unusual delay between cards

[ipalfrey]

Hi.
I am running a field logging system using a PXI Chassis fitted with a PXI 6052E and a PXI 6071E card. We are logging at 4kHz on 8 channels on the 6052 card and 28 channels on the 6071 card, and it has been noticed that there is an unusual delay between the channels on the 2 cards. I have since tested this in the lab and have found that the delay is 515 samples or 0.12875 seconds. Have you any idea what would account for this delay? This seems to be much higher than would be accounted for regarding interchannel delay.

Many Thanks in advance

Ian Palfrey

[PNHorn]

Hi Ian,

What version of NI-DAQ are you using? Depending on what version you are using you will get different interchannel delays on the different cards.

First of all, if you are using NI-DAQ 7.3 or earlier, then it bases the interchannel delay on giving the most amount of settling time per channel. So, lets take your sampling rates and number of channels, on the PXI-6052E card, you are running at 4 KHz over 8 channels and you will have the 8 channels equally spaced over the 250 us giving each channel an interchannel delay of roughly 31.25 us. On the PXI-6071E card you will have the 28 channels equally spaced over the 250 us. This will give you a interchannel delay of about 8.29 us per channel.

If you are using NI-DAQ 7.4, then it will base the interchannel delays on getting the sampling per device done as quickly as possible. So, for the PXI-6052E card it will do the 8 channels as quickly as possible at the start of the 250 us (probably in about 20 us) and then it will have a delay of maybe 230 us. The PXI-6071E should sample the 28 channels in maybe 40 us and then have a delay of 210 us.

Neither of these explainations really cover the delay you seem to be getting of over 0.1 seconds, so could you possibly send me a bit more detail about exactly how you derived this figure and what two points you are measuring between. Hopefully then I can provide you with a bit more useful information about where this delay could be coming from.

Best regards,

Peter H.
Applications Engineer.
NIUK.

[ipalfrey]

Many thanks for the response.

On that system I am using Labview 6.1.
I tested the delay by applying a square wave to individual channels on each of the cards, and then comparing the results to determine the delay. I had calculated the delays due to settling times, and was therefore confused by the size of delay. Hope this helps

Regards

Ian Palfrey
System Engineer

[PNHorn]

Hi Ian,

Sorry I didn't mean the version of LabVIEW, the version of DAQ is more important. You can find this out by opening Measurement & Automation Explorer then looking under Software>>NI DAQ.

Best regards,

Peter H.
Applications Engineer.
NIUK.

[ipalfrey]

Hi Peter
Sorry for the delay before replying. Out in the field. The version of NIDAQ is 6.8.0f9

Hope this helps

Application Engineer
University of Plymouth

[SachaE]

Hi,
are you talking about the delay between say ch0 on card 1 and ch0 on card 2?
How is your code starting the cards off?
Are you using the trigger lines to synchronise the scan's i.e. card 1 outputting it's scan clock
and card 2 using that since they're running at the same rates 4kHz.
Can you post your simple example code?

Thanks

Thanks
Sacha Emery
National Instruments (UK)