1
\$\begingroup\$

I'm trying to build the power supply described here.

enter image description here

R1: 1K8 1/4W Resistor

C1,C2: 4700µF 25V Electrolytic Capacitors

D1: Diode bridge 100 to 400V, 1.5 to 4A

D2: LED Any type and color

SW1: SPST Mains switch

T1: 230V Primary, 30V Center-tapped or 15 + 15V Secondary, about 45 to 60VA or 1.5 to 2A, Mains transformer

PL1: Male Mains plug with cord

In fact, I already did it and it's working. The problem is that the output voltage is higher than the voltage supported by the IC (ne5534), that should be 22V (maximum).

After I built it, I measured +-24V. I would like to limit it to +-20V. How do I do this?

Searching the internet I found that I could use LM317/337 (since I need +/- output). Or I could use 78xx/79xx (it seems to have a 18 or a 24 version, but not a 20V). Or I could use a resistor. Or diodes (regular or zener).

I would like to use the resistor/diodes solution, since I have some spare/salvaged ones In hand.

If I calculated (V = I*R and P = I^2 * R) correctly, to use resistors, I should use two 4R 4W (one for each output).

If I use standard silicon diodes, it will require a minimun of 6 (in series) for each (12 total).

The approach that I prefer is the use of a zener diode, but unfortunatelly (I'm a hobbyst) I dont know how to calculate it... My problem is with the current. How much power the diode/resistor should accept (considering 1A). I found many papers describing zener theory, but I didn't manage to solve it by myself... It seems that a 20V 1W zener diode would fit, but it will require the same 4R4W resistor to regulate the current...

\$\endgroup\$
  • 2
    \$\begingroup\$ Your power supply is unregulated. Its output likely to fall somewhat just from loading. Try just putting a resistive load on it (drawing a current similar to your amplifier) and see what the output is then. \$\endgroup\$ – The Photon Feb 16 '15 at 5:43
1
\$\begingroup\$

Your root problem is that either you are using the wrong transformer or your mains voltage is too high. Replacing your transformer with a lower-output unit is your best bet.

Assuming that you really are using a 30 VCT transformer, rather than the 36 VCT unit which I suspect you have, replace it with a 24 VCT unit.

None of the solutions which you are considering will work terribly well. A zener can do the job, but you need a high-power device in order to handle the surge currents which occur at the mains voltage peaks. If you use 1 ohm resistor and a 20 volt, 10 watt zener you should be OK, but those things are expensive.

Something like an LM317/LM337 will not work well, since at full amplifier output you're running beyond their current capability, and the reliability of the circuit will be very poor.

Putting a resistive load on the supply to drop the voltage is actually the worst of your options. The problem is that it won't do any good if you don't have a load connected to the amplifier. Even during test you would need to maintain a load, or risk killing your op amp.

Using diodes to drop the voltage is just as iffy as using resistors, and for the same reason.

\$\endgroup\$
  • \$\begingroup\$ Actually, low-power transformers have horrible regulation that needs to be taken into account when designing a circuit. Finding one with 25% more voltage open loop than when fully (loaded) to the rated voltage and amperage is actually rather common/easy. I have on hand 24V+24V 16VA transformers that output 29-30 V (RMS) open loop [that 1V is due to main variation between measurements at various points in time], but the 5-6V "extra" are just due to low/lack of load. (Fully loaded they give just 24V.) \$\endgroup\$ – Fizz Sep 22 '15 at 23:35
  • \$\begingroup\$ If you're luck to find transformers that have this [de]regulation number in the datasheet, it agrees with what I said above. Here's an example from a different mfg, but similar product where you can find this stuff in the datasheet (15VA, 2x25V under load, but 2x29V open) tme.eu/en/Document/505ab7a151e6fa111ec7ce1e42277c4f/55xxx.pdf \$\endgroup\$ – Fizz Sep 22 '15 at 23:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.