Is it possible to simultaneously connect several PXI-2567 channels with one software trigger?

[fazekasc]

Hello,

I have a PXI-2567 relay driver card, and I'd like to know if I can set it up to simultaneously connect 8 channels with one trigger, instead of sending 8 commands to close the 8 channels one after the next.

[Philosophe]

Hey fazekasc,

Which LabVIEW API are you looking to utilize? Depending on whether you are using DAQmx or NI-Switch, there are different answers for both. In NI-Switch you cannot do this, you have to close each switch separately. In DAQmx, you can simultaneously close multiple switches by channel expansion but I am not sure about triggering each switch closed part. How are you currently triggering the switches to close?

[marshallg]

Hey fazekasc,

It sounds like you are using scanning and not immediate mode operations.  It should possible to connect multiple channels on the PXI-2567 with a single software trigger using either DAQmx or NI-Switch.

NI-Switch Scanlist documentation: http://zone.ni.com/reference/en-XX/help/375472G-01/switch/scan_list/

DAQmx Scanlist documentation: http://zone.ni.com/reference/en-XX/help/370466AA-01/mxcncpts/scansyntax/

In your scanlist you will probably end up using a single '&' between the simultaneous channel connections.  For example, an very short NI-Switch scanlist for a 2567 which connects ch0->com0 and ch1->com1 with a single software trigger, and then ch2->com2 and ch3->com3 on the next software trigger would look something like this:

ch0 -> com0 & ch1 -> com1; ch2 -> com2 & ch3 -> com3;

Take a look at the documentation, theres alot you can do with a scanlist depending on your needs.

PS - Rob - Starting with NI-Switch 4.8 there is an immediate mode connect multiple which allows you to connect multiple channels on a single switch at a time.

Marshall G.

National Instruments

[fazekasc]

Thank for the quick responses.  In this scan list example "ch0 -> com0 & ch1 -> com1" when the software trigger is sent, will ch0 and ch1 physically close in the HW on the PXI card at the same time, or will it close ch0 first, then close ch1?

[WillPwr]

Hello, Kazekasc!

With the scanlist entry "ch0 -> com0 & ch1 -> com1," com0 is not guaranteed to close before com1, nor is com1 guaranteed to close before com0.

Both connections will essentiually be made "at the same time," but the order is not deterministic. One could close slightly (in terms of <1ms) before or after the other. As mentioned in the above-linked help, "...the string ch0->com0 & ch9->com1 means connect CH0 to COM0 and CH9 to COM1 (in no particular order and with minimal delay)." 

Does this answer your question?

[fazekasc]

Yes,

That's exactly what I needed to know, but was hoping there was a deterministic way of doing it with timing to a few us.

Less than 1ms will work for our needs, so we should be good to go. 

[fazekasc]

More info to follow up.  With our PXI-2567 card and some dummy resistive loads less than 100mA each, we were able to turn on 4 channels within 10us of each other... well below the 1ms posted earlier in this discussion.

We used the software trigger with a scan list that closed the 4 channels using the "&" symbol in the scan list... using NI's LabVIEW niSwitch example out of the box.

[KovAcevich7]

So you were able to close 4 channels within 10us? This sounds exactly like what I need to use then! I am trying to turn on 4 channels, sequentially, all within 5 ms. 

I have not read the documentation quite yet (I will be doing that immediately after I finish typing this) but this sounds like I should be able to use the scanning feature of NI-Switch to connect my relays, sequentially in less than 5ms of each other, correct?

[fazekasc]

Hello,

Yes, it was possible to close four driver channels all within 10us of each other with the dummy resistance loads we used instead of actual relay coils.  So there's a few things to consider when implementing your solution;

The load, IE relay coil, resistor, etc, may affect that turn on time.

The four channels I closed were not sequential, they all closed at their own pace regardless of channel number, or one closing before the next.  The trick was all in the command sequence line sent to the card, it basically sends the command line containing all four channels you want to close... then with a software trigger, enables that command to execute at the board level.  Each internal driver circuit of the PXI card will have some of its' own characteristics as far as delay, and execution time, but as far as the PXI low level firmware is concerned it sent out four channels simultaneously.  But we all know nothing is simultaneous.  What I'm worried about in your setup is that you're mentioning 4 channel sequentially, and that's not what I was able to do.  My application was to close 4 channels, randomly, as quickly as possible.

If you need to close the channels sequentially, within 5ms, that may not be possible.  If you don't care what order they close in, then you can use my solution.

I'll have to dig up that code and post it here...

[KovAcevich7]

Yeah i need to close 1 channel, close 2 channels "at the same time" and close a forth channel within 5ms. 

How were you able to know that you had closed them within 10us?