record signal from photodiode

[abdel2]

I am using the following photodiode:

http://ca.mouser.com/Search/ProductDetail.aspx?qs=ck8ERu4TygdFXGB8JByQYw==

When I connect VDD-GND to a 5V power supply and record the output signal using either USB 6008 or PCI 6259 with a BNC 2090A board, I get a very noisy signal that goes up to 15V. This doesn't make any sense because the maximum power supply voltage is 5V. The signal I'm recording is differential and it is coming from Vo-GND outputs of the diode. Whether the photodiode is in the dark or light, the signal's pattern doesn't really change.

However, when I power up the photodiode using one of the analog outputs of the DAQ card, I can see that the signal varies depending on whether the diode is the dark or not. But the fluctuations in the signal are still there. In the presence of light, the signal will fluctuate from 2 to 2.8 volts, but in the dark, it will be almost stable at 2 volts.

What could be the problem here? Do I need to do some sort of impedance matching? In the datasheet, it says the feedback resistance is 80Mohms.

All 3 wires coming from the photodiode are also not grounded. They are 26AWG wires, all wrapped in a heat shrink. Would I need to provide shielding? I remember I got similar problem when I tried to get data from an accelerometer when using an external power supply. Currently I'm just using one of the analog outputs to power it up and it's working ok.

[BROhan]

Hi abdel2,

The output resistance of the PCI-6259 is 0.2 ohms, so if you needed to impedance match, you could put a 0.2 ohm load resistor to ground on the output of your photodiode, and measure differentially from there.

I would also suggest measuring the signal with a DMM and seeing if you get the expected voltage with both the DMM and the DAQ devices.

Based on the diode sheet, a 5V supply should be enough, but you can certainly output from the 6259 if necessary.

[abdel2]

Thanks Rohan. I'll try to use a 0.2 ohm resistor. I tried a 20Kohms resistors before and it didn't make any difference. I also measured the signal with a DMM and it gave me correct voltage.

After further testing, I have noticed that it's only PCI 6259 and PCI 6221 (We have both set up in different room) that do not record the correct voltage level from the photodiode. First I thought it was because of electric noise inside the rooms so I tried USB 6008 both inside and outside the room. It always gives me the correct voltage level and I can see the changes in the signal as the light goes on and off. In fact, when I used a shielded cable, the different in voltage between on and off sate were even more prominent (cleaner signal). I'm not sure why I was getting similar behaviour with USB 6008 as PCI 6259/21 before. The noise recorded using PCI 6259/21 seems to be close to 60Hz. The only puzzling part is the the actual voltage amplitude which is close to 15 volts.

So basically I have tried 5 different cards and ITC DAQ card, USB 1208FS by Measurement Computing, and USB 6008 by NI are the only ones that give me the correct response. I ordered some more photodiodes that are non-powered and I will be testing that tomorrow to see what kind of response I get on different DAQ cards. Hopefully PCI 6259/21 will respond better without needing additional circuitry because I have been told by a colleague that he can record from these non-powered photodiodes easily on ITC. The signal is a bit noisy but you can still see the difference between a dark and a light monitor screen.

[BROhan]

Hi abdel2,

Since you are seeing 60Hz noise (probably due to the the power going through your light bulbs), I would suggest to use a lowpass filter to remove that noise. However, it is very strange that you are seeing 15V across to output of your diode, as the diode is only supplied with a 5V supply (per your initial post). Definitely try the load resistor and measure across it differentially. Let us know how it goes.