measurement of resistance by USB 6008

[Henrik_Volkers]

The AO of the 6218 can supply 2mA 🙂  (read the spec!)

So if you connect a 10kOhm resistor in series with your impedance to the AO and measure the voltages at the AO and after the 10K you can measure the impedance in a 1k to 100k range...

For higher impedances maybe switch to a 1M resistor.

You should use stable well known resistors with low inductivity .. (Metal foil) if you want to measure with AC.

Since you have a multiplexed AI input, try to measure twice with one channel only (assuming the AO and your DUT is not changing) to avoid crosstalk and settletime errors (read the spec!).  (or use/build buffer amplifier)....

If you use sine exitation try the tone detection vi for amplitude and phase measurements...

3 or 4 wire is only needed if the error get too big..

1Ohm wire in a 200kOhm measurement doesn't make that much 😉

Shielding and leakage on high Ohms measurements is more tricky 😄

Bioimpedance measurements need well defined boundary conditions...  expect nonlinearities  .. 

[vic_grc]

Thank you for your response.

What I understood from specs is that I have voltage range output of +-10 volts and a maximum output current of 2mA. This affect when designing my circuit to make sure that I don`t need more than 2mA for my circuit or I guess I can create an AO current drive with a voltage follower.

I will use a voltage divider to measure resistance. With my values: 200 or 300 Kohm and contact impedance in textiles 1 Mohm, I need to eliminate the 1 MOhm reading from my measurement to get the impedance that actually I need to measure, the one of 200/300 Kohm. That with a 3,4 wire connection I will be able to do it right?

What do you refer when you said "try to measure twice with one channel only"? different measurement can be done on some channel? Actually I need to keep my sample rate as high as possible (250Ks/s divided by number of channels used). I think I need one AO channel for voltage source and one AI channel to measure voltage and then calculate R with voltage divider formula. I need to work with AC. What is the best option to do that?

Greetings from Sweden 

Victor

[Henrik_Volkers]

It's a good idea to build buffers for exitation (well, with your high impedances maybe not needed) and for your input.

If you build buffers , think about adding an active shield driver and get some triax cables and connectors.  your active shield driver has to drive capacitiv loads ... prone to oscillation... search the appnotes from TI, AD and Linear to get some ideas.

What frequency range do you want to use and what uncertainy goals do you have?

[vic_grc]

Hi,

What do you mean by building buffers for my inputs and output? I found this from TI http://www.ti.com/lit/an/snoa926a/snoa926a.pdf about active shield driver. I can´t follow your reasoning, do you have some links or examples VI to understand better what is the idea? For the characterization of the impedance I will be using frequency range between 5KHz and 1 Mhz 

Best regards

Victor 

[Henrik_Volkers]

With 250kSPS don't expect much above 100kHz 😄

• Have a look at Fig2 of the spec of your 6218!
• and you migth have 100pA Biascurrent
• and 100pF input impedance will give a load of ~(-j)16kOhm@100kHz 😉

That's why you migth need a input buffer with a low input capacity. One way to minimize input capacity in your cables is active guarding. 3/4 wire resistance measurement is easy and simple for DC or AC <5kHz (Well, still room for errors 🙂 )

Again: Target uncertaincy??

HP(AG/KS) & Keithley build nice LCR meters try to find out how they do it 🙂

(Hint: Open/Short calibration!)

[vic_grc]

Hi again,

The other part I think I finally fix it. I attach the VI called 4-wire measurement. (suggestions are welcome) The idea is create a Voltage controlled current source with the image I have attached and measure the impedance load (this impedance is divided in  3 impedances in serie: contact impedance 1, load and contact impedance 2), one with simultaneous measurement ( 4 input channels used) and other measurement separately ( 2 AI channels used) so my sample rate is bigger and can reach higher frequency.  Impedance measurement is calculated using digital sine correlation (attached image of the idea) Now I am  trying to create a VI that start with low amplitude sine waveform generation and increase until the value measured in VContactImpedance_1+Vload+Vcontactimoednace2 reach 8 volts. Once reached, it stops there and show the amplitude used to reach that value and the current in Zref. This is used to avoid saturation since my impedance load (total sum) is unknown and  can be maybe of 1 Mohm and need to make sure that voltage across that impedance is maximum 8 V. I have 2 errors showing in the Voltage_increase util stop condition 

Greetings

victor 

[Henrik_Volkers]

You can try it in that way, however, if your impdance is 1kOhm you can't go up to 8V 🙂

The current needed is 8mA and your source can only drive 2mA. Your OP setup doesn't change that. (same current in both resistors)

And if you have programed your signal analyses: please compare the results with the approach to simply capture more than 15 periodes and  do a tone detection.

Remember that you have a multiplexed input. you don't read simoultaniously. that will affect your phase measurement.

If you do the measurement twice with swaped input channels (and always same range) you can compensate that.

[vic_grc]

Thank you for your help. My reference impedance actually I will use a 5,5k. What you mean read twice with swaped input channels?

I have an issue with measurement. Since I have to read contact impedance 1, contacte impedance 2 and load separately and then, have an option to read all simultaneously (each option is available in a case structure), how can I do that? I would like to use the SubVIs that read each separately and place all 3 together but the fact is that I can't use more than one DAQ assistant for measurement at same time right? Is there same way to call this 3 measurements simultaneously with out writing all code together for 3 of them?

I attach two VI, one read contact impedance and other load, its the same code but different input channel thats it.

Best regards and thank you again 

[Henrik_Volkers]

for each channel you read (say a sine signal) you get

U0_read(t)  = û0_input*(Gain_Ch0+err_gainCh0)*sin(w(t+MuxdelayCh0) + Phase + err_phaseCh0)   ( + Offseterr + ...)

if you  measure twice (A and B) and swap two channels and take the geometric mean (u = SQRT(A*B)

some errors cancel out... just do the math 😉 (geometric mean because we look at RATIOS! )

and (after some partial differentiations) you migth find that your standard deviation (and uncertainty) is less compared to a single measurement.  

If you analyse your setup and determin gain and phase compensation once , that will increase to your uncertainty

But maybe that doesn't hurt you because you don't talk about uncertainties 😄

Another hint: the Gigaohms input impedance stated in the spec (for this multiplexed card) ist only valid if you don't multiplex.

[vic_grc]

Why I am getting all the time error that data is not available? I have tried several connections and different AI channels and cables but still showing same error.