count loss

[Davis1900]

The quadruature signal goes to a Windows-based PC and Linux Based PC. Our amplifier provide these signals to Windows  and Linux PCs.

The Windows PC is used to generate a squre wave with low at 0 count and high at 1000 counts with frequency at 1Hz. The Windows PC can measure the motor shaft positiion and also the shaft position is monitored by a fixed position mark. The mark proofs that it the signal to Windows PC is very well and the PC's count measurement is teh same as the demonded value.

However, the Linux machine measurements are as follows: 0 to 1000 counts, to -5 counts to 995 counts to -12 counts to 880 couts, etc... It seems that some pulses from encoder A/B are not counted by the Linux machine. The hardware we are using are SC-2345 Signal Conditioning Board and PCIe-6259 borad. The software are Scilab with RATAI lab real time support and teh C codes are using in the Scilab for the data aqusition.

What causes the count loss and how can I compensite them?

Thanks,

Davis1900 

[Sandra_T]

Hello Davis1900,

I am not very familar with Scilab with RATAI lab real time support. What I would first do is isolate the hardware from the software programming. What version of the DAQmx driver do you have? Are you using DAQmx for Linux of DAQmx Base for Linux?

If you are using DAQmx for Linux, then you can use test panels to test your hardware. You should then be able to test the counters to see if the values are giving you wierd errors. Refer to the "NI-DAQmx 8.0 for Linux Configuration Guide," configurationGuide.html, for more information. The guide is located in /usr/local/natinst/nidaqmx/docs.

If you have tried this and still see the same issue then we'd have to continue troubleshooting. Let me know what the outcome of this test.

Regards,
  Sandra T.

Applications Engineer | National Instruments