03-27-2013 03:21 PM
To anyone who might be able to help:
For my Senior Design project my team is constructing a non-invasive blood pressure monitor for dogs by using an NI my-daq and a piezoelectric film sensor.
Through troubleshooting iterations with a circuit I finally, randomly, decided to plug the sensor directly into the my-daq hardware via analog inputs 0+ and 0-. I could now see a waveform (in labview) whenever I deformed the sensor. I could even get my blood pressure when firmly placing the sensor on an area of my body that had a pulse (e.g. neck, wrist).
Here's the problem I am trying to fix: With the sensor just lying there (no input), the my-daq gives an output of 10.5V on my labview graph. As if it's saturated (because the dynamic range is +10V, -10V). But it's obviously not saturated since there shouldn't be an input and I can produce a signal by deforming the piezo sensor. I want this base value to be zero.
I called NI tech support and confirmed that there is no problem with the my-daq hardware itself. SO...what's going on? Probably some minor, simple, overlooked issue. Just really need another set of technical eyes. Attached pictures include an image of the labview graph and my-daq hardware.
Thanks,
David
----------------
David Tagawa
Biomedical Engineering
Texas A&M University
210-544-4391
03-28-2013 04:02 AM - edited 03-28-2013 04:04 AM
I don't know the my-daq nor your sensor, but I assume that you just discovered the effect of the input bias current.
The spec of the my-daq doesn't specify the max input bias current, but you can try to measure it by doing two measurements: one with the input shortcut to measure the zero offset and the second with a 1Meg resistor . The difference of both measurements is a result of a small current that is flowing in or out of your input. Do this on both input in single ended mode and try it in differential mode
(small homework 😉 ) and calculate the input bias current.
Your sensor can be seen as a capacitor that is integrating the bias current -> the voltage in your case is increasing until the input saturates.
And if you know the input bias current and connect your sensor make a short term short cut and log the voltage increase after you remove the shortcut. Now you calculate your source capacity incl. the ~100pF of the myDAQ.
Depending on your sensor you can try to provide an extra path for this current by adding two resistors like Fig.9 in the myDAQ manual.
This will implement a high pass (mean you can't measure DC signals, since your sensor capacity is continuously discharged) but in your case that might even be a feature 🙂
Try two 1Meg or 10Meg resistors .....
Another option would be using (or building) a charge amplifier ...
04-01-2013 03:09 PM
Thanks! 2 Meg resistors did the trick!
04-01-2013 06:59 PM
I had a similar problem. Thanks Henrik_Volkers
12-17-2014 12:32 PM
I'm having a similar issue, but I don't have the myDAQ plugged into anything. I'm just reading -10.5V without any input hooked up. Is there still a bias current, and can I do something about that?
12-18-2014 09:56 AM
Hi daladethman,
The behavior you are seeing is expected for an open analog channel. You can typically expect the voltage to float.
Also, we do ask that for new questions you create a new forum post. Posting on old threads (like this one) is not recommended. More information can be found in the Community User Guidelines here: http://forums.ni.com/t5/Feedback-on-NI-Discussion-Forums/National-Instruments-Community-User-Guideli...
12-18-2014 08:50 PM
@daladethman wrote:
I'm having a similar issue, but I don't have the myDAQ plugged into anything. I'm just reading -10.5V without any input hooked up. Is there still a bias current, and can I do something about that?
Let's try a thought experiment. You have nothing plugged in. This creates an open circuit.
He has a sensor just laying there "(no input)" This is also an open circuit.
Why would they behave differently?
Beyond that, if you're not giving the myDAQ something to read, why would you expect it to give you any specific output?