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, 2020 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, 2020 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\$
    – user57037
    Dec 15, 2020 at 3:58
  • 1
    \$\begingroup\$ That's perfectly safe, you don't need an over-voltage clamp. \$\endgroup\$
    – td127
    Dec 15, 2020 at 4:05
  • \$\begingroup\$ See datasheet specification of amp: absolute maximum ratings \$\endgroup\$
    – nick_n_a
    Dec 15, 2020 at 17:03

2 Answers 2

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, 2020 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, 2020 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, 2020 at 14:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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