NI 9237 sampling rate

[adrian k]

I am using cDAQ-9172 chassis with three different modules: NI 9237 strain gauge, NI 9211 thermocouple and NI 9205 analogue input.

I have written a simple VI in LV 8.0 using DAQassistant and Express VI 'write to measurement file'.

When sampling only thermocouple and analogue channels everything works fine, I can choose a desired sampling rate and no. of samples. When sampling a strain gauge channel through the NI 9237 the rate appears to be fixed at 3846.154 Hz. This is independent of the number of channels being sampled or any set-up in DAQassistant.

The application note mentions a maximum sampling frequency for this module of 50,000/no. of channels. The sampling rate I get is exactly 50,000/13, but does not change with the no. of channels.

Please could you tell me how I can sample at different rates with this device (3846.154 Hz isn't a lot of use for my application). I appreciate there will be a maximum sampling rate for this module but there must be a way of sampling at slower rates than this?

Thanks,

Adrian

[Salvador_Santolucito]

Hi Adrian,

The manual for the 9237 contains an equation that describes the available sampling rates:

f = 50,000 / n     n = 1, 2, 3 ... 13

So the minimum rate you can achieve would be 50,000 / 13 = 3846.154 Hz.

Note: There is a typo in the manual.  The manual says "n = 1, 2, 3... 31."  The maximum value for n is actually 13.  This will be fixed in the next revision of the manual.

Thanks,

Sal

[Salvador_Santolucito]

Hi Adrian,

I just spoke with some of the hardware developers of the 9237 module.  Although the current version of the driver (NI-DAQmx 8.3) limits you to a maximum divisor of 13, the hardware can actually handle up to 31 as stated in the manual (giving you a lower achievable sampling rate).  This behavior (n = 1 to 31) is currently supported in cRIO.  In a future version, the NI-DAQmx driver should also support divisors of 1 to 31.

Thanks,

Sal

[Jeremy@Thermadyne]

So there is no way to reduce the sampling rate any lower? 

 

Also, I am using the DAQ assistant and it appears that the data I am receiving is not at a constant rate.  The data is being output to a tab delimited text file and sometimes it will output only a few samples a second and others it will output 300.  I am not so concerned about what the rate is, I would just like the output to be recorded at a consistent rate.  When I do use only the NI 9211, it outputs consistently at the rate I set in the DAQ Assistant.

[Salvador_Santolucito]

Hi Bill,

At this time you will not be able to sample any slower.  This issue should be fixed in the future however.

Since you have set up a hardware timed continuous acquisition with the DAQ Assistant, you will be sampling at a constant rate.  The number of samples the DAQ Assistant returns in each iteration is determined by the "Number of Samples" input.  When you create the DAQ Assistant, there is nothing wired to this terminal so it uses the default value of -1 (read however many samples are available).  Using this typically results getting a variable amount of samples in each loop iteration.  To get the same number of samples each time, wire a constant (i.e. 1000) to that terminal.

Thanks,

Sal

[dbtestcon]

---

I just spoke with some of the hardware developers of the 9237 module.  Although the current version of the driver (NI-DAQmx 8.3) limits you to a maximum divisor of 13, the hardware can actually handle up to 31 as stated in the manual (giving you a lower achievable sampling rate).  This behavior (n = 1 to 31) is currently supported in cRIO.  In a future version, the NI-DAQmx driver should also support divisors of 1 to 31.

---

Can someone tell me if this if fixed in DAQmx 8.5?

[Brad_K]

Hi dbtestcon,

Yes, the NI 9237's minimum sample rate is fixed in DAQmx 8.5.

Brad