fpga sine generator jitter

[vakost]

Hello.

My setup is the cRIO 9082 with NI 9264 and I want to generate a variable frequency sinewave.

I am using the sinewave generator from the LabVIEW FPGA module.

The problem is that I am having a large jitter.

I had the same problem when I was working with the same module (9462) in a DAQ chassis and the solution was the setting of the "Automatic" and "Integer number of cycles" checkboxes in the "Configure Simulate Signal" dialog box of the Simulate Signal express vi.

Do you know of any adjastment like that in the FPGA sinewave generator so that I will not have any jitter?

I know that I can write my own DDS but if there is a solution it will save me a lot of time.

Thanks a lot.

Vasileios.

[vakost]

I am also attaching the project.

I am passing the frequency through the normalize function and I do see that the actual frequency is the input to the sine generator.

I still get a jitter.

Thanks.

Vasileios.

[vakost]

So, I verified that there is a jitter, a software glitch.

I am sending the generated data from the sine generator to the host application through a FIFO before I send them to the AO.

I then plot the data and export the image. These are a couple of images. The problem occures in the 1023 point every time. The size of the sine generator lookup table is 32768.

Any help???

Vasileios.

[srm@viewpointusa.com]

i don't see anything obviously wrong with your code.

since you did not post your other experiment with sending the data to the host, i can't comment on that.

what frequency were you trying to generate?

what jitter did you measure and how did you measure it.

the module referenced in project is a 9264 which has a maximum update rate of 25 kHz.  this should result in values for "One sample tick" of about 1600.

with this update rate, your sine wave will look pretty crummy at anything over 1 kHz.  and even at that rate, you will see significant "jitter" as compared to a function generator.  at 10 kHz, it is next to useless.  if you observe the output on a scope, you will see the quantization in the waveform.

100 Hz would look ok.

Even the NI 9263 will not produce very good sinewaves at 10 kHz (10 pts per wave)

[vakost]

Hello.

I am attaching the latest vi.

I understand what you say about the module maximum update rate. Although the frequency generated, as you will see in the attached vi, is not bad, even at high rates (I get quite good results at 3 kHz), except from the glitch point.

Any help would be great!

Thanks,

Vasileios.

[vakost]

I see now that the FIFO size is 1023. So I think that the glitch is not due to the sine generator.

But then, why does my oscilloscope shows me a bad signal?

[srm@viewpointusa.com]

attach an image of what you see on the scope.