0
\$\begingroup\$

In a situation where an ADC is measuring an internal voltage node rather than an external input (where one has little control over input voltage), does one need to implement some over voltage protection for the ADC (clamping diodes etc.)?

In the circuit below an opamp buffers some internal signal which cannot exceed 30V, and additionally the buffers power rail limits the output to the 30V rail. Following that there is a divider which limits the maximum voltage that the ADC receives to 3V.

schematic

simulate this circuit – Schematic created using CircuitLab

In this situation would it be acceptable to omit over-voltage protection for ADC, or what would be the justification for adding such protection?

\$\endgroup\$
5
  • 1
    \$\begingroup\$ The TL081 isn't a great choice for this, it can't go down to zero volts if the input is zero. It'll saturate at around 100~500 mV. \$\endgroup\$ – Hearth Dec 15 '20 at 2:43
  • \$\begingroup\$ Sorry, I didn't mean to put in a specific opamp, was the default in circuitlab. Will remove from diagram \$\endgroup\$ – zsky3333 Dec 15 '20 at 2:53
  • 1
    \$\begingroup\$ It seems reasonably safe to me. But if you wanted to put a Schottky diode from the ADC input to the 3.3V rail I suppose that wouldn't hurt. \$\endgroup\$ – mkeith Dec 15 '20 at 3:58
  • 1
    \$\begingroup\$ That's perfectly safe, you don't need an over-voltage clamp. \$\endgroup\$ – td127 Dec 15 '20 at 4:05
  • \$\begingroup\$ See datasheet specification of amp: absolute maximum ratings \$\endgroup\$ – nick_n_a Dec 15 '20 at 17:03
1
\$\begingroup\$

No, there is no reason whatsoever for added circuitry.

Circuits that provide a clamping action tend to add inaccuracy (for example Schottky diode leakage, especially at elevated temperatures) and they add complexity (which affects reliability) and cost, PCB area.

If you don't need it, don't add it.

\$\endgroup\$
1
\$\begingroup\$

You have a 30V supply and an 11:1 divider. So as long as "some internal voltage signal" is not so large that it causes OA1 to malfunction, you should not get a value greater than 2.73V at the ADC input. If the ADC can accept voltages in the range of 0V to 2.73V, then you don't need additional protection.

I would note that the output of your divider has an impedance of 909 ohms, which is quite high if you are feeding an ADC. You will probably get better ADC readings if you add a small capacitor (10nF to 100nF) in parallel with R2. Alternatively you could put the voltage divider before the divider, and use a 3.3V supply for the op amp.

\$\endgroup\$
3
  • \$\begingroup\$ The output impedance is 909 ohms and not 9.09 kohm. \$\endgroup\$ – Andy aka Dec 15 '20 at 12:27
  • \$\begingroup\$ @Andyaka You are correct. I fixed my typo. In any case, even 909 ohms probably would benefit from either a capacitor or swapping the order of the divider and op-amp. \$\endgroup\$ – user4574 Dec 15 '20 at 14:45
  • \$\begingroup\$ Ah but now, 909 ohms is probably OK for many, many ADCs without adding a capacitor. \$\endgroup\$ – Andy aka Dec 15 '20 at 14:47

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.