inductor seems like it'll prevent changes in current, and I'd rather allow them, as the electronics running off 5V should experience current fluctuations. Can I get rid of the inductor without messing anything else up?
No. The inductor is fundamental to the operation of a switching regulator. It's what limits the current (the transistor is operated as a switch, i.e. it's either off or on, not in the linear region; that's why it's called a "switching" regulator).

The load isn't drawing current through the inductor, it's drawing it from the filter capacitor. The inductor current is varied by the IC to keep the filter capacitor at a constant voltage.

>>1041710
>The inductor seems like it'll prevent changes in current, and I'd rather allow them, as the electronics running off 5V should experience current fluctuations. Can I get rid of the inductor without messing anything else up?
No, inductor is an integral part of figure 15 as well as any inductive based converter. If you really don't want an inductor you need to use a switched capacitor converter. You can use this IC to build one but there are much better alternatives.

Rsc goes in series with the collector of the bipolars, so it's there to sense the voltage drop through IPK pin. It just provides a voltage drop, and if the voltage drop is higher than a certain value it cuts off the supply or do something. You can tweak its value to change the maximum current allowed.

File: image.jpg (1MB, 2163x3133px) Image search: [Google]

1MB, 2163x3133px

Can I use paper clips as wire? The new capacitors I bought were too short to span the length

>>1041713
>>1041715
>>1041716
Ok, if I'm understanding correctly I can leave it as is. Thank you.

>>1041729
>paper clips as wire?
paper clips ARE wire - steel wire
copper wire would be better but that will do in a pinch

>>1041732
just checking, these are basic Quill paperclips, didnt know if they were coated in something making them unconductive. trust me I'd rather have some light gauge insulated wire, but i dont have any right now and its just a passive crossover circuit so nothing too dangerous

>>1041733
>a passive crossover circuit so nothing too dangerous
Those green caps look like electrolytics
Crossover needs non-polarized like the two paper caps you disconnected

>>1041729
make sure you clean the paper clip thoroughly. it may have a plastic covering that is inconspicuously thin

>>1041742
They are green because they are NP, no worries friend

File: 25V-1000UF-electrolytic-capacitor.jpg (73KB, 640x480px) Image search: [Google]

73KB, 640x480px

>>1041749
>They are green because they are NP, no worries friend
They are green because they are green colored.
These are green for the same reason. Notice the - stripe denoting the negative terminal.
>not all green colored caps are NP

>>1041752
I bought them specifically as NP man, do you need to see the datasheets? relax I know how to purchase capacitors

http://www.mouser.com/ds/2/293/e-ues-876267.pdf

>>1038501
don't buy from dx.com, they are expensive.
also, related: >>>/g/csg

>>1038379 (OP)
>compiling and putting code on an atmel chip
I trying to program an atmel chip, a meta32u4, via usb. I'm on linux and using avrdude and avrgcc.

I got avr-gcc to compile without errors using:
>avr-gcc -mmcu=atmega32u4 test.c -o test
I'm sure thats wrong for whatever reason, but I haven't got to the point where the output file works or not.

avrdude errors out with
>avrdude: Error: Could not find USBtiny device
, I'm using the command:
>sudo avrdude -P usb:0000:0000 -c usbtiny -p m32u4 -U /dir/test

with the 0000s being the usb id. I've tried with bus:device, and a few other variations but none of them have worked so far. /dev/*device* hasn't worked either.

>>1041819
learned that usbtiny isnt a protocol or some shit, its an actual device.
This thing can be programed over usb, what options do I need to set to do that?

>>1041827
>This thing can be programed over usb
Yeah, if it already has a bootloader programmed. If not, you have to start with SPI or JTAG programming.

>>1041827
It's called the DFU. No experience with it.

>>1041840
Yea, its got a bootloader
>>1041841
theres a couple things called flip1 and flip2. flip1 seems to want to do something but its back to not seeing the device

File: 1457327151955.gif (1MB, 480x270px) Image search: [Google]

1MB, 480x270px

>>1041501
so basically i need to desolder the audio from the coil and start soldering it onto random PCB connections? i also remember reading somewhere that i should desolder all and any pots and chips from the board, should i do this? thanks for being helpful, m8

>>1041861
> t i should desolder all and any pots and chips from the board, should i do this?

No. Those chips are doing useful things you want them to keep doing: Running the CRT and amplifying the audio. You shouldn't need to desolder anything. Just find the point on the PCB that's the input to the audio amp you rewired to the CRT deflection.

>>1041729

Depending how many amps you have running through them it should be fine. Steel has a higher resistivity than copper, but at low amps it shouldn't matter.

>>1041760

Your data sheet says specifically they are bi polarized. Also, like >>1041752 said look at the stripe AND the shorter lead denoting a negative terminal. Your data sheet denotes the same thing.

Now looking at Mouser it says they have both NP and BP caps in that product line. IIf you're sure you have NP caps then fine, but everything you said about color and shit is wrong. If they are BP they will have a shorter negative lead and have a stripe down the side.

Is there any way to tell the original capacitance of a failing capacitor? Said capacitor is unmarked, and I can't find shit about it online.

>>1042068

Is it big? If it's just a decoupling capacitor you can just assume it's .1uf or 1uf

>>1042072
It's tiny. It's used to control voltage for a watch face. I just realized that there's four of them, and the first is most likely fine, so I'll just find the capacitance for the first one and replace the other three with duplicates of the first.

I'm about to finish an Associates in Electrical Engineering Tech soon. I want to go on for my Bachelors in EE eventually but until then I want some resources to get a head start on what I will end up learning. I want some suggestions for textbooks or other resources that cover more advanced electronics topics and are geared more towards an engineer rather than a hobbyist.
Looking for stuff that covers things like analog and digital signal processing and LTI systems and whatnot. Anything else is welcome as well.
I've had up through multivariable calc and ordinary differential equations and while I'm rusty I've had enough of a math background where I should be able to dive into this material and be able to follow along. Bonus points if a book has problem sets to solve though since I need the practice.

>>1041819
I know someone here programs avrs through usb

>>1042277
Differential Equations are going to be the most useful, might want to go ahead and get into partial differential equations also. Other than that, physics, lots of physics. Luckily the DE makes physics a lot easier than it would otherwise be. You should know Maxwell's Equations inside out obviously.

Other than that, I don't know, I'm a geophysicist not an EE lol.

File: Screenshot_9.png (43KB, 688x338px) Image search: [Google]

43KB, 688x338px

>>1041885
after doing some further poking around i identified the IC, it's an AN5151N, instead of poking around the PCB could i directly solder a wire from the IC pin handling audio amplification to my audio cable? if so, what pin would it be? i've been lead to believe that the leftmost lower row one should do it, but i'd rather not fry my ass.

File: image.jpg (2MB, 4032x3024px) Image search: [Google]

2MB, 4032x3024px

Hey /ohm/, why is this iron so comically large?

>>1042277
>head start on what I will end up learning
What you mentioned is not what you're going to learn, literally. We do two signal specific courses, one is the basic fourier transform course, the other is a communications course, using Matlab for signal decoding I think ( I haven't taken it yet).

Most of what we've taken is electronic or physics related. Applied Electromagnetics (waves in mediums/air), digital logic, a probability course (math oriented), transistor amplifiers (two semesters) and labs for basic breadboarding/circuit layout. We have literally done no LRC specific intense-math oriented classes. We did do a power distribution class, but there's only one needed for the Bachelors. We used Sedra Smith's Microelectronic Circuits for the two semesters which involved breadboard labs and amplifiers. We got up to the design of the 741.

I will be graduating in Spring with my BS in EE. I am taking EE electives, some photonic related courses. I could've chosen more nanoscale science. Literally unless you're taking a math minor, Calc III and DifEq are the highest math you need I think. I took Linear Algebra before I transferred, and it's not required unless you're doing an EE Math minor

>>1042571
80W, all the heat is at the tip

>>1042571

because it's not what you want

That's probably meant for wood workers and people who need to solder big things

for electronics get something made for electronics like a fx-888d

>>1042587
To date I've already done

DC Circuits (passive component theory, Ohm's law, Thevenin and Norton eq. circuits, nodal analysis, etc.)
AC Circuits (passive component theory for AC signals, circuit analysis techniques for AC, RLC filters, dB and bode plots, transformers, single phase and polyphase systems, AC power, etc.)
Electronics I (diode theory and operation, various types of diodes and their applications, BJT and JFET theory and operation)
Electronics II (BJT, JFET, and MOSFET biasing and amplifier design, power amplifier design, op-amp theory and applications, SCR/triacs, etc.)
Digital Electronics (intro to binary and hex and performing math in base-2 and 16, logic gates, adders, k-maps, encoders, decoders, multiplexing, flip-flops, etc.)
Advanced Digital (counters, shift registers, memory, FPGA programming -> Xilinx Sparten3e, probably other shit I'm forgetting, etc.)
Control Systems (control system theory, closed loop vs open loop, ladder logic, relay logic, PLCs, etc.)
Microcontrollers (useless course, worked with PIC16F877 programming in BASIC. Learned absolutely nothing)
Applied Circuit Analysis (condensed calc 1-3 and differential equations for engineers. Another bad course that left out too much information and wasn't very "applied." Fortunately I've taken dedicated calc and diff eq. courses so this was only taken for the credit.
Electronic Instrumentation (Pretty much a primer on the various types of instrumentation you will use and how set up and read shit correctly, mostly focuses on oscilloscopes but went through volt/ammeters, function generators, frequency counters, and other things of the like)

All courses above involved labs and a great deal of simulation and breadboarding. It seems like the only thing I haven't had is any course about Electromagnetics though we probably didn't go into the level of detail for some of my other courses you went into in yours. Most of my courses were heavily algebra based and used little calc.

File: 1471128996267.jpg (256KB, 1280x960px) Image search: [Google]

256KB, 1280x960px

I know I can just google it but I am just tempted to ask here since you guys tend to be more knowledgeable in general.

Anybody here can recommend a book / source to learn about analog - digital signals etc?

File: b_12.jpg (38KB, 448x304px) Image search: [Google]

38KB, 448x304px

>>1042571
Because you forgot to resize the picture.
>4032x3024 for a damn pic like that.

No seriously if you are trying to solder large items they can be useful ie soldering coax to brass bolts or similar.


Also pic related.

>>1042769
What do you want to learn? analog-digital signals is a really vague description.

If you mean analog-digital mixed signal design, you need to start from basics, I can list 10 books at least.

If you mean analog signals, you need to include what you want, communication systems are significantly different than analog signal processing.

See even while trying to be general I got side tracked. Is your question analog - digital signals, or analog - digital, signals. It makes significant difference as mixed signal domain is weird and a profession by itself.

>>1042780

For starters, I need to do a presentation for a class at college which involves Data acquisition systems, so professor wants me to talk about Analog - digital signals and, well, just say knowledgeable shit and dont look dumb in front of everybody.

On the other hand, I have taken an interest in electronics, so I am interested in the basics too for example.

desu I dont know where to start. my knowledge is pretty basic in this stuff.

>>1042783
Depends on the level entirely.

For entry level data acquisition and analog to digital or digital to analog conversion sampling theory is a must. You also need to learn Fourier transform as that's how each of them characterized. And you need to learn about quantization error and noise, you can even talk about how 32 bit audio is a scam.

Unfortunate thing is this is all the direction I can give, it's been a long time since I started working in electronics so I don't really remember how I started or which books I've read when I was beginning.

>>1042783
I actually have a really good book on signal analysis that I can link you when I get home. It's a little on the heavy side but it'll teach you just about all you need to know.

Then if you're super interested, I have a digital copy of my random signals college book I could give you too.

So I just got a claymore clacker, and after doing a bit of reading it looks like it could be wired into an Arduino. Now I just need something awesome to make that deserved to be triggered like an explosive.

File: mms-f10.gif (4KB, 600x250px) Image search: [Google]

4KB, 600x250px

Why are the ranges on multimeters always 0.2V, 2V, 20V, 200V and... 750V, sometimes 600V or 1000V, and never 2000V?

>>1042786
>he can't hear the nanovolts

File: 80k-40.jpg (58KB, 613x313px) Image search: [Google]

58KB, 613x313px

>>1043006
>always 0.2V, 2V, 20V, 200V
Because ICL7106 and its later variants have a range of +/-200mV. It is a very popular IC in shit multimeters.
>750V, sometimes 600V or 1000V, and never 2000V?
Mainly safety and safety classification reasons. Good high voltage dividers tend to be expensive, too.
Multimeters with 2kV and higher ratings exist, but they tend to be quite limited otherwise, or are rather old.
You can also use high voltage probes like pic related with normal multimeters.

File: uxbot_.png (78KB, 400x300px) Image search: [Google]

78KB, 400x300px

Hey /ohm/ my professor wants us to race line following robots. Only the fastest car can get an A in his class. Normally, I'd do this entirely myself but there's a lot riding on being the best.

Anyway, most line follower racers I've seen have two wheels with sweep in front like pic related. My professor is strongly urging us to purchase an rc car and rig it to be a line followed racer (ie 4 wheels). The track is carpet with white tape.

Is there any advantage to 4 wheels or disadvantage to the common 2 wheel design? If 4 wheel designs are better, why is every single design I see at world championships 2 wheels?

I'm trying to design a circuit that will dim an LED strip in time with music by filtering for bass and using a peak detector with a .1s time constant to allow current to pass through another resistor as well as the LEDs. R6 and R4 will be potentiometers to allow me to adjust the brightness of the lights when they're dimmer as well the threshold for them to dim.

Does anyone have any better ideas? I was also thinking of replacing the right half of the circuit with an op-amp that adds an inverted signal from the low-pass to the LED supply voltage. That would let the volume of the bass note affect the brightness of the LEDs, which would be cool. The downside to that is that I'd have to get a negative supply voltage for the op-amp, which as I recall is possible for fairly cheap with a buck converter. Anyone got any comments?

File: file.png (18KB, 1259x710px) Image search: [Google]

18KB, 1259x710px

>>1043024
Forgot the picture like a total jabroni

>>1043017

Two wheels are way more maneuverable than four. Go with four. If you must, add in some stiff spring wires angled down that trail on the front to lift it off the ground. That should help prevent your bot from getting stuck in the carpet.

Don't reinvent the wheel; copy what wins world championships if you can.

>>1043016
>Mainly safety and safety classification reasons.
So if I measured a voltage up to 2kV with a meter rated for maximum 750V, would I actually get valid readings?

>>1042903

Hello anon, I am that anon who asked for the meme book, I just got back and saw your post

If you have the name of that one book I can look it up and see if I can find it

File: file.png (6KB, 271x278px) Image search: [Google]

6KB, 271x278px

I've got a LPF feeding into an inverting op-amp. Is the resistor in the LPF added to the input resistor when calculating gain, or does the capacitor make it not count for gain?

>>1043033
>go with four
>copy what wins world championships

The championships are won by two wheelers, but I should go with four? I'm a little confused on what you're saying.

>>1043156
> Is the resistor in the LPF added to the input resistor when calculating gain,

nope. you keep calculations of the 2 elements separate: the filter provides attenuation, the amplifier provides gain. multiply the two to calculate total gain.

>>1043025
this circuit doesnt work. the LEDs are receiving a constant voltage so they'll be ON all the time, the FET is pumping power into R6 which accomplishes nothing. if you used a less shitty schematic program, the error would be obvious to the eye in 1/10 of a second.
also, your bass filter is kinda useless, using just 2 caps and 2 resistors. you wanna use a 4-th order filter, at least.
also, the peak detector is less than ideal, coz you're losing any signal less than 0.7V. there are ways to make the diode drop disappear using a diff op-amp configuration.

>>1043156
You have a frequency-dependent gain and input resistance. Look up AC analysis and Bode plots.

I was reading on the Colossus computer and how it was used to pick German codes and then decoded them.

Also how it was able to read german coded Radio transmissions.

How do you ''read coded radio transmissions''?

>>1043083
Safety does not mean just lower rating. Better meters will have overvoltage protection circuits and shitty meters use components which can't handle the excess voltage.

>>1043156
Like the other anon said, R2 and C3 affect the gain. More specifically, assuming R2 is fed from a low impedance source, the impedance of paralleled R2 and C3 adds to R1.

>>1043219
I'm pretty sure humans received the radio transmissions (in Morse code) and then fed the computer with the character data.
Well, it's not like you can't connect a radio to a computer. That's how cell phones, wifi, etc. work. In the simplest case you compare the output of the radio receiver against some threshold value and if it's above (signal present), that's handled as 1 and if it's below (no signal), it's 0. That way Morse code can be converted to a series of bits which you can then compare against the Morse alphabet.

>>1043156
Not really. You get a transfer function for the LPF and another for the op-amp. Multiply them for the final transfer function. To calculate gain, get the bode plot for the LPF and adjust gain accordingly to your specifications. Calculation wise they're treated separately.

>>1043231
>Calculation wise they're treated separately.
Only if R1 is much bigger than R2.

Can anybody explain me how does the transistor amplify? I just cant wrap my head around the fact that more electrons are flowing due to a current affecting the base

Where are these electrons coming from?

>>1043245
A common simple explanation is that the base voltage attracts electrons (assuming NPN) from the emitter and, since the base region is very thin, most of them pass through the base to the collector. Due to this, the collector current is higher than the base current and the base current is kinda accidental. This also means that the bipolar transistor can be seen as a voltage driven component.

File: BT_explanation.png (55KB, 1293x792px) Image search: [Google]

55KB, 1293x792px

>>1043245
I'll use NPN transistor for the explanation, if you are interested in JFET or IGBT workings just say so.

So the B-E junction is basically a diode (thus the 0.7V voltage drop). If you forward bias it, current will flow from B to E. This means that electrons flow from E to B. In a normal diode electrons would come out of the B again on the positive end, but in NPN when they enter the base, they feel electric field of the depletion zone sucking them into the C region, and when they enter it the supply voltage applied to C fully takes over and they become the collector current.

There are 4 conditions for proper transistor effect in BTs which lead to high Beta/hfe (current gain):
1) the B is very thin compared to E and C, which makes it easier for the electrons to be dragged out of the B by the C-B depletion zone's electric field.
2) the "exit" of the B is far from the C and E areas, so the electrons can't easily get out of the B, which increases the chance of them getting sucked up by the C-B junction.
3) E is heavily doped (N++), more than B and C, so there are billions of electrons waiting at the B-E junction. When the B-E diode is forward biased, the energy barrier preventing them from leaving E disappears and billions of electrons suddenly enter B (not because they go where positive voltage is but because of different carrier concentrations, its diffusion current we are talking about!) and are then sucked up by the C.
4) the C-B junction area is wide so it has an easier time sucking all the electrons from the B region.

The gain comes from the fact that you need only a small current of electrons flowing from E to B to forward bias the B-E diode, which in turn lowers/removes the energy barrier for diffusion and causes billions of electrons to enter the B, where they are sucked up by the reverse biased C junction.
Its much more complicated but this should hopefully suffice.

>>1043257
meant to say IGFET not IGBT there... IGBT is basically just a BT controlled by IGFET anyway

>>1043083
>So if I measured a voltage up to 2kV with a meter rated for maximum 750V, would I actually get valid readings?
You might just get dashes in the LCD display, or you might get a pop and some smoke, or you might get fireworks.

Here's some OSHA reports to read--
1000-volt Fluke meter connected to 4160 volts:
https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id=200535078

600-volt Fluke connected to 6900 volts:
https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id=200962611

measuring 480 volts on a Fluke meter--using the amps jack:
https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id=201170289

>>1043202
> if you used a less shitty schematic program, the error would be obvious to the eye in 1/10 of a second.
you can make legible circuits in LTSpice just fine. Can you name a better spice program?
>coz you're losing any signal less than 0.7V
because schottky diodes aren't a thing

>>1043284
>you might get a pop and some smoke, or you might get fireworks.
But what would exactly be the cause of the fault of my meter if I tried to e.g. measure a 1.8kV voltage with the setup from >>1043006 ?
The current in that case would be 180uA, I think it's safe to ignore the digital panel meter because of its very high input resistance, power dissipated on resistors from top to bottom would be ~0.3W, ~30mW, 3mW, 0.3mW, ~30uW, voltage drop on resistors from up to bottom would be... ok, I see what might go wrong here: 1620V, 162V, 16.2V, 1.6V, 180mV. But what if we put a bunch of resistors with smaller resistances in series to avoid exceeding the breakdown voltage?

>>1043328
>But what if we put a bunch of resistors with smaller resistances in series to avoid exceeding the breakdown voltage?

That's what's inside a high voltage probe.

>>1043328
>But what if we put a bunch of resistors with smaller resistances in series to avoid exceeding the breakdown voltage?
Are we talking about your very own multimeter design, which can be just as awesome as you want to, or are we still talking about applying 2kV to a 600V rated normal multimeter?

If it's still the latter, then there's no reason to assume your 9M resistor can handle 300mW or 1600V. There's no reason to assume the switch, insulation or any other input side components are designed to handle 2kV, either. There might be (will be, unless it's a really bad meter) overvoltage protection circuits as well, which are likely to distort your readings at higher input voltages. Some fancier meters might also warn you about the excessive voltage and refuse to display the actual value.

Nothing prevents you from using an external voltage divider (or series resistor) to measure >600V voltages. Such dividers are available ready-made.

What exactly is an analog signal?

Is a digital signal composed only of data as in 0 and 1,s?

>>1043338
I was just wondering what exactly prevents those multimeters from measuring above those 600V, so that's why I used that example. Thanks for the answer.

>>1043328
>But what would exactly be the cause of the fault of my meter if I tried to e.g. measure a 1.8kV voltage with the setup from >>1043006 ?
There is two potential problems with extremely-high voltages, even if the current is not enough to do damage.
One is that the electricity will arc across PCB tracks or component wires inside the multimeter.
There is a test video online where a guy puts a couple thousand volts pulse into one of the super-cheapo $5 DT830 multimeters. The meter isn't damaged but it produces no reading, and [with the multimeter PCB removed from the case] arcing is observed across two of the bare PCB knob traces. So it fails at ~2KV, because even tho the 2KV pulse doesn't exactly HURT the meter--beyond that amount of voltage, it won't obtain any volts reading.
The rule of thumb is that arcing requires roughly 10KV per centimeter depending on other factors (humidity, dust in the air).

Another problem is that resistors can break down internally under too much voltage/current. Most typical resistors are only rated for about 200v maximum drop. You can buy resistors that are rated for 1000v or more; these are usually power resistors made of the "sand-stone" type stuff.

,,,,,,,

You can easily make a high-voltage divider probe for measuring tens of thousands of volts if you want; people online show how to do it for directly measuring voltages from Tesla coils and the like.
You can buy them too but a voltage divider is all it really is inside.
Multimeters don't come already made like this (capable of measuring tens of thousands of volts) because of the internal-arcing reason.

>>1043339
>What exactly is an analog signal?
>Is a digital signal composed only of data as in 0 and 1,s?
An analog signal is one where any value is presumed to be (possibly) significant.

A digital signal only has two significant values. Usually the dividing point is around 4 to 4.5 volts, but it can be set at any level within the circuit's acceptable range given the components.
Anything equal to or above the dividing point is considered "high" and anything below it is considered "low".

The part you use for finding out if a digital signal is low or high is called a comparator.
A common logical comparator example is the LM393. For a high signal it will output 5v, and for a low signal it outputs zero volts.
https://www.google.com/search?q=lm393+comparator

>>1043339
Analog = "how much"
Digital = "how many"

I.e. analog is measured, digital is counted.

File: 2-2.png (19KB, 431x541px) Image search: [Google]

19KB, 431x541px

>>1043339
Digital signals are pulses of 1s or 0s, so they look like square pulses with varying time of the high or low states on an oscilloscope. They are interpreted as such, so they can easily send binary information, i.e. 1011 or high-low-high-low is interpreted as number "11" etc.Voltage level of low is usually 0 V, voltage level of high depends but common values are 5V or 3.3V. To read digital signals you need a clock (a symmetrical 50% duty square pulse that never changes a bit) that will go along the signal so the interpreting circuit knows when the next state of the signal is supposed to start. Imagine you get one milisecond of continous 5V, how else would you know how many HIGH states in series is it supposed to represent without a clock telling you how many ticks you went through?

Analog signals don't have any significant values, or rather all of them can be significant. Typical analog signals are audio signals that are sent to the speaker: its not a bunch of 1s or 0s, its a very complicated superposition of waves of varying frequencies and amplitudes that directly converts into sound. Sound itself is analog signal too, though mechanical, not electrical. Reality itself is analog and in electronics we try to approximately convert it to digital because digital processing is much more clearer and less prone to noise, losses, parasitic properties and other unwanted effects. However some information is inevitably lost in the analog-to-digital conversion process.

>>1042589
I already have one you bastud
>>1042778
I figured that, but the packaging shows it being used on a PCB.

>>1043352
Technically, most protocols have an undefined regions - specifically that a range cannot be used as a signal. Like 0-1.5 low, 1.5-4 undefined, 4-5 high.

>>1042597
Sounds like what you're doing is a lot more Electronics Engineering or something. Electronics and Electrical have a lot of overlap I guess, and in some places are "interchangeable", but I've done very different courses (the physics for example).

Nothing wrong with what you're doing, and you've done a few more specifically related to hardware and logic than I have.

I forgot to mention I had an Embedded Systems course. I programmed Pac Man on the Landtiger LPC1768. That was a trip because I am not very good at programming.

>>1043310
Multisim 13 is nice. I prefer it to 14, but they're identical 99% otherwise.

>>1043352
Digital doesn't mean two discreet levels it just means quantized signal. Digital is not necessarily binary it's just discontinuous while analog is continuous

>>1043375
Probably the least useful explanation I ever heard lol

>>1043438
Again, digital is not necessarily binary.
About the most well known protocol is arguably rs232 which uses -12v for '1' and 12v for '0'.
It also doesn't require a dedicated clock signal, the coding clocks are independent and synchronised at the start of each bit. When you do need a clock it doesn't need to be regular or balanced because that's the point of the clock, to let the receiver know when to sample, if it had to be regular and fixed it wouldn't be necessary.
Look at Manchester encoding for example.

Also, real world digital signals are just as Fourier-able as analog signals and so are also a complex superposition of waves. Everything is ' analog' because our world is. Except its not its it, because everything is atoms and shit it can be quantized as such. Some things like time and space are truly analog, electricity kind of is, current can be but say the charge on a capacitor is a discreet number of electrons or whatever.

>>1043017
>Is there any advantage to 4 wheels or disadvantage to the common 2 wheel design? If 4 wheel designs are better, why is every single design I see at world championships 2 wheels?
usually the line-following contests have an overall size limit that prevents you from building anything big enough to benefit from four wheels.
also a chassis with only two wheels can do a zero-radius turn, but that isn't important at high speed

I dunno why he says to use a 4-wheel car.
what would benefit more would be just two wheels, but with some pins sticking out of the tires a bit--to get better traction on the carpet

What's the downside of a boost converter ?

>>1043352
>ethernet is analog

>>1043773
Compared to what? Flyback converter? Forward converter? Full/half bridge? Charge pump?

The main issue with boost converters is that discontinuous-conduction mode requires peak current to be at least double the mean current, which is an issue for core size and i^2*R losses, while continuous-conduction mode is complex (you basically need two nested control loops, with a full PID controller for the outer loop).

File: IMG_20160121_160613.jpg (2MB, 2448x3264px) Image search: [Google]

2MB, 2448x3264px

Can I "multi-task" on Arduino UNO?
I want to run both I2C-LCD to LCDSmartie on my PC over serial and want to control RGB LEDs at the same time. Can I just write the code in a single row and hope it works?

>>1043831

Whatever you are doing probably doesn't need to be at the exact same time

i'm doing my own guitar tube amp - should i even bother with a tube rectifier or just mount a diode bridge?

>>1043831

any CPU with interrupts can easily be set up to multi-task. you run your heavier task as the main program and the other task (or tasks) run in the interrupt. so, if you set up a 1kHz timer to generate an interrupt, then 1000 times a second, the main program is temporarily paused while the second task runs. just make sure you dont do too much work in the interrupt, make it exit after, lets say, 500usecs, so the tasks have an equal chance at the CPU.

Question concerning decoupling capacitors.

What's the frequency range I can get away with placing decoupling capacitors on the back side of a PCB directly beneath the IC?

I'm doing an AVR board with a crystal of 16MHz (had one at hand) and I want to go really small so placing the caps beneath would really help compact things. Going down in Hz and using the internal clock are considerable options since the board will be used for very simple hobby stuff.

How can I tell if the chip in an Arduino is fried? I got a few clones from China, and the one now refuses to have anything uploaded to it. After the bootloader runs, the 13 LED is on, which I didn't tell it to do.

>>1043831
This is a pretty common embedded system problem. You can try putting everything in a tight loop and see if you can get the latency low enough to hit your requirements. (This will probably work if it's just those 3 things.)

The only real issue is probably the serial, as you need to grab characters before the receive buffer gets overwritten. So you could have an interrupt handler take care of receiving bytes from the PC, and everything else in a tight loop, as such:

isr triggered by uart receive: receives byte into a buffer, sets flag for main loop

main loop:
{
check uart read flag, handle reading character;
transmit over uart if needed;
transmit over i2c if needed;
set RGB values if needed;
}

This probably isn't needed though. As long as the worst case latency for the entire main loop is faster than fastest time between receiving characters, you'll be fine.

File: file.png (69KB, 1920x1040px) Image search: [Google]

69KB, 1920x1040px

I can't get the output from this op-amp higher than 700mV. The gain doesn't seem to be as simple as R4/R10 because of the filters I've got before that. The graph shows voltage at the output and at the inverting input.

Does anyone have any suggestions before I start actually calculating transfer functions? I need to retain the same cutoff frequency on the filters but otherwise I can change the components around as much as I want. R3 is just an example load.

>>1044007
If you have a programmer like the avrispmkii, you can load up Atmel Studio and have it identify the chip/reflash the bootloader.

>>1043981
As long as the vias are still close to the pins, I don't think it'll make a difference. It's all very rule of thumb. It'll probably even work fine in 95% of cases without any decoupling caps. Atmel has an app note on hardware considerations for AVR boards with some stuff on decoupling caps.

>>1044017
Have you checked the values at the input to the op-amp? What's the cutoff frequency you're trying to achieve? Easy way out would be active filters.

File: tmp.png (12KB, 752x535px) Image search: [Google]

12KB, 752x535px

>>1044018
This is pretty much where I'm at with the vias for 5v supply, haven't added ground vias yet. I would like them inpad, but can't do it because hand soldering will become a nightmare.

>>1044027
I think you'll be more than fine.

File: Active Chebyshev Bandpass Filter Calculator.png (9KB, 858x279px) Image search: [Google]

9KB, 858x279px

>>1044017

you cant chain low-pass filters like that. they interfere with each other and you get a mystery filter. two things you can do, (1) find a Chebyshev 4-th order filter calculator on the web, or (2) add an op-amp buffer between each stage.

>>1044063
Oh, didn't know that. That explains why my waveform looked retarded. Thanks.

>>1044027
Be wary of "hiding" vias underneath the chip. If you make a mistake in the layout, these "hidden" vias may prevent you from fixing the board after production.

>>1043873
depends on how you want the amp to sound. what kind of sound are you going for?

>>1044091
Thanks for the warning mate, but I'm the one doing the production so no worries there, got the shit tight knit.

File: tankrat.png (26KB, 929x648px) Image search: [Google]

26KB, 929x648px

I'm just past the beginner threshold with embedded electronics (I can make simple microcontroller circuits to operate/read sensor modules and display on LCD panels) and I want to start this major project (pic related).

My pet rat lost one of her ear drums to an infection about a year ago and can no longer use her ball to explore the house as she has lost her balance.

I want to build a vehicle that she can control by where she stands in the cockpit of the vehcile that will control turning and forward/back using either pressure sensors as in the picture or possibly mounting the platform to a joystick module (though I can't really think of feasible way this could work).

I have most of what I need but still need to select an appropriate tank chassis that comes with servos. It needs to be big enough for an adult rat to sit in comfortably and move around a little bit.

It will definitely need to be able to store everything inside the chassis and leave the top completely bare as I will need all the room for the cockpit.

I don't have a price range in mind but I don't want something crazy expensive.

Ebay and aliexpress are pretty vague about the dimensions of tank chassis' so I can't be too sure I'm buying the right thing.

Can anyone suggest some suitable tank chassis for this project?

>>1044143
Why does it need to be a tank chassis? If I were you I'd modify an RC car.

>>1044144
I've considered this but I think a tank would be better for a few reasons:
- I feel the steering controls I'm going to need to implement are more suited to a tank-track turning system - I want it to be intuitive in case she actually does figure out how to operate it (she's pretty smart).
- I want it to be able to drive up the little lips/steps some of my doorways have.
- I want to have sensors that detect if the vehicle is about to drive off a 'cliff' and stop the vehicle from going in that direction if this is too high (probably 10cm would be too high). A tank would have a far greatly ability to actually stop and not fall off the edge (if one RC wheel went over the edge there probably wouldn't be any chance of the vehicle recovering itself).

File: method.png (11KB, 395x274px) Image search: [Google]

11KB, 395x274px

What method should I use to solve this?

>>1044154
What are you trying to solve for?

>>1044155

Voltages and currents on the Diodes, but I am too stupid I dont know why, its like I cant do anything even tho I study and study I cant even solve this simple circuit

>>1044156
Voltage drop across a forward bias ideal diode is 0.7v. For current, you can use nodal analysis at the top node.

File: answer.png (25KB, 659x447px) Image search: [Google]

25KB, 659x447px

>>1044154
Thankfully the diodes make a "virtual ground" because we can assume the voltage drop is the same.

>>1044143
You can look up inductive pads. They're used in diy DDR pads, you'd have to test for what the threshold is for "rat". If you find an easy-to-push joystick you could also use that.

>>1044163
Those are kinda annoying to calibrate/filter, though. I'd look into force sensitive resistors first.

>>1044154
Another approach is to use the exponential model in conjunction with information on the datasheet (1N4007 has a current of 10mA at 0.7V). You do some KCL at the top node and KVL on the diode branches, then sub in the exponential model equation with data sheet values and tell the computer to solve the rest because you can't solve it by hand. The answer you get will be within a couple hundredths of a volt of the 0.7V drop model.

Hey guys. I have a 5w speaker at 3 ohms, I can't find the data sheet and I was wondering if someone could point me to something that could tell me how to drive it. For example, I want to know if I need 5 watts going into it at all times or is that the max? Is it peak to peak or rms? Anything about driving speakers would be awesome.

>>1041443
Buy it and test is so other people (me) would know!

I want to buy one, Baku branded, but I am also on the fence.

>>1044027
It's not a problem with AVRs. They're pretty forgiving to begin with and might work just fine without any bypass capacitors.
One way to think about it is "would it be ok with through hole capacitors". If it is, then the via inductance isn't going to ruin your day.

>>1044063
It is kinda mystery filter even with buffers between the filter stages, as the cut-off frequency moves down as you add more stages.
>>1044084
Your amplifier loads your filter heavily and the comically high gain does not help either.

I just got a sensor and an arduino breakout board, and I'd like to try the sensor out on an mbed system at some point. Is there any way to use an adruino breakout board with mbed? Or is that impossible?

>>1044017
The low value of R10 is loading the filters quite heavily. Probe the non-inverting input to determine the total attenuation of the passive section.

Also, The value of R4 is high enough to make the op-amp's input impedance significant. What happens if you replace U1 with an ideal op-amp?

Practical filters use an op-amp per stage rather than just cascading passive stages and boosting the signal at the end. A large part of the reason for that is that with passive stages, a change in a component value not only changes that stage's transfer function, it also changes its input and output impedances, meaning that each component value affects every stage. The net result is that the filter's transfer function can change significantly with small changes in component values, i.e. you'd need unrealistic tolerance and temperature coefficients.

>>1043831
> Can I "multi-task" on Arduino UNO?
You can multi-task on anything.

It will be simpler if you can just merge the main loops. Concurrent programming requires a certain amount of knowledge regarding race conditions, critical sections, etc and the ability to apply it.

And don't try to do too much work in interrupt handlers if you have multiple interrupt sources enabled. You need to correctly handle the situation where multiple devices all generate interrupts concurrently.

>>1043873
> should i even bother with a tube rectifier or just mount a diode bridge?
Neither. Use a switching PSU.

Unless you're actively trying to get some mains hum on the power line for that retroshit feel, DC is DC; it makes no difference how you generate it.

>>1044229
Obviously he wants that retroshit feel, or he wouldn't make a tube amp.

>DC is DC; it makes no difference how you generate it.
The reason why he would use a tube rectifier is that their output sags noticeably under load.
Also, if a retroshit PSU gives him mains hum, then your switcher gives him switcher noise.

>>1044245
>>1044229
>>1044092
eh, don't care too much about sound. i've torn apart a tube amp, and putting it together in a new configuration. i don't want to use a tube rectifier unless there's a good reason to do so, the sag under load does not appeal to me...

thanks for the responses.

>>1044207
>would it be ok with through hole capacitors
Thanks mate, this really made me think I'm overly paranoid since I know from experience AVRs do work quite splendid with through holes.

>>1044218
Bump for answers on this

>>1041252
Anyone? I'll be Googling hard anyway, but there must be someone with practical experience here (or not) with some tips.

>>1044326
Yes it's possible, grab the sensor datasheet, connect the necessary pins and code the mbed to interface with it.

File: file.png (30KB, 1598x1054px) Image search: [Google]

30KB, 1598x1054px

I'm trying to design an inverting second order low-pass filter. Following this design,

http://www.learnabout-electronics.org/Amplifiers/images/op-amp-2nd-order-LPF.gif
my op-amp output rails.

Am I doing something wrong? Is it even possible to have two filter stages and invert with one op-amp?

I'll reply to this post with pics of what I get by swapping the inverting and non-inverting inputs on the op-amp, and by grounding the non-inverting input and moving the R2/R4 junction to the inverting input, as well as the signal.

File: file.png (46KB, 1599x1099px) Image search: [Google]

46KB, 1599x1099px

>>1044538

File: file.png (59KB, 1599x1104px) Image search: [Google]

59KB, 1599x1104px

>>1044541
Pretty sure the gain resistors are messing with the filters or something. The output looks bad here.

>>1044545
And here's the signal from V1.

File: file.png (44KB, 1598x539px) Image search: [Google]

44KB, 1598x539px

>>1044548

>>1044245
> The reason why he would use a tube rectifier is that their output sags noticeably under load.
Which achieves what? And even if that's useful, a resistor does the same thing more simply.

> your switcher gives him switcher noise.
Only the very cheapest crap PSUs allow sub-MHz noise in the output. In practice, switching noise is generally limited to noise well into the MHz range, which can't simply be filtered out because it's capacitively and inductively coupled (i.e. broadcast as radio waves).

>>1044538
A Sallen-Key filter requires a non-inverting amplifier. You can't just replace the output (which is part of the circuit) with a 180-degree phase-shifted version and expect it not to affect the transfer function.

Also: you'd typically use larger R and smaller C (only the R*C products and R/R and C/C ratios matter; multiplying every R and dividing every C by the same factor leaves the frequency response unchanged). Otherwise, at least increase the value of the blocking capacitor so that its impedance isn't significant.

>>1044511
Thx

>>1044581
Not necessarily true regarding SMPSs. You /can/ get away with it in something as sensitive as a guitar tube amplifier, but you may have to put the SMPS in it's own metal box.

>>1044538
Do a .ac analysis and check your gain stability.

>>1044581
>Which achieves what?
It achieves that all-important retroshit feel, duh.
Doubly so, when the listener knows it's a tube amp with a tube rectifier.
>a resistor does the same thing more simply.
Probably in double-blind tests, but I saw an article just a couple of days ago where some faggot tested different brands of rectifier tubes (no, not different types) and claimed that they sounded different.
That said, the voltage loss in rectifier tube is not entirely linear, like you'd expect from a series resistor.

>Only the very cheapest crap PSUs allow sub-MHz noise in the output.
I don't know what is your idea of the "very cheapest crap", but off-line switchers commonly run at relatively low frequencies to reduce switching losses. And then enter discontinuous mode at light loads.

This point is probably moot, though, since that amp builder is taking the components from some old tube amp. I'd assume he has a suitable transformer and the filter capacitors already.

>>1044698
> That said, the voltage loss in rectifier tube is not entirely linear, like you'd expect from a series resistor.
The thing is ... the rectifiers in a PSU are unloaded for most of the cycle; it's only near the peaks where they're actually conducting. If that's a significant proportion of the cycle, the filter cap is way too small and you aren't going to hear that "warm" tube sound over all the mains hum.

> I don't know what is your idea of the "very cheapest crap", but off-line switchers commonly run at relatively low frequencies to reduce switching losses. And then enter discontinuous mode at light loads.
Right. But any decent PSU will filter that out. The biggest problem is with very high frequencies which get capacitively and/or inductively coupled, bypassing line filters. But that's not a problem for audio unless it's being modulated by low-frequency oscillations in the control loop (which is also something a good switching design will avoid).

30 years ago, switcher noise was "uh, well that's just what switchers are like". But those issues have since been solved. Or, at least, we know how to solve them if you actually care about them, rather than about cost and cost alone. Nowadays even high-end test gear uses switching PSUs, and much of that is a good deal more sensitive to noise than a guitar amp.

>>1044228
thanks :)

File: IMG_20160825_104539.jpg (456KB, 1600x1200px) Image search: [Google]

456KB, 1600x1200px

Anyone knows what component is this?

>>1044834
I think it's a capacitor but the packaging seems kind of strange

>>1044834
pretty sure its an inductor. 101 means 10 and 1 zero of microhenrys (similar to pFs with caps), so 100uH

>>1044840
I see, thank you.

>>1044834
default terms search worked.
1. its surface mount so SMD
2. its a coil of some sort, inductor transformer whatever, COIL
3rd marking is 101
SMD COIL 101
worked, google it

>>1044753
>you aren't going to hear that "warm" tube sound
When a guitarist wants a tube rectifier, he generally wants it due to the distortion effect he gets when he drives the amplifier into clipping. That, plus the effect the quickly dropping voltage of the filter capacitor causes.

>But any decent PSU will filter that out.
Pretty often smaller switchers have just an output filter capacitor. Better ones have an extra LC filter, but it's normal to see ripple after it. It's users job to add more filtering and shielding if that's not enough.
>modulated by low-frequency oscillations in the control loop (which is also something a good switching design will avoid).
When the load is light enough, it is normal that switchers start skipping pulses - simply because producing narrow enough pulses to maintain regulation is not possible.
>Nowadays even high-end test gear uses switching PSUs
Yeah, sure, it's not an unsolvable problem. Just like the mains hum of the old-fashioned PSUs wasn't an unsolvable problem.

File: experiment11.jpg (75KB, 308x720px) Image search: [Google]

75KB, 308x720px

This is experiment 11 on Make Electronics book.

I noticed that It requires me to connect two 3.3 microfarad.

But I noticed that both leads on the capacitors are connected to positive through the resistors.

Does this mean I could only use ceramic capacitors in this experiment? Or can I use electrolytic?

>>1044890
> can I use electrolytic?
Yes. The collector side is +ve, the base side -ve.

File: 1472345297955.jpg (112KB, 2296x1440px) Image search: [Google]

112KB, 2296x1440px

Stupid question, but what is my scope picking up

>>1041909
they're BP you can see in the photo
I have the same ones

no idea why this guy is using old paper caps though

Reposting from /g/ because I didn't get an answer in their questions thread. I'm hoping this is an appropriate thread since my question is about interference in a circuit (I think...?)

Unfamiliar with vehicle audio but bear with me it you will.

I connect my phone via 3.5mm to my car's audio system and at full phone volume I still have to turn the car audio to 60-75% when the radio is "loud" at 15%. Eventually the interface panel gets hotter and there's a whine noise in the audio; this persists so long as the 3.5mm jack is plugged in to the car (even if the music is paused or the phone is unplugged).
1) is the heat something I should fix?
2) if the heat isn't the problem, what would I need to get rid of the noise? An add-on or a repair?
3) should I get some kind of portable amplifier so I don't have to crank all the audio or is that no big deal?

>>1045062
>via 3.5mm
cable wired wrong?
are you powering the phone from a 12vdc charger when it's plugged to the car system?
>unplug the phone charger

>>1045092
Just a 3.5mm aux that carries right+left signals with no mic signal.
It does occur without being charged.

I really think it has something to do with the whole box heating up but I have no clue if that's normal behavior.

>>1045124

you need to check the docs on your phone to determine if the jack is a 4-way jack, L+R+mic+ground. if so, there's a good possibility the cable you're using is shorting 2 pins in the phone, and you need a special 4-way to 3-way cable, which you may have to wire on your own.