2
\$\begingroup\$

After ordering and assembling one of those¹ "red" 0-30V/0-3A power supplies from eBay/AliExpress and taking a bit of a closer look at it, I'd like to replace the TL081CP op-amps. It's working fine with the specified 24VAC input, but two of the three op-amps are being powered at a total of almost 37V, definitely beyond their Absolute Maximum Rating of ±18V.

Schematic as drawn by Mile Kokotov at https://www.qsl.net/z33t/dc_0-30v_0-3A_eng.html

The only op-amp I've used before and know to be cheap and readily available and that handles higher VCCs (up to ±22V) is the NE5532, and that's a pin-incompatible 2-in-1.

How do I find another op-amp that is (a) still available, (b) affordable, (c) pin-to-pin-compatible, and (d) happy with, let's say, up to ±20V, so that it would work as a replacement for the above-shown circuit? Edit: as long as the other pins are a fit, I don't much care about the presence/absence of the null-offset pins (the ones which have RV1 connected to them).

\$\endgroup\$
5
  • 3
    \$\begingroup\$ While this is a well formulated question and very clear as well, I fear that it is still asking for a product recommendation. Question about those are considered off-topic here. That being said, with some parametric filtering in Mouser or other Distributors you might find what you are looking for. \$\endgroup\$
    – Arsenal
    Commented Sep 9, 2020 at 12:04
  • \$\begingroup\$ I came here after trying parametric filtering on various distributor and manufacturer websites and feeling overwhelmed by the number of parameters, half of which I don't even know if could understand them yet, even if I tried. Sad to hear that's considered off-topic here, considering that unlike with software, there could be no electronics without manufacturers producing products and giving them names. \$\endgroup\$
    – Sixtyfive
    Commented Sep 9, 2020 at 12:14
  • 4
    \$\begingroup\$ NE5534 is a single opamp version of the 5532 with a similar pinout to the TL081. Check out its pinout and voltage ratings. \$\endgroup\$
    – user16324
    Commented Sep 9, 2020 at 12:30
  • \$\begingroup\$ @Sixtyfive asking the question in a way which helps you to learn how to choose a suitable substitute would be on topic and might generate a nice answer. Spehro Pefhany's answer is pointing in that direction. The problem with recommendations is that the parts go out of production and people looking for hints end up with a replacement part which also doesn't exist anymore. I'm not sure I agree with that approach because I think some things can be learned by selecting a replacement with a good answer, but that is a question for meta. \$\endgroup\$
    – Arsenal
    Commented Sep 9, 2020 at 14:39
  • \$\begingroup\$ Hi @Arsenal! I did so; would you re-open the question or else, let me know what else is still lacking so that I have a chance to learn from it? \$\endgroup\$
    – Sixtyfive
    Commented Sep 14, 2020 at 12:40

2 Answers 2

Reset to default
6
\$\begingroup\$

Search parametrically, say at Digikey for op-amps in the same package, with one op-amp and that are in stock and active.

Then select ones that have dual supplies of +/-20V or more or single supplies of 40V or more. That's using the column "Voltage - Supply, Single/Dual (±)".

I suggest doing it in that order, because selecting the individual entries in voltage range is rather tedious and you want to cull the number of entries down as much as makes sense first.

When I do that, not so carefully, I get 23 types of op-amp. Sorting by price, there are only a few that cost C$5 or less each, all from TI either originally or via NS.

If you decide to make the substitution you have to look at the circuit characteristics to make sure that the new op-amps won't cause new problems and verify from the datasheet that the characteristics in the distributor parametric search are correct (sometimes they are a little wrong sometimes very wrong). Also note that not all offset trim circuits connect the pot in the same way and connecting a pot in the wrong way (even if you don't intend to use it) might destroy the chip, so again reading and understanding the datasheet and circuit is crucial.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for the specificity of your guide, Spehro! I followed that procedure and arrived at the UA741 of which I ordered two dozen, since they could be had very cheap over eBay. \$\endgroup\$
    – Sixtyfive
    Commented Sep 13, 2020 at 20:42
-2
\$\begingroup\$

If it is working, don't fix it.

It is very strange that something rated to 18 V is working at 37V, but it is not strange that something rated to ±18 V works at 37V.

The datasheet says that the supply voltage of OpAmp is ±18 V: this is a spec for a symmetrical power supply.

If you probe the IC's supply pins with a multimeter, ±18 V reads as 36 V (DC). So, it does not seem that this OpAmp is working "definitely beyond" the specs, not in such a way that someone should bother to "fix" it.

If it is working, there is no need to fix it.

enter image description here

\$\endgroup\$
6
  • \$\begingroup\$ 37V > 36V, which is the absolute maximum rating. That's very bad practice, even for a hobbyist. And ripple is on top of that. And mains regulation is on top of that. \$\endgroup\$
    – Spehro Pefhany
    Commented Sep 11, 2020 at 18:45
  • \$\begingroup\$ @SpehroPefhany Ok, you're right, 37> 36! :-) But still... OP measured 37 V from + 12 V to the negative rail; this measure doesn't seems reliable, since the negative rail is feeded from the bridge-rectifier AC input (digital multimeters aren't good for that kind of measure, and there is 99% chances that OP uses an DMM). I wish we could put this circuit in a scope to see the true voltage in those OpAmps... I somewhat agree that I deserve a downvote, but I still keep the answer, because I don't see the reason for fixing something that it's not broken. \$\endgroup\$
    – mguima
    Commented Sep 11, 2020 at 18:52
  • 1
    \$\begingroup\$ No. OP says the total voltage was 37V. Your argument sounds like the argument for doing nothing about the space shuttle O-rings because nothing had blown up yet. \$\endgroup\$
    – Spehro Pefhany
    Commented Sep 11, 2020 at 18:55
  • \$\begingroup\$ @SpehroPefhany :-) OP said: "two of the three op-amps are being powered at a total of almost 37V" probably he measured the +VCC to -VCC pins. There can be measuring errors. I don't want to blow the shuttle, but first I would check zener diode D7 (or even change D7 for an lower value) before replacing OpAmps in a functional circuit. \$\endgroup\$
    – mguima
    Commented Sep 11, 2020 at 19:04
  • 1
    \$\begingroup\$ OP here, I did mean there's some O-rings being used out-of-spec and I don't want to see them fail. FWIW, after changing some other parts in the circuit that were underspec'd, it's now a good 37.4V between two of the op-amps' -VCC and +VCC. Somehow impressive but sooner or later the magic smoke will get out. \$\endgroup\$
    – Sixtyfive
    Commented Sep 11, 2020 at 21:38

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