NI 9215 phase delay between two channels

[ASakhno]

Hello,
I use Ni9215 with ENET-9163 to measure phase delay between two differential sine signals. First signal connected to AI0 and second to AI4 at sample rate 100k. I know that 9215 has simultaneous ADC, but it seems to me that NI 9215 gives additional delay between channels. Is it possible, or l am wrong?

Alexandr.

[Hillman]

Hi Alexandr,

Good afternoon and I hope your well today.

Thanks for the post.

You are correct about the NI 9215 - you shouldn't see a delay between channels readings - so are you reading the same signal into both channels, also try to get the wires the same length, as it will take a finite amount of time for the signal to travel from the sensor to the 9215. ( NI 9215 has a group delay of 0 samples(not a delta sigma ADC)).

If its not the same signal, how do we know the phase delay isn't in the two signals? For example Connecting a Floating Differential Voltage Signal to the
NI 9215 with Screw Terminal requires a bias resistor.

User Manual

http://www.ni.com/pdf/manuals/373779e.pdf

You can account for the delay - if you haven't got the time to find the cause of the delay. 

Something to keep in mind is that the external circuitry can introduce additional skew. In what sort of region is your phase delay? - I'd have to check it up but I think you'd be looking at ns for interchannel delay. 

Are you using code to generate the external signals? Maybe the labVIEW code is generating a phase delay..

Please let me know your thoughts, and I'd be more than happy to discuss this issue with you. 

Hope this finds you well, 

I would also make sure you've input circuitary is correct as per the manual,  

[ASakhno]

Hi James.
Thank You? for Your reply.
Thanks for the post.


I'am reading not the same Grounded Differential Signal into both channels. I have the long wires with different lengths (60 - 200 meters) , an You are right, that I'am looking at very small delay 0.0001-0.001 ms. I don't use labVIEW to generate signals. This is real signal from sensors. I think You are write about wire length. So if there is no source of delay in the 9215 module, I think source is in wire length. Do You have more ideas?
My scheme in the attachments.

Best Regards,
Alexandr.

[Hillman]

Hi Alexandr,

Thanks for the reply.

Your setup for the C Series module looks correct - which is good. 

I would suggest cable length being the main cause - you can add a filter to increase a delay in the shorter cable, here is a knowledgebase article which mentions this... it is more with adding a group delay for modules with different specifications, but same sort of thing.

How to Compensate for Different Group Delays with C Series Modules in LabVIEW FPGA 

http://digital.ni.com/public.nsf/allkb/74EB238E1BCADD528625735300681A7D?OpenDocument

Otherwise, I would maybe consider altering the phase in software. 

Note: The delay of around 1ns from what I have read is pretty good - but there is no group delay with the module, so it must be your setup I am afraid.

Sorry I didn't have any other thoughts from the cable length.

Hope this helps, 

[ChrisR.]

The different channels in the 9215 will have slightly different delays, since they have slightly different bandwidths.  I would expect variation in the 10's of nanoseconds, but it could be more.  However, 200m of cable would introduce delay on the order of 1 usec, so I suspect that's your primary culprit.

You could try connecting the exact same signal to your two channels and measure the phase delay between them to see how much of the error is due to the module.

Chris 

[Hillman · Accepted solution]

Hi Alexandr,

Please let us know how you get on,

Chris is correct - with the idea of using the exact signal for two channels, I was trying to suggest this earlier on.. 

I hope this helps resolve your issues with the delay, 

[ASakhno]

Hi All.

I can't try to connect same signal to the two channels, because it is already working system. But I will try to calculate delay, and use it in program.

Thanks for all.

Alexandr.