Zero for VCP1000 - 20-bit (+-40V) Voltage Input Phidget

[alexeialexandrovski]

Cannot get zero voltage reading. For the open circuit VCP1000 measures 2V. With 50 K shunt it is 0.18V. With 21 K shunt it is 0.08 V. I am trying to get voltage for the photodetector, Si photodiode. It is a source of current (means its internal impedance is closer to Gigaohm). I was using 50 K shunt all the time with 1 Mohm voltmeter to not overload it. VCP1000 has same impedance 1 Mohm but somehow adds its own voltage which depends on the shunt resistor. Is it a defective unit or a "normal" response?

[mparadis]

Are you using the Phidget Control Panel and getting this behaviour?
Is it connected to a VINT Hub, or another Phidget?

[fraser]

This is expected behaviour, but you will only see these voltages with a shunt alone, or when floating. The floating voltage is 2V (there is a 2V reference connected to the positive input terminal through 1MOhm). When voltage is applied to the terminals this will be negated and the measurement will be correct.

[alexeialexandrovski]

It still shifts zero when I am trying to measure the laser power with the biased, Si Thorlabs photodetector DET10.
Experiment:
1. 1 Mohm DC input of the SRS lock-in SR850 shunted with 50 K resistor to avoid overload (the photodiode current is about 0.1 mA, the output voltage of the photodiode will be limited by 12 V which is the bias voltage).
Laser on : 5.51 V
Laser off: 0.000 +/- 0.001

2. VCP1000 in parallel with the shunted SRS lock-in:
Laser on: 5.23 V (SR850); 5.28 (VCP1000)
Laser off: 0.160 V (SR850); 0.1618 V (VCP1000)

3. VCP1000 only with 50 K shunt.
Laser off: 0.1686 V
Laser on: 5.50 V

Obviously, the shunt cannot be used with VCP100 when lower current from the photodiode is flowing.

[alexeialexandrovski]

Are you using the Phidget Control Panel and getting this behaviour?
Is it connected to a VINT Hub, or another Phidget?

Yes, it is connected to VINT Hub.

[alexeialexandrovski]

This is expected behaviour, but you will only see these voltages with a shunt alone, or when floating. The floating voltage is 2V (there is a 2V reference connected to the positive input terminal through 1MOhm). When voltage is applied to the terminals this will be negated and the measurement will be correct.

I posted the experiment. It looks like no shunt can be used even when the voltage source is connected. I will check now how it works with lower current from the photodiode and no shunt.

[alexeialexandrovski]

Are you using the Phidget Control Panel and getting this behaviour?
Is it connected to a VINT Hub, or another Phidget?

And, yes, I am using the Phidget Control Panel to get the DC voltage.

[alexeialexandrovski]

This is expected behaviour, but you will only see these voltages with a shunt alone, or when floating. The floating voltage is 2V (there is a 2V reference connected to the positive input terminal through 1MOhm). When voltage is applied to the terminals this will be negated and the measurement will be correct.

With no shunt and attenuated laser power I am getting wrong voltage from the photodetector with the VCP1000. When the attenuation is strong, the VCP1000 shows 2 V as if no voltage source is connected. I can understand that because Si photodiodes have huge internal impedance, may be Ghoms.

That opens the following question: which impedance of the voltage source should be OK for the VCP1000? Or maybe, since you say "when voltage is applied to the terminals this will be negated", which minimum voltage that would be? 2 V? Then the measurement range of the VCP1000 is limited by 2 V from below, correct? It is 2 v to 40 V then?

[alexeialexandrovski]

Figured out how the measurements are affected by the 2 V "floating" voltage. Have a spreadsheet to share if anybody is interested.
The simplest model worked well in experiment where 2V are in the loop together with a 540 K resistor, not 1 MOhm. The result is a bias. You may neglect it when the impedance of the voltage source is low. You cannot do it when you connect Si photodetector which is effectively a source of current which means the internal resistance is very big.

As a numerical example, for the 47 K shunt across the Si detector output the bias is 0.16 V calculated, 0.17 V experimental. For the 148.5 K shunt the numbers are 0.43 V both. At 1 K the calculated shift is 3.6 mV only.