FTDI VCP driver NI-DAQmx driver problem...

[abraham@spinalmodulation.com]

Hi,

I am making use of FTDI VCP driver on Windows XP to talk to a FT245R device (through USB interface).

I also make use of NI USB-6251card to capture signal data [pulse wave form] and make use of NI-DAQmx driver S/W to talk to the device connected to the NI card.

I observed that when there's no read/write happening on the FTDI VCP driver,  DAQmxReadAnalogF64() returns the right data. But where there's a read/write happening on the FTDI VCP driver,  DAQmxReadAnalogF64 returns corrupt data.

This happens even if the read/writes to FTDI VCP driver and the  DAQmxReadAnalogF64 to the NI-DAQmx driver are in two different processes.

Could you please let me know if there's any workaround?

Thanks a lot!

-Abraham

 

[Ali M]

Hello,

By corrupt data, did you mean e.g. wrong data points? unreadable data? please clarify. A good starting point would be to launch Measurement & Automation Explorer (MAX) and observe the pulse waveform input of your NI USB-6251 during the time this behavior occurs. If this data is correct, I would recommend reviewing the code for appropriate changes. You can observe the data in MAX by navigating to Devices and Interfaces»USB-6251»Right click and select Test Panels. Set up the Mode e.g continuous; Rate e.g. 100 Samples at 1000Hz. Alternatively: Devices and Interfaces»USB-6251»Right click and select Create Task»Acquire Signals»Analog Input»Voltage»Select the physical input channel, configure the Task and start your acquisition.  

Thank you.

Regards,

Ali M

Applications Engineer

National Instruments

[abraham@spinalmodulation.com]

Hi Ali,

Thanks for your response.

By Corrupt data, I mean wrong data points. So if I request 100 data point, I will get the data count as 100, but the values are junk values.

I took your suggestion:

I selected the test panels and aquired samples. When the FTDI VCP driver was not being exercies, I can see the correct wavefor [regular bi-phasic pulse] in the amplitude vs samples graph. But when there's activity on the FTDI VCP driver (read/write), then the graph shows junk waveform (most of it is close to zero).

The Systems is a dual-core CPU.

Initially we though it was due to CPU bandwidth, but looks like CPU is used some 60%.

A collegue of mine suggested that it could be that when both the USB [NI and FTDI] draw power at the same time, we could get junk values.

I also tried a USB hub and did not see any change - same bad values.

I really appreciate your response.

Thanks!

-Abraham

[Ali M]

Hello Abraham,

To investigate the DAQmx calls further during the Read/Write of the FTDI VCP driver, post an NI Spy capture image of this behavior. NI Spy Can be found by navigating to Start»Programs»National Instruments»NI Spy. Run the program and observe the DAQmx calls during this behavior (post a screenshot of this information).

Thank you,

Regards

Ali M

Applications Engineeer

National Instruments

Message Edited by Ali M on 03-10-2010 10:03 AM

Message Edited by Ali M on 03-10-2010 10:04 AM