Here is an idea for a data acquisition system that I would like to construct for a research laboratory: any ideas on whether or not it will work?
I am using ion-selective electrodes, devices that generate millivolt signals. Accuracy needs to be within +/- 0.01mV. The electrodes attach to BNC connectors and are measured vs a reference electrode.
I planned to attach 8 ion-selective electrodes and a common reference electrode to the BNC-2110 connector block. From there I would use a cable to attach to a NI 6713 board and use LabView to collect the data.
My main questions are: Does this setup have any obvious flaws or problems that I am overlooking? Can this setup support low mV signals?
I am not sure about the NI-6713 that you mentioned since that is the name for an analog output board. But that's fine! Whether that is a typo or you intended some other product, I have a few ideas that you could use.
Two products that you can look at are the NI-4060 (High Resolution DMM) in combination with an SCXI-Switch module, and an E-series data acquisition board in combination with an SCXI signal conditionaing module.
The first option is the NI-4060. It is a high resolution DMM so it has one channel of measurement. The reason for having one channel is that the NI-4060 is intended for use with a switch that can provide very configurable and expandable multiplexing options. You would probably use the SCXI-1127 or SCXI-1128. With one one of these modules , you can switch or multiplex between 32 differential channels.
For a 20mV range, you will get an accuracy that in short, is about +/-8 microvolts. For a better description, you can look at the user manual for the NI-4060.
Another option is a high resolution e-series data acquisition board (for instance the PCI-6052E) in conjunction with a millivolt signal conditioning module (for instance the SCXI-1102B). The combination allows you to use a standard data acquisition board, lowpass filter your signals, and apply a gain so that the signals are not as susceptible to noise.
For the software, I think LabVIEW is a great choice. If you prefer text-based programming, you could go with Measurement Studio.
I hope these give you a starting point for finding what you want. Other people might have different suggestions for you, and I'll try to answer questions if you want to reply to this post.
If you want help configuring a sy stem, one of our internal sales folks might be able to help you even more with picking parts.
Hi again. I forgot to address the BNC connectors that you have from your sensors. You could use a BNC-2095 that provides lots of connection options for the second option (e-series with signal conditioning).
However, since these are such small signals that are susceptible to noise, I would suggest connecting the BNC's on the sensor side, running the cable as far as you can to the equipment, and then breaking out the wires to connect directly to the differential channel inputs for either option.
Once again, the sales department at NI could probably help you the most with cabling and parts configuration.
Hi -- Thanks for the suggestions! I'm working with an academic budget and need to keep things fairly cheap. Do you think this might work? A PCI-6052E board attached to a BNC-2110 connector block? Is the signal conditioning module entirely necessary? Thanks-Erich
Hi EDS, pretty little voltages are these, which you want to handle. Are you aware about the Seebeck effect? This is the effect, which gives thermocouple sensors their output voltages. For instance a connection of Fe and cu-ni gives you a voltage of 54?V per ?C of temperature difference between the connection point of the different metals (at the point of connection from your electrodes to the wire feeding your DAQ-card)and the connectors of your DAQ card. There are different methods to meet this problem. I myself often handle with voltages like this, using thermocouples and strain-gages. You will get pretty messy voltage measurements if a part of your connection changes temperature by airlow, if somebody opens the door or a window, even if you get near to your constr uction (without touching) and the temperature of your body changes the temperature of your connections.
Answer me, if you want more information. Above I simplified a bit, but I could describe more precisely, if you want me to.