/diy/ - Do It Yourself

Do any of you have experience with voltage inverters? I'm using a TC7660S with 12V input and it's only outputting -7V in my design. If I isolate it it gets all the way down to -9V. I've got a 10uF polarized cap on it just like it asks for. Why is it not outputting the right voltage?

>>1050988
It needs an output capacitor. They also tend to have pretty high output impedance.

>>1050988
How much current are you drawing? The output impedance is around 50 ohms, i.e. it will drop a volt for every 20mA drawn.

Also, ensure that your caps have low ESR. The RC time constant needs to be much less than the switching period. Electrolytics intended for filtering 120Hz ripple won't necessarily suffice at 45kHz.

>>1050995
Shouldn't be anywhere near enough to bring it that low. I've got it supplying ~6 op-amps that need 600uA each.

The cap is 3.2 ohm SER.

>>1050992
Isn't that just decoupling?

>>1051009
Come to think of it, it's actually only 3 op-amps because they're two-circuit chips.

>>1051009
>Isn't that just decoupling?
No, it's the output reservoir capacitor and the output is half-AC without it. Read how the chip works.

>finally try to learn how to work with AVR microcontrollers
>use an Arduino Uno as an ISP, programming an AVR328P
>set the Arduino as a ArduinoISP, then set Arduino as ISP
>try to upload the bootloader because apparently that's a thing you need to do if you don't have an oscillator handy, get a not in sync error or something

I'm retarded, how do I use an Arduino to program a standalone AVR?

>>1051009
> I've got it supplying ~6 op-amps that need 600uA each.
Does that include output current or Is that just the stated quiescent current?

> The cap is 3.2 ohm SER.
That's fairly high. It gives T=32us, fc = ~5 kHz

The graphs in the data sheet are all based upon 1 ohm ESR.

It's not immediately clear whether the charge pump is driven at the oscillator frequency (10 kHz with boost and osc unconnected) or half that. At 5 kHz, a 32us time constant will cost you ~4%, at 10 kHz it's ~20%. I'm unsure whether the output cap ESR will reduce efficiency or just add ripple.

Wanting to make a simple 12V AC -> DC converter
bought some 47µF 16V capacitors.
Will this be enough ?
How can I seize the needed µF capacitance needed ?

>>1051023
Your connections are likely not right if you get not in sync(it cant see the chip). Also you can use avrdude/avrdudess to easliy set fuses which i think is what you mean by bootloader. also if your current fuses are expecting an external clock, you will need to connect one before the chip will work! Or use a HV fuse reseter to reset the fuses to not need an external clock

>>1051254

Vripple = I / (f*C)

Also you need a diode bridge before the caps, obviously. Get higher voltage caps too, 12V RMS will have ~17V peaks when you rectify it so 25V will be enough.

Would a capacitive mic like a CMB-6544PF work with the terminals reversed?

>>1051272
> Vripple = I / (f*C)

Note that f will be doubled, i.e. 100 Hz for 50 Hz mains, 120 Hz for 60 Hz mains.

At 120 Hz, 47uF gives you 177mV/mA ripple (or ~5.6mA/V). IOW, those are probably far too small.

Filter capacitors for transformer-rectifier PSUs are sized in thousands of uF. Which is part of the reason that switch-mode PSUs have taken over (the other, larger part being that the transformer can also be smaller by a similar factor).

> 12V RMS will have ~17V peaks when you rectify it

But you'll lose 1.3V to the diodes (assuming that they're silicon; Schottky diodes only drop ~0.2V each), so 16V would be just enough if you could guarantee that the RMS will never be more than ~12.23V. Which you probably can't (e.g. if it's 12V out for 110V in, it will be over at 120V in), so use 25V instead. And the stated output voltage normally assumes some load; it will be higher at no load.

>>1051265
What even are fuses in an AVR context? Apparently I'm completely lost when it comes to lower-level programming.

>>1051023

You need an oscillator and to set the serial rate depending on the frequency.

2 questions about this:

1) What does the brass terminal to the far right do? This is a switch for DC supplies so not sure its a ground.

2) This is a 20A 14VDC used feor switching a 12VDC supply. Can I turn on the LED with that? Doesn't seem to work but maybe it has to do with the brass terminal.

>>1051371
If it has a built-in lamp, and the lamp is supposed to be on when the switch is closed, it needs 3 terminals. When the switch is closed, there's no voltage across the switch terminals, so the lamp power has to come from somewhere else.

So it's probably 12V in, 12V out (switched), and ground.

small question, i want to install a subwoofer in my car (peak 1300w rms 300w 12v) and i want to put a potentiometer in the signal line so i can turn the bass up or down, but im stumped on what resistance potmeter i need?

>>1051347

They're special, non-volatile bits of memory that govern some touchy settings. Stuff like clock output/division/source, watchdog timer, and reset disable.

A number of them can soft brick a chip if set wrong, and require HV programming to fix.

>>1051371
It also could be just a SPDT switch, are you sure it's not that? Most switches are like that

>>1051381
Use a low pass filter in combination with whatever you're using as an amplifier

Do any of you have experience building circuits with microphones? I finally got a chance to troubleshoot mine with a scope and the mic is apparently doing nothing at all. I'm totally stumped.

http://www.cui.com/product/resource/cmb-6544pf.pdf

There was basically nothing happening on the scope. The only thing that's different from the test circuit is that I used a voltage divider instead of a dedicated 5V supply but the mic needs .5mA so I don't think that should be a problem. Any suggestions? Anything at all. I'm so desperate.

making a buchla lowpass gate-
i don't have the 3pdt on-off-on yet. but it works fine in "both mode"(all leads open)

>>1051428
i didn't have a 3.9v zener, so i improvised.

>>1051429

>>1051417
>file.png
I don't see the voltage source for the mic.
If you had any output the majority of it would be directed to ground by R16 abd C3.

>>1051417
The 1k needs to go to a stable voltage. Get a 7805 or add a big cap to the divider.

>>1051417

so your software is so fucking retarded it doesnt even have a mic symbol?

anyway, your mic may be backwards. usu the lead that connects to the case is negative. also you need to set your scope to the 1 or 2mV scale to see anything.

What would be the practical way to build an USB charger (1A) from a 12V input? Would I be fine with an LM2576 buck converter, connecting one of those to an LM7805 to avoid any possible voltage change (at the cost of dissipating at least 1W at max current) or a different way altogether?
I want to make something useful and learn something, plus paying 6-15€ for thirty cents of parts sounds retarded

>>1051390
That makes sense, thanks.

I need some babby's first microcontroller programming resources, everything I find about programming AVR's seems to assume a fair amount.

>>1051455
Triple-checked my leads today. They were right, but you've got a good point, I never bothered to check when I was building it.

Software is LTSpice and I couldn't find a mic symbol, but it might well have it.

>>1051435
E.g. 33uF cap in parallel with 712R?

>>1051485
Yeah. Also, 0.5mA _is_ going to sag your voltage divider.

>>1051495
How do I avoid that? Scale down R15 and 16? I'll check if I have any smaller resistors that'll maintain the ratio, but if not, I think I'll get a buck converter.

What's the ideal, as low as allowed by the power rating? I don't really care about a few watts wasted.

>>1051507
The range is 4.5 to 10, so it doesn't strictly need to be 5V.

>>1051475
>dissipating at least one watt
Uh, anon, there's a 7 volt differential * 1 A. You lose 7 watts.

>>1051380
It works, thanks!!

Anyone have a suggestion for a cheap handhand scope (about 50-70$) mainly for automotive work (injector pluses, pwm signals), audio work, and maybe switchmode power supply work? Im considering the dso201(60$) and the dso138 (21$). The bandwidth should be enough for most things i would need, just wondering if the dso201 is worth over double compared to the 138 or if theres anything better available in that price range?

>>1051381
A 5k-20k pot would work fine. The two outer terminals get connected to ground and input signal, the center is where your volume controlled signal comes out.

>>1051475
I use LM2596 adjustable boards readily available online cheaply and they work great. You could probably add an LC filter( or just more capacitance) to the output to reduce noise/ripple, but ive been using them without for years with no problems.

>>1051532
You might want to check what data can be captured if you're into that. If you're just scoping, and you don't care about battery, you can get the cheaper one. I have a DSO 112 and it's alright. Technically I can get data off it (it uses a usb-serial internally, so you can use custom software to dump a spreadsheet).

Personally I'd like to have dual-channel, but single is fine for observing. I have adapters for the MCX connector to fullsize, and then generic 10x probes. I was testing some caps for repairing a computer PSU by charging them with rectified AC. God damn that's stressful (I could watch the voltage on the scope rise to the proper levels).

>>1051532

Check out DSO068. It's a kit scope but you won't have to fuck around with a crappy touchscreen and it has larger bandwith than those two and also FFT.

>>1051540
Data sniffing would be interesting, maybe for i2c data. But i dont think id use it except to play around with. Multiple channels would be useful though.
>>1051545
Looks to be the same price as the dso201, but better bandwidth and ability to use it with a pc application. I would guess the pc app makes it much more usable for sitting at the bench. Any hands on experience? Apparently the ds201 has custom firmware that fixes many bugs, i didn't find anything like that for the 068, but its not a dealbreaker. Overall it looks good for the price!

>>1051548

>Any hands on experience?

Mine should be here next week, I can write a quick review here when i put it together.

>>1051525
I meant using an efficient converter to about 7, maybe 8 volts and giving that to the linear voltare regulator

>>1051588

No need for the linear regulator, just set up the 2576 for 5V output.

>>1051381
Potentiometer resistance ideally needs to be large compared to the output impedance of whatever is driving it and small compared to the input impedance of whatever is being driven from it.

Too low a resistance will load the preceding stage. Too high a resistance will cause non-linearity, reducing the amount of variation at the low end of the knob's range and packing it into the high end.

The non-linearity isn't too bad provided that the potentiometer resistance is no higher than the input resistance of the following stage.

>>1051475
No need for a linear regulator. The buck regulator will be quite sufficient.

If the "12V" supply is automotive, be sure to add a reasonable LC filter on the input to remove short, high-voltage transients which could fry the chip. L doesn't need to be particularly large provided that C has low ESR/ESL.

I want a soldering station for my birthday and I have about 50-100€ as a budget
I want to go temperature-controlled as I'm sick and tired of my 20€ variable-power one
What do you suggest?

>>1051711
if you cant gid gud on your small one, there is no point in investing in a better one t b h

>>1051715
I just need more juice for TO-220 packs and the like, I do ~fine on smaller stuff

How bad an idea is cutting off a 110V cord and soldering a 220V plug to it?
Asking for, uh, a friend who ordered a 110V PSU from China

>>1051720
Very bad, it's a full retard idea. You need some transformer to use 110V stuff with 220V mains.

>>1051720

In general it should be fine but if you're using a plug from a donor cable be sure to splice it well since there won't be any real strain relief. Also use heat shrink tubing after soldering.

>>1051717

Hakko fx-888d, it's a no brainer.

>>1051727

Most SMPS will take either, I'm assuming he checked beforehand.

I need to lern2FPGAs. In my class we use alterra's Cyclone V. And the software we use is called quartus. The thing is the software is a bit of retarded. Model sim has no functionality to alter the state of all inputs so we need make a .do file with a script in it telling to change it manually. The thing is that feels retarded to me. Does anyone know of a better way?

>>1051741
You're not using testbenches?

>>1051761
I don't think that Quartus 2 has Verilog HDL or VHDL testbenches.

>>1051763
Quartus 2 definitely has testbenches mate. I've written several in the past.

Wait until I get home and I'll walk you through how to write one and upload it. On phone right now, would be too much typing and can't take screenshots.

>>1051769
Ok. Thanks. I never used a FPGA before. I have no idea of what I'm doing and my classes are only once in a week and short. Thanks in advance.

Where do you guys get your components, I'm a hobbyist and I'm having trouble finding a place with good qc

>>1051864
Digikey, Mouser, Farnell, etc. offer good, reliable components from known makers - at a price.
Ebay's Chinamen offer sometimes ok stuff, sometimes total shit at low price.

>>1051867
I'm fine with going online but just incase, do you know any retail in the us?

>>1051870

RadioShack used to be good, but you'd be better off googling for electronics shops in your local area these days.

Do any of you guys have a degree in Mech. E but work mostly in electronics? My interests lie in electronic junk, but I'm 3/4 of the way through ME and starting to worry I'll get trapped in a job away from EE topics

>>1051916

You don't need an EE degree to get into this.
With your ME background you're likely comfortable with numbers, something you'll deal with a lot when you dive into the deep end of EE.

Just get some parts and start farting around.
Get a book as the OP suggests, you'll do grand.

>>1051916
I'm 1/4 in EE and as >>1052109 said you definitively don't need a degree to do that. I have very little experience and knowledge in electronics and I'm learning about that in my course.

Hey /ohm/

Is it possible to rectify a waveform that has a +ve offest? As in no part of the waveform goes below 0V.
Thus far I have not been able to think of any circuit that can achieve this.

>>1051711
Hakko FX-888D

>>1052179
> Is it possible to rectify a waveform that has a +ve offest?
Add a DC-blocking capacitor before the rectifier.

If you want to retain the DC offset, then it gets more complex. It may require a an op-amp.

Note that you can't do that passively (without some power source), as it would violate conservation of energy.

>>1052179
Nothing stops you from putting a diode in series and a capacitor at its output. The output will be what you'd expect from a half-wave rectifier.
If you want to achieve the frequency doubling effect in your pic, you could simply high pass filter the input first.
Or if the frequency is fixed, you could phase shift it 90 degrees and then multiply the original with the phase shifted version.

>>1052179
Complicated suggestion? Make a filter that block low frequencies. Easy suggestion? Put a capacitor. Both are the same.

>>1052207
>>1052185
>>1052183
I should have mentioned that what I'm trying to do is rectify a +ve sinusoidal waveform with an active rectifier circuit like the one in pic.
The OPAMP's are single supply (+5V & GND), I find that if I feed in a 1Vpeak 2Voffset sine wave then the output is basically the same as input. However if I make the power dual (+5V & -5V) and then feed in a 1Vpeak waveform (no offsest) then the output is rectified like the picture I posted before.

>>1052244
Yes, of course. The suggestions you got were valid.
In that particular circuit's case you could tie the bottom amplifier's + input to your offset voltage. If you then feed in your offset sine, the output would be like the bottom curve in your previous pic, offseted.
Or just use that capacitor / high pass filter in the input.

>>1051916
You're not going to get hired for a straight EE job like circuit design. But stuff like robotics, if you have the coursework, you should be good. Might need to get a MS or PhD though.

Why are compliance certification such a bitch?

It's just impossible for any smaller projects to get certified. It would increase my NRE tenfold. I'm just a student trying to make a little buck on my work, I don't have tens of thousands of dollars to rent a pre-compliance testing lab and then do the actual certification. So I'm just gonna say fuck it and break the law and sell a product without a CE and FCC and UL and whatever other compliance marks are out there.

>>1051916

If you do a EE project that is legitimately impressive and you know how to use something like that to your advantage, sure

>>1052424
>sell a product without a CE and FCC and UL and whatever other compliance marks are out there.

And who is going to buy it?

>>1052424
So you want to become the equivalent of the Chinese and sell a shoddy product?

>>1052424
If you're selling Arduino shit to hobbyists, no one cares. If you're selling motorized dragon dildos to middle america, then yes, get your shit certified before it burns down somebody's house, or fucks with their wifi.

>>1052424
> Why are compliance certification such a bitch?
Because however much you believe "this is my world, the rest of you just live in it", it isn't actually true.

>>1052506
People buy things from China all the time what are you talking about?

>>1052582
And are the Chinese known for their quality?

>>1052244
Righto, thanks for that everybody, you were right the opamp circuit works just fine I must have not wired it up correctly...

>>1052317
Righto, thanks for that everybody, you were right the opamp circuit works just fine I must have not wired it up correctly...

>>1052607
the chinese are known as the #1 supplier of goods to the NA region as well as most of the world and lead on pretty much every economic front

>>1052582
> People buy things from China all the time what are you talking about?
Uh, what does China have to do with this? He was talking about items lacking CE/UL/FCC certifications.

>>1052506
I've used some similar products from competitors and I've seen that many of them also lack any certifications. Yet I know people are buying them. Maybe not enterprise B2B but that's not my target market anyway, I make small-volume products for hobbyists. 90% of them just want something that works for as cheap as possible.

And it's not like I don't give a shit about good design practice. I've taken several EMC and ESD precautions, and the product is assembled completely lead free (RoHS). It's just that I can't afford to actually test it to verify the limits.

>>1052424
If you want to sell stuff to the consumers in EU countries, your product needs a CE mark. The lack of the mark is very obvious and gets cops to your door quickly. At least no-one sane wants to resell it. So, as a wannabe criminal, you obviously put CE marks to your products. The good thing is that this is not illegal in any way - in itself. No-one forces you to go to tests either, and skipping the emission and immunity tests might be even justified with stuff like flashlights.
By CE marking your product your company guarantees that it fulfills the requirements of all the applicable directives. If it doesn't and you get caught, you have a problem.

>>1052582
Even the Chinese stuff has to have CE marks. Whether they're actually CE compliant is an another thing. The local authority here has a long list of CE marked stuff they've tested and which have failed.

>>1052716
Addition: stuff sold to the other companies has different regulations and is not always required to fulfill the emission and immunity limits.

>>1052716
In my experience it's different, as said I've seen many products that don't have the marking. And I've sold uncertified products myself on a small-time basis for years without anyone asking. And I suspect even if you get caught you just get a slap on the wrist as the product isn't illegal itself, you've just not been paying attention to where you can export it to or not. However if you actually put the CE mark on there without it being actually certified and THEN get caught, you're in much deeper shit as you've actively committed fraud and forgery, which is a serious crime anywhere.

>>1052717
Components aren't required to satisfy EMI requirements, only complete products. Many finished products would fail EMI requirements if you remove the case or line filters.

The same applies to many safety requirements, which are often satisfied by specific components (fuse, TVS, etc) protecting the system against faults, rather than every component being fault-tolerant.

RoHS regulations don't apply directly to components, although if a component violates RoHS, there's nothing you can add to a finished product which will make that violation go away. But you can still ship non-RoHS components for use in exempt products.

>>1052721
Components don't need CE marks to begin with. Motor control shield for Arduino might go to that class. Well, I'm not too sure about the classification, but in general the legal requirements are much easier if your stuff isn't an end product sold to the masses.
Obviously selling a handful of products personally carries much lower risk than selling shitloads via a retailer network.

Have fun optimizing your crimes for maximum profit and lowest risk.

>>1052735
True. That was badly said from my part.

>>1052647
Great job at dodging the question.

>>1052740
I don't mean components, I mean I've seen fully fledged proprietary products sold to end consumers without the CE mark

>>1052745
Sometimes the mark is quite unnoticeable and/or in the packaging only. Not all consumer products are required to have it, either.

Well, I don't know where you live. Maybe things are more lax there.

>>1052751
In the UK, enforcement is largely the responsibility of local Trading Standards authorities, so it's fairly weak. At most they might confiscate items from specific locations if they lack the CE mark, they have it but are obviously non-compliant, or they've received specific complaints.

Other countries are somewhat more rigorous; if you get caught, you can basically give up on trying to sell anything else there.

The other issue is that if the customer (whether the end user or a reseller) isn't happy about being sold a dodgy product, they can just cancel the payment, and you won't exactly have the law on your side.

I need a SP4T SMD multiplexer. Does anyone here know a nice one or where I can find one? The best I can find on Google is a two-in-one, but that's too much.

>>1053175
You know, "nice" is about as vague as it can be. For example, 4052 is cheap and easy to get in many different packages, which is nice.
You could try your luck with Analog Devices' parametric search.

>>1053179
Yeah, sorry about that. I should probably tell more about what I want to do:

I have 4 lines, one of these is at 3.3V, while the others are low, depending on the 2 control pins. I plan to manually place the SMD component and reflow it with hot air. I prefer small parts (so a double SP4T mux would be annoying because a single one is sufficient), but not too small, because I still have to do everything by hand.

>4052 is cheap and easy

Are we talking about the 74HC4052? That's a dual one, unfortunately.

>>1053191
https://www.youtube.com/watch?v=s1X2b0kq9EE&feature=share

>>1053191
>Are we talking about the 74HC4052? That's a dual one, unfortunately.
Yeah, and many others. It is available in QFN, which is pretty easy to solder with hot air. While the unused part wastes some pins, it is still smaller than anything with actual leads.
Funny enough, ADG1204 has only one switch, has less pins and comes in a QFN, but it still needs slightly more PCB area.

>>1053204
So I guess I just go with the dual one? Appears to be the easiest option, though I'll end up with a bunch of unused pins on the board.

>>1053219
If you're really obsessed with size, you should spend some time checking what other options you have and consider using BGAs.
If you actually need a demultiplexer (driving unselected outputs actively low), there might be better options.

What are some good two pin connectors?

They need to be reasonably sized (1cm) and locking or difficult to pull apart.

I like tamiya connectors but they're expensive.

How would you guys use an AVR to determine if a fuse is blown on a 12v electrical system?

I'm looking to monitor up to 30 ATC fuses, ranging from 2 amps to 30.

I was planning on just using a bunch of opto-isolators connected to the digital pins but was wondering if you guys think there is a better way.

>>1053333

Do you require isolation, and can you tolerate a small quiescent current (<100µA per input) for each fuse?

Given your application, I'm going to guess that, yes, you can. In which case, my suggestion is a few daisy-chained parallel-in, serial-out shift registers. Use a resistor divider placed after each fuse to get the 12V down to something the shift register can tolerate.

>>1053351

Isolation probably isn't required but I think it would be a good idea. Mainly because it will be going in a car, so there will be a bit of variation on the "12v" side and unkind transients, but that isn't difficult to deal with.

Also thinking of going with the Teensy 3.2 instead, just learned they now have a CANbus library and good support.

>inb4 using a 72MHz Cortex-M4 to monitor fuses

It's going to be doing a hell of a lot more than that.

>>1052424
>CE

See >>1052716

>FCC

https://www.sparkfun.com/tutorials/398

Basically $1k to $1.5k if you don't meet any of the exemptions and don't want to sell it as an unassembled kit.

>UL

Only needed if it plugs into AC directly (i.e. not through a UL listed plugpack), technically not needed at all in some places.

Is there something like Louis Rossman's channel but for PC motherboards? It would be nice to start repairing them instead of heating circuits and calling it repair.

Anyone have any experience with the Leader LBO-516 Oscilloscope, and is $100 a good price for it?
Taking other suggestions too, I don't expect to buy for another week or so, Thanks in advance.

Finally they arrived.

>>1053899

>>1053899
what is this magic?

>>1053673
No experience, but it looks like a pretty standard old 100MHz analog scope. $100 isn't a bad price if it's in decent condition.

>>1053280
Connectors are expensive, that's the nature of the beast.

In general, how do I make my electronics more robust without using an etched board? For example I have a box with some absolute rotary encoders connected to an IC, but the encoders are on the top part and the IC on the bottom part, and every time I open it up to do whatever I'm afraid the wires will crack.
Been thinking of doing stuff on a circuit board as much as possible, and then using a ribbon cable from the top board to the bottom board, any other ideas?

>>1053940
Good strain reliefs on both ends of the wires or decent connectors plus good cable management taking strain away from the joints/connectors are the usual ways to go if you want a case you can open repeatedly. And of course you should use flexible cable/wire.
There's also the option of not opening the box all the time.

>>1053899
DAT shitty polarity marker for the diode, how will someone know if the component is the right way one its been soldered down. Reeeeeeeeeeeeeeeeee

Cool pic though

>>1053944
>There's also the option of not opening the box all the time.
Yea this is what I was planning on lol but the guy wanted some extra features. I'll look into strain reliefs, I see they make some nice clamps and things for that, thanks anon.

>>1053940
> without using an etched board?
Uh, why? Making PCBs isn't rocket science.

Babby's first breadboard here. I'm a student currently learning electronics for the first time, trying to make a simple circuit I've seen a few times as an example for how transistors work. But I can't figure out what I'm doing wrong. Don't know if it's easy to tell what's going on in the pictures, but if someone could help I'd be super grateful.

>>1054045
The point of the circuit is to use the two leads (green wires) to make a very small charge to switch the transistor and power the LED. But I've got nothin, tried making it a few times now.

>>1054046

One transistor won't cut it, you need a darlington pair.

Why FPGA shit only has proprietary software to use and why all of them are utter and complete shit that looks like it came from the 90s and made to make your life worse? Also VHDL and varilog are shit.

>>1054052

Aren't there a bunch of open source FPGA out there?

>>1051429
The horror

>>1054050
That depends upon the resistance between the probes. The 2N3904 should have h[fe] of at least 60 for Ic in the 10s of mA, so you'd only need 150uA of base current (~60kOhm resistance) for the LED to be clearly lit.

>>1054052
VHDL and Verilog do the job they're meant to, far better than most languages. SystemVerilog is a mess but I'm sure a more experienced Digital ASIC guy will set me straight.

>>1053333

maybe use an output connected to an input across the fuse

>>1054061
Open source platforms have shit IDEs and simulators.

The companies like Altera and Xilinx like to give their software away for "free" but they only work with their boards which cost an arm and a leg.

>>1054045

circuit should work. there's an error in the wiring where you tried to put 220 and 820 ohms in series to get 1K but, in reality, the 820 is shorted, so you only get 220. still, it should work. the 2n3904 is being overdriven by the 220 ohms on the base, but i dont think it's enough to kill it.
measure voltages to see what's what. you should see 0.65V on the base, 0V on the emitter, 0.2V on the collector. make sure that blue wire is inserted properly.

also, 220ohms in series with the LED is bad news. that will drive 33mA into an LED that has a max rating of 20mA.

>>1054213
> also, 220ohms in series with the LED is bad news. that will drive 33mA into an LED that has a max rating of 20mA.
From the photo, it appears to be a green LED, which may drop as much as 4V. In which case, there'll only be 5V across the 220R, which would be 23mA.

I'd still suggest making it larger.(e.g. 680R). The LED doesn't need to be driven to its maximum current.

>>1054069
its heatshrinked now.
i also put some laquer on it, so it wouldn't short out if it got squached.

its only there to limit the amount of juice the vactrols see, so it would't really do much harm if it did short.

i COULD have ordered 50pc of those zeners for 0,99£, but then i'd have to wait.
i whipped this up in minutes.

I've seen worse bodges on safety systems on ships, during my time as a ship technician.

Anyone know anywhere that does custom wound toroidal transformers at a decent price preferably in Europe.
I've a partly built valve amp lying under the stairs for the last year because I'm having trouble finding a power transformer for it.

I need a 160VA toroidal with both 37V and 15V secondaries.

I found one once but it was from an Australian site and they wanted over 100 kangaroo dollars for delivery

>>1051484
This is the tutorial I used to program my ATTiny85 and it worked no problem. Pretty sure the steps are the same for other AVRs, you just have to wire it correctly and use the correct setting in the IDE.
http://highlowtech.org/?p=1695

Can you help me out diy?

I have a home theater receiver that gets quite toasty during normal use. I want to add some fans to help cool it off.

I picked up a pair of 12 volt PWM fans with three wire leads. Thing is, I can not find a circuit to build to drive them and give me the control I want at 12 volts.

I need the circuit to have a thermistor that starts at low resistance when cool (the fans just barely turning) and climb up while the system is warming up. This is so I can set a high speed cap using a pot wired in parallel to the thermistor.

>>1054668
555 astable with the thermistor to control DC would be really bootleg but should work

>>1055051
Can confirm, it works fine. Made a 12v fan 555 controller myself last month. Make sure to select the charging capacitor value that makes the PWM go at >20kHz, so something like two nanofarads or so. Otherwise the coils in the fans are gonna screech like hell.

>>1054668
Sounds dumb, what your skin thinks is hot and what electronics think are hot are two different things.
If by 3 wire fan you mean computer fan then its not pwm the third wire is an output for speed sensing.
A thermistor will keep your fan at low speed for how many second before the equipment heats into steady state? Pointless.
If 12v is too fast run it on 9v and be done with it, find a real project to do.

>>1055068
> If by 3 wire fan you mean computer fan then its not pwm
It is.
> the third wire is an output for speed sensing.
Correct, but doesn't change anything.

PC fans are designed to be driven from PWM.

The original PC fans just used a brushed DC motor. Newer ones use a 4-phase (quadrature) BLDC motor; they filter the 12V line to generate a smooth 12V DC supply and measure the duty cycle on the 12V line to determine the rotation speed.

> If 12v is too fast run it on 9v
9V is probably enough but you can't count on it When connected to a PC (which is what they're designed for), they always get 12V, so that's what the circuit is designed around. Any ability to work on a lower supply voltage is coincidental.

Where does one order chip antennas? I somehow can't find any. Usually I just go to ebay to order parts, but no luck there or on Google.

>>1055184
Suppliers like Farnell and Digikey have them.
Or just use a wire or pcb antenna instead.

>>1055195
But I'm not an electronics magician, I don't know shit about RF and antenna design (besides putting a chip antenna on it based on a provided example). I just want a damn small antenna that I can just put on the board.

>>1055202
Chip antennas suck shit, unless you follow maker's instructions exactly or you actually know what you're doing. If you're capable of doing that, then you're capable of following instructions for making PCB or wire antennas.
Well, the choice is yours. Mouser has them too, btw.

>>1054601
TME sells cheap ass toroidals from Indel. Got one there for my gainclone

>>1050981
>order 15 PCBs from Seeed
>already 7 days overdue, haven't even shipped yet
>other guy orders from PCBway 2 days after I ordered from Seeed
>Already have their boards.

Fuck me, bloody Chinese.

Anyone have experience with this website? Any good?

>>1055331
My experience has been that you can get whatever they're selling for cheaper elsewhere.

>>1055350
Can't seem to find cheaper prices in Australia. What websites should I use?

>>1055331
It's fine but expensive.

>>1052721
Sell these products to the fastest growing/least regulated market in the world: China.

>>1053362
Nice poo-in-loo photo.

>>1055331
Website is okay. That soldering iron is shit. get a used Weller and fix it.

>>1054052
Digital design is dick easy. 75k starting next year.

>>1055470
Everyone says the Hakkos and Wellers are the best. Are Wellers that much better than Hakko?

>>1055476

once again, DIY's endless quest for THE BEST in every fucking category. it generates endless arguments, but is about as useful as a ''tits vs ass'' thread on /b/. both Hakko and Weller are more than perfectly adequate for hobbyists and professionals. neither is the best. if you want the best, then you gotta pay thru the nose for it: like, replacement tips costing $120.

>>1055476

Everyone must be wrong since neither of those options is JBC.

If a USB toy iron is a 1 and a dollar store iron without temperature control is a 2, then a cheap Hakko is about a 5 and a busted up Weller is a 6.

The question is, is the difference between a 5 and a 6 really worth the price difference if you don't need a 9 or 10?

>>1055550
The main reason I need it is for soldering a Z8 QFP in an upcoming project, along with a bunch of 0805 components.
I just need something reliable with interchangeable tips.

>>1055331
Adafruit charges rather high prices, but then,,,, some of the stuff they sell, they designed and build themselves specifically to work better than the typical low-cost China versions.

Another difference is that they have real actual support files (libraries and working code examples) for what they sell.
Some other places like Sparkfun do that too, but at least 98% of the China sellers don't bother with those minor details.
So if you're new at this stuff, it's not a bad choice to spend a bit more and know you will get instructions and code that will [almost certainly] work.

>>1054601
>I need a 160VA toroidal with both 37V and 15V secondaries.
what sort of setup requires this?
are these heaters for some super serious valves?

I'm looking for a decent oscilloscope for looking at radio AM signals (so atleast 1MHz).
eBay has a lot of these portable DSOxxx scopes (I don't care for portability). Are these any good or should I look at a second hand one?

>>1055630

the DSOxxx scopes are toys. a used 20Mhz CRO is infinitely better, as long as everything still works, and the knobs and switches arent all noisy and intermittent.

>>1055630
I wouldn't use a scope for looking at radio. Get a RTL-SDR dongle and make recordings that you can view in Matlab or Python notebook.

I've got a question for any HAMs or RF wizards that might be lurking...

So, I've got an First-Person-View headset for flying quadcopters and planes, and I want to get better reception out of the receiver for flying through obstacle-ridden areas. Having seen the results that can be achieved (https://www.youtube.com/watch?v=DeSuXK1noDQ)), I think I want to try and add an LNA preamp just upstream of the receiver module itself, but I'd like it to be small and integrated into the headset if possible (or the antenna, I suppose).

So, what do I need to know for designing a clean, compact LNA pre-amp circuit? Can someone point me in the right direction?

>>1053280
Not locking, but XT60s are nice and have largely replaced Deans Ultra and Tamiya connectors in the hobby. How much current do you need to carry?

>>1056018
You can buy MMIC amplifier chips that just slap down onto a PCB for which you've calculated the proper stripline impedance, but without pricey test equipment there's no way to debug any problems that come up. And all sorts of weird things can happen at 6 GHz, even to experienced RF guys.

The thing you posted is reasonably priced. About the same as an RF design textbook.

>>1056029
>without pricey test equipment there's no way to debug any problems that come up. And all sorts of weird things can happen at 6 GHz, even to experienced RF guys.
I was kinda afraid of that. Ah well.
>The thing you posted is reasonably priced.
Perhaps, but it's a bit too clunky for what I'd prefer to do. I was hoping something significantly smaller could be built or purchased, so I could integrate it into my headset directly.

One of my RC receivers (a Spektrum AR610) has this itty bitty PCB towards the end of the longer antenna feed which, I assume, is an LNA (the receiver itself is advertised as having an "amplified antenna"). Kinda made me hopeful that something small (perhaps not THAT small, but small) could be integrated directly into my headset or antenna.

Hello /ohm/

My friend gave me 2 50 Watt speakers. I already ordered a 2x50W amp on ebay, though it has no volume control or such. I want to put a female minijack as input, so I can either use an aux cord with a phone or... the thing I want ot ask about. I have old Bluetooth headphones, can I use their speaker output with amp input? The input has common ground and the headphones do not so I'll add a 32 ohm resistor to the empty speaker spot I won't be using, and the other one goes to a male jack to be connected to the amp. Will this blow up in my face? I don't listen to loud music so I'll keep the headphone volume at a minimum.

Any tips for building a microcontroller output amplifier for switching higher power

I've tried and failed at this probably 3 times before with different types of transistors

I think my designs have been too simple and half assed for such a big gain

>>1056141
What voltage? How much current? Push-pull or one-sided? What switching frequency?

High-current BJTs have low gain, so you need to use Darlingtons. High-current FETs have high gate capacitance so you need high-current gate drivers if you need to switch them fast. Also, "logic-level" FETs don't necessarily saturate if the gate is driven from 5V; if you're switching a lot of current you need to check the actual Rds at that voltage.

>>1055593
>Just looking at the schematic again after months and what they actually have is 2 toroidals.
One has 2 12V outputs and they use one of these outputs to drive shit like LEDs and relays.
Then they have the 2 37V and 2 15V connected in series with the remaining 12V from the first transformer to get 116V with is then rectified to make a 310V HT rail.

I've just gone and ordered a toroidal with 2 55V outputs and will connect them in series to make 110V and leave the 12V windings from the other transformer unused.

The small difference shouldn't have too much of an effect should it?

>>1051428
Did you make those rails and faces or did you get that box from somewhere?

Just got an arduino and I'm fucking around with it.
I'm basically just trying to use a shift register (74HC595N) to control a common anode 7 segment display with a decimal point.

So I've basically connected the 2 Vcc pins on the display to the 5V rail and the segments are then wired to the shift registers outputs through 120 ohm resistors.

I've segment A wired to output 0, B to 1, C to 2, etc.

I've pins 8,10 and 13 of the shift register connected to ground and 16 to the 5V rail.

Clock, latch and serial are connected to the arduino.

As the display is common anode, I assume I have to set the output pin on the shift register to low if I want it to light and high if I don't.

My code basically has an array of 10 bytes. Each byte represents a number to display in the format ABCDEFGDp.
I always want the decimal point to be on so for instance the byte to display the number 8 is 00000000 (every segment and decimal lit up).

The program is basically just a loop that sets the latch, shifts out the nth entry in the array of bytes and then closes the latch pin.

When I compile and upload it though, all the segments come on and just stay on permanently.


Code is below


const int latchPin = 10;

const int clockPin = 9;

const int dataPin = 11;
byte symbols[]={10011110,00100100,00001100,10011000,01001000,11000000,00011110,00000000,00011000,00000010};
int counter=0;

void setup() {

pinMode(latchPin, OUTPUT);
pinMode(dataPin, OUTPUT);
pinMode(clockPin, OUTPUT);

}

void loop() {
do{

digitalWrite(latchPin, LOW);


if(counter=9){
counter=0;
}
else{
counter++;
}


shiftOut(dataPin, clockPin, MSBFIRST, symbols[counter]);


digitalWrite(latchPin, HIGH);
delay(1000);
}while(true);
}

>>1056208
If nothing else,
>byte symbols[]={10011110,...
This reads ten million and something, not binary 10011110. Prefix them with 0b, like 0b10011110.
>if(counter=9){
This is always true. Replace with counter==9 for comparison.

Also, when your shit does not work and you don't know why, removing/commenting out the unnecessary shit usually helps. In this case, loops and delays aren't needed to see whether the display part works.

>>1056223
Oh, and pin 10 on 74hc595 is reset, active low. Tie it high instead.

>>1056223
>Oh, and pin 10 on 74hc595 is reset, active low. Tie it high instead.
Cheers.
I just assumed that a byte data type could take a string of bits as data

>>1055331
I love Adafruit. Sure some stuff is cheaper but I'm usually there for unique stuff not the over priced packs of leds. If its not a board they design themselves you should shop around but I make sure to check their page weekly for new stuff.

Also they do a 10% off code every Thursday for their live stream that they post on their twitter.

>>1056060
I'm sure you could make it that small. You're just looking at a substantial amount of learning, and then having to beg/borrow/steal a spectrum analyzer and signal generator that work at a few GHz. It may take a while.

>>1056237
Consider: how does the compiler know you want binary as opposed to decimal?

>>1056261
I thought that if I defined it as a byte (like as in 8 binary digits) instead of an int that it would be acceptable.
Also, what's the difference between upload and upload using programmer in the Arduino IDE?

Is it possible to write my programs in notepad, compile them and transfer to the Arduino without the IDE? I've always hated using IDEs

>>1056223
Have it working now but had to do a bunch of if else statements as the array wouldn't work either as a byte array or integer array (I had elements prefixed with 0b in both).
Also it only works properly if I shift out the byte least significant bit first.
Does this mean the first bit I shift out goes to Q7 on the shift register?

I might have the ability to grab an old (possibly not working) tektronix 7704a scope with a logic analyzer.

Or I could get like a 3 channel 500mhz scope slightly easier, it would be a bit newer, from the 80s instead of the 70s but no logic analyzer.

Should I stick to trying to get the 7704a?

>>1056353
>tektronix 7704a
grab the 3 channel 500 newer scope

>>1056362
*checks more*
It's a tektronix 475, and I lied, seems to only be 100mhz..

Going through ancient storage rooms at work and boss said I could probably pick one as long as it's from an area he manages.

I'm not even into /ohm yet, but I want to hack the a new display into my car's pop up dash and hack bluetooth audio into the navigatoin system's fiber optic system..

>>1056341
> Does this mean the first bit I shift out goes to Q7 on the shift register?
After 8 bits have been shifted out, the first will be on Q7, the last on Q0.

>>1056353
>>1056372
If you've got room for the 7704A and the ability to move it around, it probably has better specs vs the 475. However it's a real boat anchor. I have a 475 myself but haven't used it in years.

>>1056207
lifted the rails and the bare aluminium plates from an old ship system at my old job.
rails are available on ebay and all sorts of synth websites.
google eurorack

the black plates i got for cheap from GMSN.
at 1£ each because they where a bit to shitty for their other products.
dunno if they still have them.

the "box" as it is now, is just a temporary solution.
i got 3 sets of rails total.
I'm planing on building a console style case with two 1U blinds between the rails for utility stuff.

>>1056208
good way to start.
made this display many years ago in the same way, before i knew much about electronics.

the SIPO's i used at the time wast latching, but they way i code it updates it in such a way that i dont get any flimmering.
they fact that they are incandecant probably helps too.

I found it in an old project bin and turned it into a clock this summer.


next challenge for you would be to get a multiplexed display running with two shift registers
VFD's are cool, and also very cheap, like the IV-21

>>1056155

I need about 24 volts with a couple hundred mA, it's one of those rotating mechanical counter displays kind of like this picture, so it isn't really specific

The frequency is probably less than 5 hertz

I have tried darlingtons and other larger transistors but I don't think my microcontroller outputs enough current or voltage to drive it so it probably needs to be more complex

>>1056644

>>1056644
> I need about 24 volts with a couple hundred mA,
High-side, low-side, or push-pull? I'd assume that a counter doesn't need push-pull. If it can be placed on the high side, then it's trivial: you just need a transistor with enough current and gain. You can probably get a BJT with the right specs, but a Darlington or logic-level FET will definitely suffice.

If the counter has to be on the low side (because it has a common cathode and the mechanism is sensitive to polarity), then you need something like pic related.

>>1056644
Darlington with something like 1k base resistor should be more than enough. So, +24V -> counter -> collector, emitter to ground, base to microcontroller via resistor.
Couple of typical errors: people forget to connect the 24V PSU ground to microcontroller's ground and/or forget the reverse diode (cathode to +24V, 1N4001 for example) in parallel with the counter.

>>1056690
Rather than using an emitter-follower, the circuit shown in pic allows Q2 to be part of an open-collector buffer such as a 74LS07. That halves the number of discrete transistors if you need multiple outputs.

But first check whether you can put the load on the high side; that would allow you to get away with a single Darlington or FET.

>>1056309
Why the hell would you want to avoid the IDE of you want to do the exact same thing it does?

>>1056791
>Why the hell would you want to avoid the IDE
Perhaps because it's a bloated piece of shit that is only just tolerable for making slight tweaks when debugging?
Jesus, they still don't support tabs being entered as tab characters instead of X number of spaces.
(Yes, I know it can read files using tabs just fine, but I'd like to actually be able to add them myself).

>>1056309
>Is it possible to write my programs in notepad
Yes, quite easily. Just make sure you use the .ino extension, and put your main source file into a folder with the same name.

E.g. your main file is called "My Project.ino", so it's full path would be "~/Documents/My Project/My Project.ino".

I'd recommend using notepad++ over notepad, then you can use the C/C++ syntax highlighting. There is a plugin for Arduino but it's not very useful.

>compile them and transfer to the Arduino without the IDE? I've always hated using IDEs
Compiling for Arduino and flashing via command line is extremely painful, just open in the IDE and use that.
Besides, any compiler errors will actually show you where the problem in the code is.

There is actually a setting to make Arduino open all source files as read-only. I have stuff opened in Arduino for compiling/flashing, and notepad++ for actual editing.

I recently got my hands on 2 programmable PLCs made by Mitsubishi, some relays got stuck so that's easy to fix.
But I'd like to build something with it, what's a cool thing I can I incorporate a PLC into?

What are some good (relatively) heavy duty connectors that are pcb/panel mount?

I'm looking to make a kind of "backplane" design for a microcontroller project, but I need several removable modules, similar to the hot-swapable power supplies in server racks.

So module 1 would look like a metal box with a connector on the end that slides into a receptacle. Pic related from my old server rack.

What software do you guys use for simple dc analysis? All I want is the voltages across each resistor, currents, simple stuff like that, LTSpice takes it to a whole new level and its complicated!

>>1057017
I'm not sure if that connector in your pic is an eurocard connector, but they're very similar. They're available in different sizes, with different amounts of pins and with different current ratings.

>>1057093
If it's simple enough, I do it by hand. If not, with PSpice (which is comparable to LTSpice).
You might want to try Falstad's simulator.

you can get old card cages on ebay and such places. search, for example, for ''S-100 card cage'' or ''card rack''. same for STD bus, Multibus, etc. this guy, for example sells this (pic) for $60.

another alternative is RJ21 type 50-pin phone connectors.

you can get old card cages on ebay and such places. search, for example, for ''S-100 card cage'' or ''card rack''. same for STD bus, Multibus, etc. this guy, for example sells this (pic) for $60: http://www.retrotechnology.com/herbs_stuff/cardracks.html

another alternative is RJ21 type 50-pin phone connectors.

>>1057017
you can get old card cages on ebay and such places. search, for example, for ''S-100 card cage'' or ''card rack''. same for STD bus, Multibus, etc. this guy, for example sells this (pic) for $60: http://www.retrotechnology.com/herbs_stuff/cardracks.html

another alternative is RJ21 type 50-pin phone connectors.

>>1053280
salvage molex connectors from computers?
what sort of amperage are we talking here?

>>1057093
> What software do you guys use for simple dc analysis?
How simple?

If it's simple, by hand (outside of circuit theory exercises, you don't get resistor networks which can't be decomposed by successive applications of the serial/parallel formulae). Otherwise, with whatever version of spice is to hand.

Bit confused with my Arduino.
So I had a common anode 7 segment display and a 74hc595 shift register on a breadboard with the shift register outputs wired into the display segments via current limiting resistors.

I had 5V and ground rails on the board with the anode of the display and the VCC and reset pins of the shift register wired into 5V and the ground pin and enable pin of the shift register wired to ground.

I was powering the rails from the Arduino's 5V and ground pins.

Anyway it was working fine, I just had a program for it to cycle through 0-9.

So today, I found an old LCD screen and wanted to check if the back light was working.
So I unplugged the Arduino's 5V and ground pins from the breadboard and connected them to the LCD.

When I powered up the Arduino, the 7 segment display lit up and started counting.

The only connections from the Arduino to the breadboard are the data, clock and latch pins for the shift register.

When I check the voltage across the rails that were powerd by the Arduino, they're flopping between 1.8 and 2.5V.

Where exactly is this power coming from?

>>1057332
Sounds like it's powered via 74HC595's input protection circuits. Pic related.

So I really like hiking and sometimes on longer hikes I need to worry about how to charge my cellphone and stuff. I was thinking of making a piezoelectric generator and strapping it to my shoes. Realistically, hiking 8-15 miles a day could this charge a cellphone?

>>1057357
Typical shoe piezo generator output powers are measured in milliwatts (or millijoules per step). In comparison, cell phone batteries have tens of kilojoules of energy.

>>1057357
No
Also you are fucking stupid.
Get a phone with a battery lasting longer than an hour or get one of those power packs that dumps it's charge onto the phone.
If the Chinese haven't built one then you know it must be completely garbage rather than just a bit crap

>>1057371
http://www.memsjournal.com/2010/04/microstructured-piezoelectric-shoe-power-generator-outperforms-batteries.html

>2 mW per shoe during a regular walk
>2000 steps in a mile
>4 watts per mile
>average 10 miles a day, 40 watts a day
>iPhone 6 consumed 10.5 watt-hours to charge
>enough juice to charge cellphone 4 times from 1 day of walking

Is this right? This doesn't seem right. I must have messed something up.

>>1057357
Buy a power tool battery and USB adaptor for it.

>>1057385
>>1057389
I'm currently using powerbrick. But on long hikes (longer than a week) I have to ration my phone usage which means I can only use it for maps. If I could charge as I walk I could listen to music all day. Plus powerbricks are heavy, this might be lighter.

>>1057394
>But on long hikes (longer than a week)
Then you need to step up and get solar.

>>1057397
Thought of that. Those are not as good as you'd think. Gram per energy wise you're better off just bringing a 2 or 3 large batteries.

>>1057387
> Is this right?
Not even remotely. Learn the difference between energy (Joules = Watt-seconds) and power (Watts = Joules per second).

What it says in the article is 5 mJ per step, 2 mW per shoe (4 mW total) average power (after conversion losses).

4 mW = 250 seconds per Joule (Watt-second) = 250 hours per Watt-hour. So 250*10.5 = 2565 hours = ~107 days = ~3.5 months for 10.5 Watt-hours.

Anyone have experience with those cheap chinese bench supplies from ebay/aliexpress/etc? Seems like fun thing to have around without spending too much, but I'm a bit cautious about how long they last.

Is there a particular type of glove I should wear when disassembling old electronics? Just curious what other people might recommend.

>>1057399

...the hell are you getting your numbers? And for what locale? A panel the size depicted in that image should produce at least 10W in good sunlight. I've got some bare LiFePO4 cells lying around that each hold 8.3Wh, and there's no way that panel weighs more than 4-5 of them. The panel wins after lying in the sun for just a few hours in a single day.

Even if, for some reason, you never actually put it in direct sunlight and only got 10% of the rated output, it would still, at worst, break even with the batteries over a hike that long. You'd have to be almost completely without sunlight the entire time in order for the batteries to be competitive over that span of time.

>>1057125
>>1057149

Thanks m8s. I'm looking into some of them, but I fear that my power requirements are too much (100 or so amps), so I think I'm going to redesign.

At a minimum I know I can get it down to 30a (@12v) per pin, so I think this might be the way to go.

>>1057646
Eh, you don't want to do 30A/pin on a backplane. Bad contacts will kill you pretty quick. At work we have Metral power connectors that are fookn expensive and they only do 3A per pin/tab thing.

>>1057652

Well shit, I'll just find a better way then, or redesign the whole damn thing to be lower power for the removable modules.

If I remove the fuse box being a module then I can get it down to less than 1/4 amp per pin.

>>1057408
i got a 2x 0-30v 5A digital chinapower from farnell.
works fine.
being able to adjust on a milliamp level is very useful.

the inrush current is quite huge when turing it one though.
i have electronics bench on a 6amp automatic fuse.
its tripped that fuse quite many times, with only soldering irons on the same line.

>>1057658

Irons could draw a bit of current, but still.

How much inrush current are we talking?

Something that could be negated with some decent capacitors?

>>1057660
>Irons could draw a bit of current, but still.
not really. an idling iron doesnt really pull much at all. especially not on the mains side.

>Something that could be negated with some decent capacitors?
how on earth do you figure that?
its the huge capacitor bank in the powersupply that is creating the inrush in the first place.

i suppose you could put a huge inductor to limit the inrush, or build a zerocrossing start with a solid state relay.

>>1057646
Eurocard connectors are available up to 40A per contact and then there are connectors specifically meant for "backplane-like" power distribution, but if you can avoid huge currents in backplane simply by placing the fuses elsewhere, then that sounds like a much better option.
Also, live insertion/removal requires a bit of thinking at 100A, or you just destroy your connectors.

Can anyone recommend me an entry level bench power supply?

Is eBay a good place to pick one up? Or should I stick to a normal shop?

>>1057670
i looked at ebay first. but living in norway made it as expensive as ordering new on farnell due to shipping costs.

it might be different for you, might even find one close that you can pick up.

>>1057674
Price wise I'll just look for the best value. What I'm really asking is whether or not eBay can be trusted for powers supplies, or if they're generally shitboxes that people are trying to get rid of.

>>1057688
well if its analog, it should be simple to fix^^
cant really comment on if people on ebay in your area are trustworthy or not..

>TFW you ordered your hakko fx888d

Brehs it's my first big boy soldering iron....I.e. not 7.99 from China.

>>1057699
>hakko fx888d
looks like a fucking toy, but i've heard they are good.

How does the pic related circuit works and what is it for?

>>1057754
It looks like some trash you threw together to troll /ohm/.

>>1057746
Who gives a fuck what it looks like. I'm soldering with it, not fucking it.

I'd want some cheap (hopefully <$10) ARM devboards, anyone got good suggestions? Need solid foss toolchain support, other than that pretty basics (few pwm channels, handful of gpio, one or two adc channels, i2c, uart, etcetc)

Teensys look nice, but $20/board +$20 for shipping is bit salty for me

>>1057834
I'm asking no- ironically - have you considered doing one yourself?

>>1057841
answer to that question is yes. At this point I don't a custom board would be worthwhile. I might roll a board once my design is finalized, but now I don't want to bother with that and I don't think attempting to compete on price with mass-produced items makes sense.

How would you reorder this in order to remove remove the crossing in the middle?

>>1058105
I think this it, also the loops for voltage/current.

>>1058105
I think this it, also the loops for voltage.

Sup /ohm/ I see no ham radio thread so I thought I'd ask here. Want to get into SDR, is this the best choice for begginers?

>>1057754
>How does the pic related circuit works
>pin 4 not connected
It doesn't, it doesn't makes much sense either way.

>>1057834
There are some ST Discovery boards around that price. gcc compiler and st-flash support, but the oss libraries will do basic stuff. There's also supposed to be a BSD-licensed HAL from ST but I haven't looked at it.

>>1058284
one that is cheap, but not too cheap.
seriously, there must be a SDR forum around.
> google first

>>1057754
>>1057763
>>1058289
>trash you threw together to troll /ohm/.
>pin 4 not connected
I know, right? I found it in electronics magazine as a riddle to solve, but I have no idea what it might be. I'm not even entirely sure what that symbol on the far right is - some sort of a meter?

>>1058312
Yeah, most likely a meter.
Question "what it is" boils down to "what is CX" and "what changes the value of CX".
What comes to the other anon's comment about the reset pin: IIRC bipolar 555 works even if you leave it unconnected.

I'd like to build an electrically-assisted bicycle. What kind of motor should I use? I've been looking for cheap DC motors in 200..250 W range.

>>1058300
Somehow my post ended being deleted, but yeah, I've been googleing for a while and that seems to be the best "bang per buck", enough for what I want to do.

>>1058312
>I'm not even entirely sure what that symbol on the far right is
It's a gender.
I myself identify as a MOSFET

Can anyone tell me where emf interference would be coming from in this setup?

Top left is 5v PSU powering L298N motor controller
Bottom left is Arduino nano setup
Middle is L298N with 6 data wires going to arduino, as well as 1 common ground line going to arduino
Right side is the two motors

>>1058455
pic of the arduino pinout

>>1058456
Forgot to mention, the issue is that when the motors are firing, my button gets triggered and overridden, meaning that whenever the motors fire the button automatically clicks then becomes unresponsive until the motors are turned off.

>>1058457
Probably a software bug, double-check your ifs

>>1058465

Don't think so, it's very basic and straight forward right now

const int buttonPin = 2;
const int rotorMotor1 = 7;
const int rotorMotor2 = 8;
const int rotorPwm = 9;
const int vibMotor1 = 10;
const int vibMotor2 = 11;
const int vibPwm = 12;

volatile int count = 0;
volatile int setPwm = 200;

void setup() {
pinMode(buttonPin, INPUT_PULLUP);
pinMode(rotorMotor1, OUTPUT);
pinMode(rotorMotor2, OUTPUT);
pinMode(rotorPwm, OUTPUT);
pinMode(vibMotor1, OUTPUT);
pinMode(vibMotor2, OUTPUT);
pinMode(vibPwm, OUTPUT);

attachInterrupt(digitalPinToInterrupt(buttonPin), cycle, FALLING);

Serial.begin(9600);
}

void loop() {
if (count==0) {
Serial.println("0");
analogWrite(rotorPwm, 0);
analogWrite(vibPwm, 0);
delay(2000);
}

else if (count==1) {
Serial.println("1");
digitalWrite(rotorMotor1, HIGH);
digitalWrite(rotorMotor2, LOW);
analogWrite(rotorPwm, setPwm);
digitalWrite(vibMotor1, HIGH);
digitalWrite(vibMotor2, LOW);
analogWrite(vibPwm, setPwm);
delay(2000);
digitalWrite(rotorMotor1, LOW);
digitalWrite(rotorMotor2, HIGH);
digitalWrite(vibMotor1, HIGH);
digitalWrite(vibMotor2, LOW);
delay(2000);
}
else if (count==2) {
Serial.println("2");
digitalWrite(rotorMotor1, HIGH);
digitalWrite(rotorMotor2, LOW);
analogWrite(rotorPwm, setPwm);
digitalWrite(vibMotor1, HIGH);
digitalWrite(vibMotor2, LOW);
analogWrite(vibPwm, setPwm);
delay(2000);
}
else if (count==3) {
Serial.println("3");
analogWrite(rotorPwm, 0);
analogWrite(vibPwm, 0);
delay(2000);
}
else {
Serial.println("resetting count");
count = 0;
}
}

void cycle() {
static unsigned long last_interrupt_time = 0;
unsigned long interrupt_time = millis();
if (interrupt_time - last_interrupt_time > 500)
{
Serial.println("incrementing cycle");
count = count + 1;
}
last_interrupt_time = interrupt_time;
}

>>1058455
>that unbalanced counterweight on the rotor
are you building an assvibrator, anon?

>>1058470
Mabes. Now help me

>>1058455
> Can anyone tell me where emf interference would be coming from in this setup?
From pretty much everywhere. But the wires from PSU+->controller->motor->controller->PSU- would be the prime suspects, as the current will be many orders of magnitude higher than the wires carrying the power for the uC or control signals.

Also, ensure that the power rails are adequately decoupled.

>>1058488

Can I throw a capacitor on it or something to absorb excess charge?

>>1058509

Also, should I buy a $22 oscilloscope from amazon? that would at least let me trace the surge of current right

>>1058509
You need to bear in mind that wires start to look like a resistor and inductor in series at high frequencies, i.e. the topology of the power network matters.

Split the power rails into uC and motor branches early on (i.e. the power to the motor controller shouldn't be running through the breadboard's rails), place a capacitor across the breadboard's rails close to the uC, place another capacitor across the power wires near the controller board. Use low-ESR caps (e.g. tantalum).

Also, form the power wires to the controller and between the controller and motor into twisted pairs (total flux is proportional to the area of the loop, so keeping the wires close minimises it).

>>1058515
Uh, I can't imagine it would be much good for $22. For looking at transients, you need decent bandwidth, much higher than the switching frequency.

>>1058563

Thanks I'll try to clean it up when I'm back home. Also the oscilloscope would be this well reviewed kit, still nah?

https://www.amazon.com/gp/aw/d/B00WAQGGZA/ref=mp_s_a_1_2?ie=UTF8&qid=1474494367&sr=8-2&pi=SX200_QL40&keywords=oscilloscope+kit&dpPl=1&dpID=510KVyDdWKL&ref=plSrch

>>1058573
>Analog bandwidth: 0 - 200KHz

What can I use for signal-level ground isolation for audio? A simple DC choke where the input is connected to one coil and the output to the other coil will suffice?

>>1058606

Additional info - a shitty 6 dollar amp and a bluetooth headset as input, both powered by an ATX power supply.

>>1058606
Audio transformer.

I know that you can get them with 1/4" jacks or XLR connectors, I don't know about 2.5/3.5mm jacks.

>>1058615

Connectors are a trivial matter. I know about the transformers you speak of, but I wonder why a DC choke won't work since it has a 1:1 turn ratio.

>>1058618
It may work, but it's not designed to be used as an audio transformer.

An audio transformer is designed to have a flat frequency response over the audio range.

A common-mode choke doesn't really care about its frequency response when used as a transformer. A "DC choke" may even have intentional leakage inductance to block differential-mode ripple; that would tend to make it act as a low-pass filter if used as a transformer.

>>1058650

I suppose I'll just have to wait until the amp arrives and see. I can always upgrade the bluetooth headset battery - 400mAh now and it lasted a day when it was powering 2 32 Ohm speakers. So that should be acceptable. Thank you for your responses!

Anyone know where I can get a decent scrap motor for a project, something around 1/3 hp

>>1058869

old trash compactor? old vacuum? old sink garbage disposal?

>>1058573
I've got it, it's not bad and it could be pretty helpful here
Worst case you get a basic oscope to start

Roll for project ideas, friends

>>1058980
Fuck that's the wrong pic

Does anyone have the electronics one

>>1058982
Nvm i have it

roll

hey /ohm/, i recently dismantled an old color TV set and pic related is pile of components i got
the problem is they are dirty as fuck and i want to first clean them somehow before putting them in the parts box, some are so dirty i cant even read the markings
can i just wash them with water&soap? will old electrolytic caps and inductors/potentiometers be damaged by this? i also have this one component which is basically two inductors packed as one (4 pins) which has on its bottom some sticky, bitumen-like matter (it was stuck to a board, and the shit is actually liquid, when i rip it apart after a while it flows down flat). Can i remove this and is washing all this with water, soap and say an old toothbrush safe?
any help would be much appreciated

>>1058985
roll

>>1058990
reroll

>>1058989
> pic related is pile of components i got
Check whether any of those TO-3 transistors are useful. The rest can go into the trash.

>>1058989
protip: don't reuse parts unless they are worth it or you know what you're doing. When you're making some circuit you want to make sure you minimize the amount of things that can go wrong with it, or else debugging it is going to be hell. Using some subpar parts that can be bought cheap and easily is totally not worth it.

>>1058997
thanks for the advice, but could you answer my question? TO-3s should be cleaned as well, but can i do it with water?

>>1058985
i'll play

>>1059002
Yes, you can wash TO-3s, but the thermal compound is not water soluble. Dunno if soap has any effect. Rags/tissues work and you can try alcohol as well.

>>1059025
i know, ill just wipe it with some toilet paper, water is for the dust
thanks for answering

>>1058989
maybe keep the trafos too? those tend to be relatively costly components that can be useful

>>1059035
ok, but what about washing them? can i use water on them?

>>1059013
>stepper driver
ok, after finding out what this thing actually is and seeing as how there's a myriad of different off-the shelf driver modules and chips for driving stepper motors i decided to build one using a ghetto /diy/ way: NE555 and 1N4007 with 2N3904 transistors and some diodes for the inductive 'spikes' protection (a thing i don't fully understand yet since i have no exp with motors, but i trust the seasoned veterans lol). and now i have to get a stepper motor from somewhere.

first draft version probably coming up in next /ohm/ edition

i understand old printers, floppy and hard drives have stepper motors inside.. right?

>>1059044
Only very old HDDs (30MB or so and smaller) had stepper motors.
Only very old floppy drives (5.25" and bigger) had decent stepper motors. "Modern" FDDs have very small and very shitty motors. Try to use the whole assembly with ball screw, bearings, etc. instead of taking just the motor.
Dunno about printers. Apparently some have them, but the few I've taken apart had normal motors.

>>1059042
> some diodes for the inductive 'spikes' protection (a thing i don't fully understand yet since i have no exp with motors
The current through a coil cannot change instantly. If you try to switch it off, it will keep flowing through the switch. With a mechanical switch, this equates to a spark across the air gap as the contacts separate. With a transistor, it equates to a dead transistor.

Adding a diode allows the coil to "pull" current through the diode when it can no longer get it through the transistor. Note that this means that the coil remains energised for a while after the switch is closed, which is an issue if you're stepping quickly.

An alternative is a snubber network involving a capacitor, a resistor and (optionally) a diode. This allows a moderately-high negative voltage to develop which reduces the current more quickly than allowing it to circulate through a diode, but not so high as to damage the transistor.

>>1058985

Fuck it, let's do it

>>1057754
ESR meter? By applying a frequency to Cx and measuring voltage across it with an ammeter.

I could be wrong, though.

>>1057834
Cypress PSoC boards are cheap but the ide is Windows only. You should really be OS agnostic though if you want to get anything done.

Look up the cy8ckit-059. Up to 80Mhz cortex m3, lots of peripherals, 4 built in opamps. Internal digital and analog routing matrices. My favorite feature is its UDB capability. There are 24 "UDBs", each one of which has an 8 bit ALU and two PLDs. It's nice if you're doing state machines because you can completely offload it to hardware. Also, many of the peripherals can be implemented with the UDBs (the ide gives you a config GUI). For example, if you need 7 timers instead of the two built in ones you can *make them*. Pretty dope and super flexible.

>>1059453
seems like it:
http://danyk.cz/esr_en.html
mystery solved, nice job!

So a while back I made myself a VR headset with a little Arduino for tracking and a Raspberry Pi screen. It was only 1280x800 but it worked alright. Then the screen died. Anyone know what a good replacement would be? Something between 5 and 10 inches with a decent resolution that doesn't cost an arm and a leg.

So I have a 12v DC motor running off 5v right now. I'm controlling the speed with PWM but the rate of the motor doesn't seem to change much. Is this expected because I'm under-powering the motor? Also, currently on 5v the motor stops when I get down to around 110 (out of 255) for PWM. Will I be able to use the full range (0 - 255) when the motor has full 12v power?

>>1058909
purchased. thanks m9

>>1059565
Are you driving the motor directly from your Arduino by chance?

>>1059601
no got it running off an l298n, which itself is getting power from a separate 5v psu

>>1059608
Try driving it through a transistor and see if anything changes

>>1058985
test

>>1059612

the l298n is h-bridged transistors

Is there anywhere that sells resistor assortments in smaller value ranges? (me=USA location)

Like for example:
most of the stuff I buy, I get of aliexpress.
the resistor assortment packs are not real good quality, but they work. Problem is tho that the smallest resistor is 2.2K, and then they go up to 10meg or whatever.

I also have a sparkfun resistor assortment that is nice, it starts down in the single-digit ohm values, but still goes up to the megs. I don't need any more high-value resistors.

what I'd like to find is a 1/4-watt and 1/2-watt assortments that ranges from 1 ohms up to around 1K or 2k ohms at the MOST.... even if I gotta pay first-world / sparkfun / digikey / newark prices it'd be okay. ???

>>1056849
Home security system

>>1059749
Just get any pack that has the low values you need and keep the high values, never have enough

>need an HF ATU for amateur radio
>see an ex military automatic one at a car boot sale today
>bloke knows nothing about it
>hmmm a bit and decide not to get it
>find out it's for 30-76mhz so would be a bit useless anyway

Dodged a bullet but I dunno maybe I should have bought it cheap anyway.

Pic related is one.

Any Germans here?

>>1059903

how much was the bing bong going for

>>1059903
Just an FIY military radio stuff is generally useless to civvies. They have their own frequency bands and they generally don't overlap much. Their antennas are decent, though. The old OE-254 mast antenna, while a pain in the ass to setup, gets great reception and can last for decades without maintenance.

Rate my PCB

>>1060075
I didn't ask desu
>>1060078
Yeah I don't know anything about military radio really, is it channelised?

Those antennas look decent though.

>>1057399
TEC.
if you make a fire once in a while.

>>1060244
there is just so much i could bitch about here in terms of layout, component placement, traces, vias, uhghhh just everything seems so wrong. does it work? reliably?

>>1060318

tell us more senpai

When doing Mesh analysis in this circuit, I am curios as to how to consider the current in the red part of this circuit.

I take that from blue point current goes down on both branches, then how does it behave when it reaches the one in between (red)?

>>1060318
>there is just so much i could bitch about here in terms of layout, component placement, traces, vias, uhghhh just everything seems so wrong.

Really? Can you give some concrete examples, I'd be grateful

>does it work? reliably?

I've only received the boards a few days ago I don't know about long term reliability but they've worked great at all testing so far

>>1060394
The current only flows in one direction obviously, either from a to b or opposite, or maybe no current at all. Just pick a direction at random and use it consistently for all meshes. Doesn't matter if you guessed the wrong direction, then the equation results will just show a negative current

>>1060394
Given the 20V source, it's a fairly safe bet that the current through R4 will be from a to b.

OTOH, it's far from a safe bet that those green and blue arrows have the correct direction; specifically, the ones on the top right and bottom left, as the 20V source will try to make the current through R2/3/5/6 from right to left.

But as >>1060405 says, none of this affects the calculation. Choosing the opposite direction just means that the calculated value will have the opposite sign.

>>1060405
>>1060458

Thanks anons, this stuff confuses me a lot and I just returned to it.

This one been bothering me for a couple of days. If I wanted to solve it by using Nodes method.

I am wondering how is it going to behave right there in the middle (purple) is there going to be a V3 there?

I know that I can get a Equivalent resistance in that part, but I am confused as to how it is going to look in the end and who is going to be affect the middle branch (V1 or V2)

>>1060458
>>1060405

Thanks anons, this stuff confuses me a lot and I just returned to it.

This one been bothering me for a couple of days. If I wanted to solve it by using Nodes method.

I am wondering how is it going to behave right there in the middle (purple) is there going to be a V3 there?

I know that I can get a Equivalent resistance in that part, but I am confused as to how it is going to look in the end and who is going to be affect the middle branch (V1 or V2)

>>1060471
Yes the purple node will have its own voltage. Both V1 and V2 will likely affect it.

>>1060400
1. layout, placement

you placed the crystal closer to i/o connectors than the ic it's supposed to clock, you didn't reserve a ground plane around the crystal with vias littered all around to keep coupling to a minimum (can't even tell what the return path is here), you turned it into 45 deg angle for literally what purpose? there looks like enough board space to place it in a manner that it is not interfered with. your manufacturer was probably like "the fuck son? but ok, you payed the money"
you placed the decoupling caps and vias to gnd so far from the pins that they are effectively useless. you used small enough packages to make them fit a lot tighter and functional, why didn't you make use of that? if it works, you don't even need them, if it fails.. replace, reroute, reorder the pcb

2. traces

your vss/vcc traces are too thin. bulk them the fuck up to provide good juice to your chips. use regions of board space to run polygon spaces to power pins etc.
you're using meanders for what reason? there doesn't seem to be any digital signal lines that would need meanders for fine-tuning the signal times. the lengths are not long enough to justify wasting board space borrowed from better grounding via more ground plane around your signals etc.
you're running traces in a way that creates a lot of useless space between them. this adds to islands of copper that need to get etched and adds to production time/cost. if they are not high speed signals, tighten them up, they wont interfere with eachother.

>>1060400
3. vias

there's not enough of them. your ground return paths are way too long and fucked up in all sorts of ways (in some cases i don't even see a return path) and your whole board is a gigantic multi-antenna for all sorts of funky signals coupling in. i'm seeing a lot of them placed in ways that disable a better placement for others too... what the fuck is up with that???
they are too big. your manufacturer has minimal drill specs. go as minimal as you can and place them as close as you can to you components and keep the traces minimal. especially those that connect to active components and to gnd.

4. random

why didn't you include mounting holes? do you expect the connectors to last when you connect/disconnect over 2 times? it's mechanically shit. always mount your board with at least two holes (since you have i/o at both sides 4 mounting holes).
i'm not seeing any diodes to protect from static discharge... might be a good idea to protect your ics from people who like to, before touching your device, dance around on a carpet or so on..

5.. dunno i think my post might already be too long and i'm a bit drunk so i'm gonna ask you if you received my anonymous email?

>>1060400
4. random

why do you need such a bit ic anyway? by the looks of it you're only using a couple of i/o pins there... wouldn't it be cheaper to use an ic that is what.. 4th the size?
was the firmware too big? maybe an external memory would help to get the space factor down by 2-4?

>>1060549
Thanks for the thorough feedback. Comments:
>crystal
The usb connector is shielded so I didn't think it would matter too much if I placed the crystal next to it, and also it was the only way if the mcu is oriented the way it is, and that was the logical orientation. If I turned it other critical traces like the usb data traces would have to be much longer and clunkier. I turned it 45 degrees because that minimized the distance from the crystal to the mcu, a distance which was already dangerously high. Datasheet says max 12mm recommended, there was no way to achieve that with a straight crystal. Also at just 16Mhz I don't think things like separate ground planes and guard rings are really necessary. I could have done it, but it would have just further required space in an already cramped area.

>decoupling caps
Couldn't really have been placed any closer. They are all already just 0.1mm away from each other's keep out zones. Maybe I could have ignored the keep out zones but I don't think there was a need. I should have maybe included some object in the image for scale reference, they really aren't placed far away from the IC at all. I think maybe you think the PCB is bigger than it really is. Most other PCBs I've ever inspected, including like PC components that run at a hundred times the speed of my product uses pretty much the same distance from the pins. And the gnd vias are also places as close to the relevant gnd pin as I couldn't, but fab design requirements made it impossible to get them optimally close without crashing with other traces or pads.

>your vss/vcc traces are too thin.
Here I disagree completely, they are actually more than thick enough. I did the calculations and the temperature in the traces won't even rise 0.1 degrees even at continuous max power consumption. I could probably have made them the same thickness as the data traces even and it would have been fine.

>>1060549
>meanders
The meanders are for a ~20MHz SPI signal. Sure, it would have probably been fine without, but I actually added those in at the end of the design because I happened to have that space there unused anyway.

>vias
I tried to place the vias as close as possible to all gnd pins and let the ground plane handle the return paths. There are few and short traces on the other layer so the ground plane is clear and unobstructed to have an easy return path for everything that way.. I've taken care to not let noisy ground paths pass over critical ones so I think it should be fine.. I also don't really see how I could have placed the existing vias better. I tried really hard to get them as close to the source as possible but the design requirements make it hard, most of them are really squeezed in there. Yes in retrospect I could have made them a little bit smaller indeed but just like 0.2mm smaller at most so it wouldn't have made a huge difference really. And vias are some of the things that could most easily fail with a pcb production, so a larger size decreases the chance of faulty boards.

>mounting holes
The PCB is designed for a specific enclosure. It is mounted on slots on the side where the white silkscreen lines are. There are no place in that enclosure to use mounting holes for it anyway. The side panels are closed with two screws each to lock everything in place tightly. I think it is mechanically sound.

>protect from static discharge.
The usb port is indeed protected with an IP4220CZ6. The DB25 port is not, but all the ICs it is directly connected to are relatively robust, with some built-in protection verified in the datasheets. True that I could have explicitly protected all the pins like on the USB connector too, and I considered it, but it would have increased cost quite a bit and taken a lot of PCB real estate so I decided against it.

>>1060553
>ic size
The microcontroller IC was carefully chosen. Yes, I could not have gone with a smaller one. It contains several built-in modules that I need, like USB, a DAC and a lot of memory. Having all of these in separate ICs would not save space and also be more expensive and take longer to code for. That IC was the best choice.

Also yes I got your email hehe, thanks for the warning but I just assumed that the people who cared about random doxxes are all in /b/, not /diy/ so it's not a problem probably

>>1053899
I'm having some problem with my pepe motor controller. Everything works when tested individually but when ever I try to run it the high side mosfet driver voltage will not float with the motor phase voltage.

As soon as it starts conducting gate to source voltage drops to 0.5v and only 13V will come out of the mosfet regardless of drain voltage. I tried many bootstrap configurations and even an isolated supply directly into the gate drivers high side VDD but nothing changes.

I can't understand where the voltage drop is happening, I have run lights and fans off the 13V output and I know there's more current than the isolated 15V supply powering the gate driver can supply, there's no obvious shorts anywhere, none of the components are heating up.

This is what the gate looks like on a scope before and after it starts conducting, the peak is about 14V.

>>1060725
> I can't understand where the voltage drop is happening
Nor can we without a schematic.

>>1060759
Just a standard half bridge, gate driver is fully isolated across input and high / low sections. I tried isolated power supply in place of the bootstrap with no success.

Ordered a 230V toroidal transformer off ebay.
It arrived today and turns out it has 2 115V primaries that I assume I need to wire in series.
The pick is from the data sheet.
Should I just wire the grey and violet to each other and connect blue and brown to the mains?

>>1060855
> Should I just wire the grey and violet to each other and connect blue and brown to the mains?
Yes.

>>1058952
I'm up to old tricks and I thought I might as well post it here.

experimenting with a single ended power now, trying to learn this time instead of just following a schematic.
this is a 6p14, a russian equivalent of el84.
just as good, just cheap as chips

here I got a 6.66v bias over a 120 cathode resistor, meaning 55mA
at 250v platevoltage, this means its dissipating 13.4W ,which is over spec(12W)
I've run an diy orange ad5 with a similar output stage just as hot before and working fine.
I guess it just won't last as long biased this hot.

I'm going to try to bias it a bit colder tomorrow by increasing the cathode resistor.
lowering the bias current this way also lowers headroom, making distorting the outputstage easier. mite b cool

>>1061457
I made a dummyload to aid testing.
you might spot the variac behind the meter, used for safer testing.
not in frame is a 6amp automatic fusebox that runs the whole bench.
I use this to switch the whole thing on and off, avoing to fiddle too much with it while its running.

any tube guys got any tips for scoping the interstage signals without blowing my scope?

I was thinking a 1/100 resistor divider with zenerdiodes in antiseries camping down the voltage on the scopeside to 10.1v

afaik, my scope is grounded.
making tripplesure that i connect ground to ground should be save enough.
does anyone groundlift their scopes just to be sure?
We did it a few times at my old just as a maritime service tech, but it was never really necessary

>>1061457
hmm.
according to this, I'm within standard operations for the 6p14p with 250 and 55mA(1mA away from max)
I'm struggling to find a datasheet that isn't russian^^
http://www.jacmusic.com/nos/6P14P-EB/

i think I'm also considering plate dissipation wrong, as I'm taking the total current and multiplying by voltage, forgetting the the screengrid is also dissipating some of this.
i'll need to look into this.

I'll look into messing with the bias regardless, i need to get a better understanding.
anyone ever use a pot for adjusting bias? (plastic knob for safety)

also, the glow isnt really as bright as it looks in pick.
i got small 6v3 bulbs on the heater rail to indicate power.

Hey /ohm/, this AC power in is absolutely fucked. I want to replace it, and I can't figure out what it's actually called. It does take the PSU6-UL power cord, but I'm at a loss.

someone make a new thread

>>1061668
>>1061668
>>1061668
>>1061668

new thread