NI USB 6009 Low Voltage Meaurement Accuracy

[Laddee]

Hello everyone,

I am new to this forum but I have been using LABview in one of my classes. Currently I am involved in a project about Image Mapping and we are measuring the voltage from the photodiode usually around microvolts to millivolts. I tried to use the USB 6009 but I think it is not accurately reading the voltage aside from not giving stable reading. I have compared it to a Keithley 2000 Multimeter and this is the result. USB 6009 = 5.365-5.367 mV DC while Keithley 2000 = 4.532 -4.533 mV DC. This is a big difference. I am using differential connection where I connect the positive photodiode to AI0 and the negative photodiode to AI4, thus using AI0 physical channel in LABview. Could this be because there is a limit to the USB 6009 measurement capacity? and I need to use more higher/powerful NI USB? 

Thanks a lot for helping out.

[crossrulz]

For issues like this, we go read the device specification.  So assuming you are using the 1V range, your typical accuracy is 1.53mV.  If you are only reading the one voltage, I would just use your Keithley 2000 DMM.  I think it has a serial port interface you can use and a pretty simple command structure (it does use SCPI) to get your readings.

[Laddee]

Hello Sir,

Thank you so much for your reply. As of the moment I am using Keithley because it is much more accurate than USB 6009 and it is much stable however we need faster data acquisition and currently with Keithley I can get a 100x100 pixel voltage data in 2hrs. With USB 6009, I can have it to get the same data in 15 minutes but due to the inaccuracy and unstable data I cannot use it. Is it okay if I use USB 6210 with a typical accuracy of 88 μV at ± 0.2V?  I really appreciate your response for this. Thanks.

[santo_13 · Accepted solution]

Between Keithley DMM and USB-6009 or USB-6210, you will be compromising between speed and accuracy.

USB-6210 is nowhere comparable with a 6.5digit DMM in terms of accuracy. It is expected that measurements need more averages to get better accuracy and hence it will be slow.

Now, you could approach this problem in two ways,

• Analyze the scope for improvement in the Keithley 2000 automation and get the best accuracy and moderate speed
• Develop new automation with USB-6210 to get a moderate accuracy and better speed

[Laddee]

Hi Mr. Santo, thank you for this great suggestion. Sorry for the late reply, as I have been not well for a couple of weeks. I have talked with my colleague and he said I forgot to mention about the 3D Stage Actuator that also affects the speed of data acquisition.

I may have not explained clearly how the Image Mapping of our system works, but it involves x-axis and y-axis actuators (Z825B Thorlabs) with their KDC 101 motor controller to move from point to point in a scan area of a few millimeters. As I have said previously that with Keithley's speed in acquiring 100x100 pixels voltage data is at 2 hours while NI USB 6009 is at 30 minutes using 1 sample on demand acquisition mode (tried it again recently). We may have neglected that it takes also some time from moving to different points of the scanned 100x100 pixels. With this realization, do you think if we opt to change USB 6009 (48kS/s) to USB 6210 (250kS/s) using 1 sample on demand acquisition mode or with your approach, will it be much faster? or any NI USB Multifunction device despite having higher sampling rate will not affect the speed of acquisition because of the x and y axis actuators movements?

[santo_13]

Still, a lot of information missing in your description of the setup, is there only one signal you measure or is there a multiplexer involved?

How do you synchronize both Z825B after it has made movement and your instrument measurement?

I would still prefer to use the Keithley DMM with fine-tuned settings to acquire at a faster rate than using a DAQ, you tweak the aperture time, auto-zero, settling delay, add triggering, use onboard memory etc.,