Analog input task with special sample sequence

[VladimirZav]

We use Pcie-6323 card. I suppose it it a X series.

[santo_13]

Yes, it is an X-Series card, which means, the dummy channel technique will be on your side.

So, you would have to create 2 additional virtual channels for each of the real channels, then while creating the task, group the virtual channels together with the original channel.

[Kevin_Price]

Santhosh,

Do you have systems where you can test out the duplicate channel names?  Here's what I got when I added the same channel twice to a shipping example:

real PXIe-6361, DAQmx 16.0 : no error 

real PCIe-6321, DAQmx 20.1 :  error 

simulated PCIe-6361, DAQmx 20.1 : error

-Kevin P

[santo_13]

I tried with actual hardware (PXIe6368), software LV2020, DAQmx20.7

Tried creating separate virtual channels in NI MAX and it did not work. Looks like that feature is obsolete.

[Kevin_Price]

Hmmmm.  Strange.

I *do* hear you with the "obsolete" declaration, but feel further compelled to ask, "WHY?"

The X-series hardware (and IIRC, other generations of hw before X) are *capable* of supporting this.   So WHY was this capability TAKEN AWAY?

-----------------------------------

Dear NI:  Knock it OFF!  As far as I'm concerned, EVERYTHING you've done in the changeover of the last couple years makes you WORSE!!!!   I just sat down and gave serious thought to the matter and you know what?  I cannot identify a single positive change in that time frame.

Hey!   I know.  I'm too old to be part of your target demographic.  You do you.  And I'll do me.  And me doing me looks like this:

Actively removing support for the ability to sample the same channel multiple times was a STUPID and unnecessary decision.  The hardware can (obviously, based on actual test runs with older drivers) handle it, WHY (literally WHY WHY WHY) would you commit development effort into REMOVING such support?!?!?!

Tech wise, I'm definitely an old guy now.  I have almost 10x as many working years behind me as I have in front of me.  I used to think I might freelance a bit in my golden years but the new subscription policy dictates otherwise.  I've been a fan and advocate of NI since before Y2K, but not so much now BECAUSE OF these last couple years.

Soon enough I'll be out of the picture.  So, you know, I *do* realize you (NI) don't have to care all that much.   It's been a mostly nice run these last 20+ years with LabVIEW.  Too bad you've decided that NEW users are so much more valuable than us long-time EXISTING users (and advocates).

- Kevin P

[VladimirZav]

Hello,

It looks that first idea with three samples on each channel will not work.

Then I will try to to scan all sixteen channels three times and calculate average values. Maybe it will improve measurement.

Is there any example (in C)  how to do in this?

On one rising edge on PFI0 task will make three scans of sixteen channels.?

Thank You for  advice

[Kevin_Price]

Back again and in less of a mood this time.

I can't offer any help with C syntax as I do all my DAQ programming with LabVIEW.  But I can give you an outline for what to do.

You'll need to pair up a counter output task with your AI task.   The counter will be a finite pulse train of 3 pulses.  It will trigger off a rising edge at PFI0 and be configured to be retriggerable.  You should also choose a different PFI pin as its output terminal.

The AI task will be configured to use the counter output pulse signal as its sample clock.  All 16 channels will be in the task.  You'll be free to call it a finite or a continuous sampling task -- either way it's the counter task that's *really* in charge of sampling.

You'll have to plan out the timing to be compatible.  But you may find yourself unable to meet all the desired constraints.  The more settling time you want between channels, the slower your max sample rate can be, and thus the longer the time before you can be ready for the next PFI0 trigger.

Honestly, it may even be worthwhile to *downgrade* your LabVIEW version so you can use DAQmx 16.0 which supports the use of duplicate channels in tasks.  (In the meantime I've found that DAQmx 18.1 gives errors with an X-series board.)

-Kevin P