I'm having very bad latencies and results when I'm sampling from 2100Hz to 4000Hz with my DAQPad6020E, no matter what the buffersize of function type is (LabWindos/CVI)

[TRG]

I'm using the DAQPad6020E (usb) for my data acquisition, and latency is an important issue. It seems al samplespeeds are working fine when I'm using the double buffer methode or when I'm working with my own buffer wich uses the event message function.

Now I have noticed that at 2100 Hz the latencies are extremely bad (>700ms), and it is getting better when the samplespeed is going up to 11KHz (<200 ms). But this is strange because at other samplespeeds I'm having latency below 100 ms.

What is te problem here? Why am I having bad latencies from 2100 Hz to 11 KHz?????
I have uses all kinds of combi
nations from buffersizes to different DAQ functions. And always bad results especially at 2100 Hz!!

I'm using LabWindows/CVI 5.5 on WindowsXP

[RikPrins]

Hello,

These boards use rather large on-board FIFO buffers. When you do a continuous Analog Input transfer you can get these kinds of latencies at the relative lower speeds. This is due to the large buffer. What is happening: The buffer on the board is designed in a way that it only transfers its data collected to the Data Acquisition Buffer on the PC when it is half full. At the lower speeds this can be resulting in these kinds of latencies.

Best Regards,

RikP
Application Engineering
National Instruments

[TRG]

I understand the limitations of the FIFO buffer according to the samplespeed. But the strange thing in my situation is that at samplespeed below 2100 Hz (from 100 till 2000) I get great results (>100 ms latency).
The difference between the 2000 hz and 2100 hz results is dramatic. I really can't figure this out.