0
\$\begingroup\$

I have a circuit input that I need to measure down into the nanovolts, so it gets fed to a 24 bit ADC via some buffer amps. However, in order to protect the circuits from overload I am thinking of putting zener clamps across the unconditioned signal input lines. What, if any, noise problems might this cause?

Signals are low frequency (<1kHz) and voltages up to 15V

\$\endgroup\$

closed as too broad by Andy aka, PeterJ, Daniel Grillo, Fizz, jippie Nov 1 '15 at 10:26

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    \$\begingroup\$ Presumably it's an AC voltage you wish to measure? Why not show the proposed circuit? \$\endgroup\$ – Andy aka Oct 26 '15 at 9:40
  • \$\begingroup\$ @Andyaka First, it is part of a complex analog PID loop and second because of confidentiality issues. It is an attempt to put some basic ESD on the inputs to the board. However, we do use more than 20 bits of resolution so any noise contribution from the (supposedly) non conducting zeners will show up \$\endgroup\$ – Dirk Bruere Oct 26 '15 at 10:03
  • \$\begingroup\$ @DirkBruere You could at least show an equivalent schematic; things like input impedance may have an impact on the answer. \$\endgroup\$ – Nick Johnson Oct 26 '15 at 10:04
  • \$\begingroup\$ Input impedance is around 10K, with two zeners being used back to back across the inputs. \$\endgroup\$ – Dirk Bruere Oct 26 '15 at 10:21
  • 1
    \$\begingroup\$ I'm voting to close this question as off-topic because proper details can't be supplied about the circuit. Those details that can be provided make the question too broad to answer IMHO. \$\endgroup\$ – Andy aka Oct 26 '15 at 11:09
2
\$\begingroup\$

Low voltage zener diodes (<5v) work by a different breakdown mechanism to those above 5v, and are very much softer, conducting significant nA at quite low voltages, which could interfere with your readings. However, you say that you will use 15v signals, and zeners to handle these voltages are much sharper. Whether they are sharp enough in your application, you will have to measure.

The input clamping ESD technique I always use is to use silicon diodes to voltage sources, ideally the rails for convenience, but if it has to be lower voltages, then a biassed zener diode in parallel with a huge electrolytic to quiten any noise. The reverse bias leakage in a good silicon diode should be much less than the sub-threshold leakage in a zener, so interfere less with the input.

\$\endgroup\$
2
\$\begingroup\$

First, just because an ADC is 24-bit, that doesn't mean you're getting anywhere near 24 effective bits. My experience with high-res ADCs, especially the sigma-delta's, is that when you go to the data sheets, you'll find that you're actually getting closer to 17 bits- So, you're not getting that nanovolt reading you think you are without some amplification, unless you're taking advantage of an on-ADC PGA. If you're using a PGA with a gain of 256, you're talking hundreds of nanovolts, which may be close. Otherwise, you probably should consider a 32-bit ADC. Preamplification might even be preferred, unless there's a good reason why that isn't possible.

The standard way to protect the inputs is through stacked diodes to the rails.

Don't forget-- you still need to anti-alias filter.

\$\endgroup\$

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