0

I have these circuits: enter image description here Their goal is to protect the +vcc and -vcc pins of an opamp. They limit the voltage to +30 V (-30 V) and 25 mA in normal conditons.

The 1200 ohm resistors act like 25 mA load at 30 V.

When there is a short on one rail, this circuit drops the Vout voltage. Which means that the supply isn't "symetric" anymore. And i don't think thats good ( more about that here and here). So what I would like is that when a short happens on one rail the other one is also shutdown until the fault is adressed and cleared. I was thinking of some sort of voltage monitor that would trigger a BJT or MOSFET on one rail when the votlage on the other rail is not high enough. But the negative rail makes thing a bit more complicated for me.

OPA445 datasheet

Neeko
  • 51
  • 7
  • Hmm. Maybe use two [window comparators](https://en.wikipedia.org/wiki/Window_detector) (one for each *supply*), then sum (AND) their outputs and use that signal to turn on/off both outputs. – rdtsc Apr 20 '21 at 14:16
  • 1
    I am trying to understand what you mean by "protecting" the op-amp from undervoltage. Does something bad happen if the voltage is too small? What if a signal is applied while the power is turned off? Wouldn't you rather turn off the _signal_ if the voltage is too low? I'm just trying to understand what you are aiming at. – Math Keeps Me Busy Apr 20 '21 at 14:41
  • 1
    *Q5&Q6 and Q8&Q7 are undervoltage protection because my opamp is rated for +/-10 to +/- 45 V.* Why all this complexity? All other circuit desigers including myself would simply make a +/- 30 V supply and feed the opamps with that, **DONE**. Even though I've commented on your previous questions about this that **you don't need this**, you're still "protecting". Explain us **WHY**? – Bimpelrekkie Apr 20 '21 at 14:50
  • I've downvoted this question, that can turn into an upvote if you convince me that these protections are needed. For that you will need to provide one or more sound (technical) arguments. – Bimpelrekkie Apr 20 '21 at 14:54
  • @Bimpelrekkie it's just that those are the opamp's specs (opa445 first line)... I don't absolutly want it. But if this is useless, I'll drop it I got it now, I'm not trying to make you angry. But what i do need is to "ensalve" each rail to another if one goes off. Maybe I'm wrong but i believe it's not great if +vcc is supplied and -vcc is 0(or the other way). If thats wrong i don't get why the datasheet says +/-10 to +/-45... I'm more than happy to learn. – Neeko Apr 20 '21 at 15:10
  • @MathKeepsMeBusy it's the opamp datasheet that indicates +/-10 to +/-45 supply votlage. OPA445 form TI – Neeko Apr 20 '21 at 15:12
  • 2
    *I'm not trying to make you angry.* Oh, I am not angry at all! :-) I am just challenging your design choices so that we can have a discussion and maybe someone will learn something (you or me, if you have good arguments). *opamp datasheet that indicates +/-10 to +/-45 supply votlage* And that means that you should design your circuit such that the supply is between +/-10 to +/-45 V. If you use voltage regulators to provide +/- 30 V **you would be done**. There is really no need for (extra) protection. – Bimpelrekkie Apr 20 '21 at 15:24
  • @Bimpelrekkie ok, so now I get it. The minimum voltage is useful when designing a circuit but for the normal operation of my circuit it won't be an issue. But i still have one issue. I edited my post about that. Is my new thinking also wrong? Because asymetric supply can have an impact on the outpout signal of my opamp. (And it's also good that the user knows that something is wrong, which is easier to realize if there is no signal rather than a signal that is wrong). – Neeko Apr 20 '21 at 16:55
  • If it's audio their ears will tell them very quickly due to the terrible distortion. – Transistor Apr 20 '21 at 17:06
  • 1
    The datasheet (which I added to your question) in the table "Electrical Characteristics" in the row "Operating Voltage Range" in the column "Min", gives the value +/-10V. This means that _proper_ operation is not guaranteed if the supply voltage falls below that level. However, the IC will not be _damaged_ by operating _simply_ with a low supply voltage. [There may be other issues, like input signals exceeding the supply rails, but barring these other issues, there won't be _damage_]. So what are the negative consequences you are trying to prevent with cutting off a low supply voltage? – Math Keeps Me Busy Apr 20 '21 at 17:10
  • @MathKeepsMeBusy Now i get what Bimpelrekkie tried to explain. So i withdrawed the UVLO. But asymetric supply will have an effect on the OpAmp's outpout signal. And it's not audio as Transistor suggested. So I think it's a good thing not to let that asymetry. I guess the term "protection" isn't the best one because I want to do several things at a tiem... And thanks for the edit, i might be useful if someone stumbles on this thread. – Neeko Apr 20 '21 at 17:15
  • *But asymetric supply will have an effect on the OpAmp's outpout signal* **Why**? It depends on **how you use the opamp**. If the opamp is in a circuit with feedback and the output voltags stay within the opamp's limits at the given supply voltages, then **there is no issue**. Example: suppose my opamp circuit needs to output a sinewave of +/- 5 V. I can power the opamp with +15 V and -10 V and there would be **no problem**. You should stop writing un-funded claims like *asymetric supply ...* as they are not always true. It is better to say "I do not know" and then do research or ask here! – Bimpelrekkie Apr 21 '21 at 12:05
  • @Bimpelrekkie I do not know which is why I'm asking. I guess i didn't right it as a question. I don't want to mislead people when I'm talking about asymetric supply. But when I have 30V on one rail and 0 on the other it does sound asymetric to me. Which is why i uses this term. Now, when I go to [this post](https://electronics.stackexchange.com/questions/85683/what-is-the-effect-of-supply-voltage-asymmetry-in-opamp-amplifier-topologies) I understand that there can be an issue with asymetry in supply (or what ever name it's called). I based my previous comment on that idea. – Neeko Apr 21 '21 at 16:07
  • Again, if it's wrong, I'll do my best to try to understand. I just want to learn and understand. (And technical discussion in a foreing language isn"t my strong force). – Neeko Apr 21 '21 at 16:14
  • *But when I have 30V on one rail and 0 on the other it does sound asymetric to me* OK, that's called a "single supply". In that case the opamp's inputs and outputs are **biased** at a voltage "in the middle", like +15 V using extra resistors. Then **to the opamp** it looks like a +/- 15 V supply. I suggest that you read "Opamps for everyone" (a free Ebook) and also search in Google for opamp circuits. Just to see how things are done. – Bimpelrekkie Apr 21 '21 at 17:53

0 Answers0