Problems with a LINEAR power supply

[Francesc V.]

Hi:

I bought an ebay kit to assemble a linear power supply. The following one that is explained in this page:

https://www.qsl.net/z33t/dc_0-30v_0-3A_eng.html

The power supply seems to work, but I have the following problems I would like someone to help me.


1. The input must be about 24-26V AC. My idea was to reuse a 30VAC 160VA toroidal transformer. But with this input the output doesn't go from 0-30V......it remains with some mV. And I guess will stress/burn some components. Do you see any possible modification to not increase the costs by buying a new transformer?

2. Using a 24V AC thanks to a variable transformer, it seems to work, but voltage goes up until 26V.......not reaching 30V. How could it be fixed?

Thanks in advance,
Francesc.

[CapLeaker]

Re: Problems with a LINEAR power supply

Yeah, I believe s little thing would struggle at 30VAC. 30VAC x 1.414 = 42.42VDC. Plus the voltage would be even higher if there is no load on the transformer. No wonder she wasn’t working quite right!

You got 2 or 3 choices:
A: Get a proper 24VAC transformer
B: unwind the secondary or lower the input voltage of the 30VAC transformer (does the transformer have taps on the input and output?)
C: junk the bridge rectifier on the linear supply board and feed it 34VDC.

[harp]

Re: Problems with a LINEAR power supply

I think that you first must be sure that your psu is working corectly regardles to your point 2. max v of 26v...
So, disconect track where X and measure voltage at point A with 24vac input.
If it is about 33vdc, then conect ampermeter between point A and B to check what is lowering your power.

Posible modification is to drive opamp through some zener regulated circuit, also managed to drive fan, and change electrolitic capacitors on any rail to fit newer higher voltages.

[eccerr0r]

Re: Problems with a LINEAR power supply

For the full 30V output for a linear regulator, the input needs to go at least 1.4V over the output voltage, so at least 31.4V at the positive plate of C1. So 24VAC RMS input to 31.4V DC? 24v RMS - 1.4V (for 2 diode drops of the the bridge) * sqrt(2) = 31.95V

This is really pushing it. Plus the power factor will be miserable - your transformer will only push the voltage at the peaks, running at probably 50A or so, and the rest of the cycle it's just sitting there doing nothing because its voltage is lower than the output voltage.

So no, unless you're drawing milliamps at 30V, a 24V transformer will not cut it.

Using a 32VAC RMS transformer is what you need to use. However with a 32V transformer, you will dissipate more heat. First off now at 32VAC your rail will be 43.3V which now exceeds the voltage limits of the op amp which will fry those. (TL081H might work as a sub, but it has a limit of 40V.)

So I think the 30V output is extremely optimistic with a 24VAC RMS transformer and if you really need the 30V output, a lot will need to be modified to be able to handle it...

[eccerr0r]

Re: Problems with a LINEAR power supply

oh and you _DO_ need the AC for this circuit, if you want to feed it DC, some major circuit hacks is needed beyond dumping the bridge rectifier.

[redwire]

Re: Problems with a LINEAR power supply

These 0-30V 3A power supply kit boards are pushed to the limit.
The 1978 original circuit used LM741's with max. Vcc 44V.
The kits IC's TL082 op-amp max. Vcc is 36V, and don't forget the -5V rail so the IC's are run over their max.

I would consider wiring the toroid for 1/2 the voltage, or removing secondary turns.
Why does everyone want 30V+ output when it's rare you need that unless you work with 24VDC equipment.

[Francesc V.]

Re: Problems with a LINEAR power supply

Hi. I decided to remove secondary turns from the transformer to reuse it......... I have pending to measure parts of the circuit and put it here for your review/opinions.

Do you think that by just removing the Op. amps and put 3 LM741 will fix the problem?

[eccerr0r]

Re: Problems with a LINEAR power supply

heh, funny, after all the effort to stop people from using 741's, but they're better for this one application!

[Francesc V.]

Re: Problems with a LINEAR power supply

I've used an old ATX case.......

[sam_sam_sam]

Re: Problems with a LINEAR power supply

I've used an old ATX case.......

Very nice job reusing a computer switching power supply for your enclosure