I have a PSU similar to this (this is just a Google image) and I've broken it.
When I was charging a battery with it a while ago, I noticed that once I turned off the PSU, while the battery was connected, the voltage needle indicated the voltage on the battery. Based on that, I assumed that you could also use a PSU as a voltmeter.
Then I tried to do the same thing again some weeks later and, I don't know why, but the wires immediately went super hot, burning my fingers and melting the plastic sheathes.
Now the thing doesn't turn on. Why did this happen? And why this time and not the last time?
Also can it be fixed? What am I likely to have burned out? I've looked inside and can't find any signs of a burned component.
>>901412
It was a 12V battery btw.
Bump
Maybe you connected it backwards. Or maybe you were just unlucky this time.
Assuming the PSU uses normal components: no particular reason why it couldn't be fixed.
>>901418
I'm willing to buy and fit the new component but I don't have a clue where to start. What is it likely to be?
I would guess a fuse first but I couldn't see any broken fuses inside. A transistor then maybe?
Also does anyone know why the amp meter appears to go up to Pi amps?
Are they just using the same scale as they use on some sort of phase meter or something?
>>901416
Don't.
>>901419
Output protection diode (if present) and the pass transistor (bolted on the back heatsink) would be the first things I'd measure after checking if anything looked obviously burned.
>>901423
Thanks. I'll have another look. I'm getting a multimeter for Christmas so I'll probably have to wait til then.
I need that negative rail for doing stuff with op amps.
those things generally have two 2N3055 transistors in the back. it's likely you burnt them up. i have a schematic somewhere i might try to dig up.
> I need that negative rail for doing stuff with op amps.
there's no negative rail. GND is just a direct connection to the 120V ground, as is usually ignored.
found the schematic for a knock-off; it should be similar to this. hopefully, only the reverse polarity diode blew and not the transistors.
>>901440
It has a + - and GND, so isn't that a negative voltage too?
>>901442
My version (I think it's LG - it came from an LG factory I worked at anyway) goes up to... er, as I say, pi amps. No idea why. The voltage goes up to 10piV.
Don't know if that would invalidate this schematic at all.
>>901502
No. +/- are the output pins, GND is protective earth connected to wall plug earth.
>>901505
I'd been using the + and GND pins and it worked fine. I tried using the + and - once out of curiosity and it didn't work.
It was only when I read about op amps that it occurred to me that it might have been for that. Is it not?
I don't really understand.
>>901506
Forget this. I just remembered it's the other way around. I'm a bit drunk now.
In any case, how do I perform simple experiments with op amps when they need a positive and negative supply?
>>901508
You make a split supply with virtual ground.
For instance, you set your PSU to 10 V, and then connect a 7805 to it, using the 5 V output as a ground point for the circuit, making the real ground appear as -5 V in relation to it.
>>901540
I had to Google a 7805 but I think I get the idea.
>>901544
yes
'ground' is arbitrary
if you make -5v your 'ground' and measure from it you get 5v where you used to have ground and 10v where you used to have 5v
this is why using the word 'ground' is fucking retarded because actual ground is outside, the electrical potential of earth.
what you call ground you could call 0v because its not connected to actual earth ground outside.
then it makes a little more sense if you think about whatever you call 0v just as where you put the negative lead of your multimeter.
you could run your circuit using 105v, 110v and 115v relative to outside earth ground but the relative differences in the circuit are still +-5v
>>901548
I see. Could I achieve this dual supply with a voltage divider with 2 equal resistors?
>>901540
While your idea is basically correct, 7805 and the most other regulators won't work, because they can only source current. A shitty kludge would be to add a load resistor to the 5V output so that the resistor provides some current sinking capability.
>>901551
Yes, you could. The downside is that your "ground" will have the impedance of the resistors, so it'll either run really hot (low value resistors) or not be able to handle much current (high values).
For opamp stuff, you usually only use the ground as a signal reference, so there's basically no current flowing through it, so you usually go with a high impedance (10k and up) divider. Because of this, using a 7805 or similar will work, despite >>901554 being true. Your ground current is unlikely to ever overcome the quiescent current of the linear regulator.
(I'm the guy who initially suggested the 7805 thing.)
>>901589
So how does this virtual ground work with being both a source and sink? Or is this what a genuine ground would do anyway in a dual rail scenario? Does it make a difference that the virtual ground actually has a net voltage?
>>901412
you are stupid.
>>901602
You mean I'm a beginner at electronics.
>>901551
in theory
in practice as soon as you add something across one of the resistors you change the circuit and the voltage changes.
the way i did it when i was playing with 741 opamps was with two 9v batteries in series. take the connection between them as your 0v and you get +-9v
you can sometimes do the same thing with a bench supply if you have dual outputs that aren't linked or anything. or if you have two of them.
>>901508
you can make a split-rail supply for OP-AMPS in about 5 minutes using 2 wall-warts. preferably not switching ones (which are very light) but good old heavy iron transformers as used by cordless phones, etc.
>>901608
>>901651
I don't quite get these. The top one gives AC right?
And why has the middle rail been grounded? Won't that turn the virtual ground into a real one?
And why +-11v from 18v?
>>902081
The top one produces DC, but a bridge instead of 2 diodes would work better.
The bottom one might be just a typo, but many (old) wall warts have pretty bad load regulation so that their no-load output voltage is significantly higher than the specified output voltage. How much higher depends on make and model.
>>902092
I thought a transformer was inherently AC.
>>902093
The transformer is REKTified by diodes!
>>902096
And then smoothed by caps to generate an "envelope"?
>>902092
> but a bridge instead of 2 diodes would work better.
requires a center-tapped transformer which is more difficult to find. a typical house will have at least a dozen single-ended transformers.
>bottom one might be just a typo
not a typo. it reflects reality; if you're using low-current stuff like OP-AMPS, the actual voltage will be considerably higher than the sticker voltage.
>And why has the middle rail been grounded? Won't that turn the virtual ground into a real one?
the ground symbol doesnt necessarily refer to earth ground. there's a specific symbol for chassis ground, or local ground, but it's ugly so not used as often.
>>901508
doesn't get much simpler than this desu
>>902154
>requires a center-tapped transformer
... which was used in your two diode schematic as well.
>not a typo. it reflects reality
So you agree that the no-load voltage can be anything within reasonable limits, instead of just 11V?
The modern wall warts tend to be regulated, though.
>>902157
Positive to negative voltage inverter ics also exist very cheaply
>>902101
Not sure what you mean by an envelope, but yes, the caps smooth out the voltage due to their characteristics.
>>902161
>which was used in your two diode schematic as well.
shit, nigga, you're right. i had this other drawing in mind.
>instead of just 11V?
11V is an actual reading taken from my unit. could be up to 14V on low-current x-formers.
>>902182
Like in synthesis, to get the volume curve of a sound you rectify it and then LPF it. The lower the cutoff the worse the transient response is though (but less ripple).
