1
\$\begingroup\$

I'm looking for a way to somehow measure voltages in the range of 60~90V with an arduino analog input. These voltages are biasing a rather sensitive APD exposed to very low light levels, which supplies very small currents (~0.8 uA) for a transimpedance amplifier. The biasing circuit is shown below.

enter image description here

I though about using a voltage divider with the lower resistor having a bypass Zener to protect from over voltage, but that would require very large resistances (8~10 GOhms, and I'm mounting this on a PCB). What kind of interference could this have on the APD measurement?

\$\endgroup\$
2
  • \$\begingroup\$ Very large resistances on an ADC input can seriously skew the results. It is generally accepted that the impedance of a device connected to an ADC shouldn't exceed 10% of the impedance of the ADC's input. \$\endgroup\$
    – Majenko
    Commented Jun 13, 2014 at 10:55
  • \$\begingroup\$ How acurrate measurements you need? \$\endgroup\$
    – Kamil
    Commented Jun 13, 2014 at 11:29

1 Answer 1

1
\$\begingroup\$

I though about using a voltage divider with the lower resistor having a bypass Zener to protect from over voltage, but that would require very large resistances (8~10 GOhms, and I'm mounting this on a PCB).

You certainly do not need GOhms. Note that the circuit you have specified uses a 47k resistor on the output. If you replace this with a voltage divider (say, 42k / 5.1k) you will get 3 - 6 volts with a 5k impedance. If this impedance is too high for your ADC, you can always buffer it with an opamp. Note that the 47k resistor shown will only dissipate ~ 170 mW at 90 volts, so it's not as if you'll save a whole lot of power by increasing its value.

And, just for future reference, zeners are not appropriate for microamp limiting. Typically their leakeage current is in this range.

\$\endgroup\$
1
  • \$\begingroup\$ Indeed I had noticed the resistor at the end, and as such it would be relatively simple to build a voltage divider to measure output. I'd also wager that a buffer between the measurement node and the analog input of an Arduino would serve as protection against overvoltage - rendering a Zener useless either way. \$\endgroup\$
    – joaocandre
    Commented Jun 13, 2014 at 16:14

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.