1
\$\begingroup\$

I'm designing a Eurorack synthesizer module which includes an MCU, DAC (TI DAC8565), and TL074 op-amp for scaling and buffering the DAC output.

In normal operation the module will operate from a +/-12V supply and use voltage regulators to get 5V and 3.3V for the digital components of the circuit.

During development, however, the module will be disconnected from the +/-12V supply and the MCU and DAC will be powered via USB. A diode protects the 5V regulator against the reverse voltage in this case.

My concern is that the DAC will be outputting 0-2.5V to one input of the op-amp, and a 1V reference voltage derived from the 3.3V will be going to the other input, and these are outside the op-amp's absolute maxima when the supply is 0V.

I'm considering two options to deal with this:

  1. Add diodes from the 5V rail to the 12V, and from ground to the -12V, so that when the +/-12 rails aren't directly powered, the op-amp gets a few volts of supply, so it can handle the DAC output. I don't care if the op-amp functions correctly here, I just don't want it damaged. enter image description here

  2. Hold the /RST line of the DAC low while the +12 is unpowered, and derive the 1V reference from the +12 instead of the +3.3, so the op-amp inputs are at 0V if there's no supply.

For implementing option 2, I see a bunch of questions here about how to switch 12V from a 3V control signal, but I want to do the opposite. The circuit given in this answer does the trick in simulation, but I note that when the control line is high, the transistor would be reverse-biased to the tune of around 8V, which seems bad. Is there a standard way of switching a low voltage with a high voltage, non-inverting?

And a final question, is it generally true that the absolute maximum voltage figures for an op-amp, given relative to supply voltage, hold true down to zero supply voltage? Will a couple of volts on the input of an unpowered op-amp damage it?

\$\endgroup\$
3
  • \$\begingroup\$ I guess that note (2) in the TL074 datasheet, absolute limit table is reason enough for adding protection. Option 1 is simple and should work although diode between GND and -12Vdc is reversed since voltage is relative. \$\endgroup\$
    – Ernesto
    Apr 16 at 18:11
  • 1
    \$\begingroup\$ I think you're being overly concerned. The ratings relate to the op amp when powered. With the supply off, it's not a supply of 0V, it's no supply at all. Excess voltages when powered may push too much current through the internals, but with no supply there is no current to force through. The inputs are JFET gates with a very high impedance. There is no significant current in the circuit to damage it. Think of it this way, are you going to disconnect everything connected to any op amp audio input before switching equipment off? \$\endgroup\$
    – Ian Bland
    Apr 16 at 18:26
  • \$\begingroup\$ Ahh, I see-- I'm not connecting the supply terminals to ground; they're floating in this mode. \$\endgroup\$ Apr 16 at 19:13
0
\$\begingroup\$

If you apply > 10 mA to any input greater than the Schottky diode protection rating of 0.5V outside Vcc,Vee then ensure a current limiting series resistor is used.

" Input signals that may swing more than 0.5 V beyond the supply rails must be current limited to 10 mA or less"

\$\endgroup\$
2
  • \$\begingroup\$ "This low impedance path may cause system upset or catastrophic damage due to excessive current levels. The Latch-Up condition typically requires a power cycle to eliminate the low impedance path." These two sentences are very confusing in conjunction. I'm concerned about permanent damage to the op-amp, and not at all concerned about a condition that can be cleared by power-cycling. \$\endgroup\$ Apr 16 at 18:58
  • \$\begingroup\$ Ok I made it more specific \$\endgroup\$ Apr 16 at 19:32

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.