05-24-2012 01:15 PM
Attached the requested files:
VI with your requested adaptations
scope screenshot Hardware timed simultaneously updated io using while loops 1; Boolean array 11110000..
scope screenshot Hardware timed simultaneously updated io using while loops 2; Boolean array 01111000..
The behavior is comparable with my previous post, only this VI keeps running without the errors, after a while it get worse. see next post with picture.
05-24-2012 01:17 PM
Scope screenshot Hardware timed simultaneously updated io using while loops 3; Boolean array 01111000.. running about 5 minutes.
05-24-2012 02:10 PM
Thanks for your input. I'll try to answer your questions.
1. I've tried polling before, which stone didn't I turn. Result: no more CPU use but results are worse and after only 10 loops it will fail.
2. The counter task is only used to establish hardware-based timing. It is not clear to me what could be gained by counting edges of the internal 20 MHz clock, but I'll give it a try.
3. This one is new and I tried it. Doesn't make any difference.
05-24-2012 06:45 PM
I've tried the example for the 6133 using correlated DIO. Works flawless up to much higher frequencies (tried 50k) and within timed loops. This is the behavior I expect from the 6602. Because this example use correlated DIO I can't use this with 32 ports on the PXI 6508/6509? How do we proceed?
05-25-2012 10:09 AM
For this there are several options: M-series cards, X-series cards or other S-series cards.
In some cases you might have to change ports to U32 inputs.
The escalation is ongoing now.
Can you also provide me with a MAX Report for The RT PXI System and the Windows System?
This can be created the following way:
05-25-2012 02:34 PM
05-29-2012 04:49 AM
I've sent you an e-mail towards your personal e-mail address that contains some extra information regarding this issue and some contact persons.
05-29-2012 06:21 AM - edited 05-29-2012 06:23 AM
As discussed in the e-mail and on the phone I'm also posting some extra information over here.
The issue you're encountering seems to be related to a limitation regarding the PXI-6602 card.
This issue can be caused by the (relatively) small buffer size on this card, together with other devices that use the PXI/PCI bus and the CPU Performance.
For this specific application it is more suitable to use a DAQ-card that supports Hardware Timed DIO (with enough Digital Outouts).
This will also simplify the code you have to use and improve latencies by just being able to use one card.
Some solutions could be:
- M-series cards (62xx) by using Correlated DIO
- X-series cards (63xx) by using Hardware-Timed DIO