Switch controller for SCXI 1128

[chericks1]

Hi holoman0888,

I don't think your current setup is correct:

I recommend connecting the negative terminal of the 140V DC source to OUT+ and the positive screw terminal of channel 3 (CH3+) to the negative terminal of the LC plate:

 

In regards to the 1 ms pulses, the NI-SWITCH Soft Front Panel (SFP) is meant to be user-interactive.  There are user interface delays associated with the SFP that will not allow you to achieve timing accurate to 1 ms.  You will have to do some software development (whether it be LabVIEW or something similar) to meet any requirement of automated timing. 

I recommend creating a simple scan list and sending a digital trigger to the SCXI-1128 each time you need the switch to actuate.

Note: You would need to create the trigger signal externally.  As you require 1 ms pulses, the trigger signal you provide would be a pulse train.  Each rising edge (or falling edge, as you choose in the program) would execute an entry in the scan list. 

For example, to connect an SCXI-1128 switch channel for 1 ms, simply have two scan list entries: the first where ch3->com0, and the second where some other channel connects to com0 (ch0->com0).  Then, create the pulse train such that the time delay between two rising (or falling) edges is 1 ms.

Let me know if you have further questions!

Chad Erickson

Switch Product Support Engineer

NI - USA

Message Edited by Chad PSE on 07-16-2009 02:34 PM

[holoman0888]

Chad,

Thanks for the diagrams. So if I were to connect it that way, would I leave the topology as 2-wire or change it to something else? About the programming, if I use LabView or LabWindows, could i just use immediate operations commands like niSwitch_connect to do something like you mentioned? Which would be creating an entry for chan3+ to connect to com0+ (to let voltage run through my plate) and also an entry for say chan1+ to com0+ (to stop voltage from running to the plate. Thanks again for the help

holoman0888

[chericks1]

Hi holoman0888,

So if I were to connect it that way, would I leave the topology as 2-wire or change it to something else? 

I believe you could do either: leave it as 2-wire or change it to 1-wire.  I would do the latter, as this may prevent confusion later (in reference to your intent).

could i just use immediate operations commands like niSwitch_connect to do something like you mentioned?

I suggest trying to use immediate mode operations.  If you cannot get the timeliness you require, then you can move to the scan list.

Hope this helps!

Chad Erickson

Switch Product Support Engineer

NI - USA

[holoman0888]

Chad,

I tried the connections the way you suggested and they worked. I used the NI-SWITCH soft front panel to connect channel 3 to common 0 and it worked. The problem is I can't get it to disconnect. Or maybe it does disconnect but the power is still running through the circuit so my LC plate stays the same. After that I tried using scanning. I set the scan list to ch3->com0; ch1->com0, for continuous scanning, on break before make mode, but didn't change any settings on the trigger part. I got an error message saying error 223135 ocurred, which apparently is an internal software error in switches software and it said to contact NI. I had one idea:

* If I don't do scanning for right now i.e. just control the paths, maybe I could set up a "dummy" signal to channel 2 for example. Basically just putting a wire in there connected to ground or something, maybe that'll actually change the connection from ch3 to another channel. 

* Or maybe I need to specify something for the trigger?

Thanks for the help

holoman0888

[chericks1]

Hi holoman0888,

The question for me to answer is, "Why can't I get it to disconnect?"

Looking at the circuit that we've created, when the switch is closed, the LC plate is charged.  When the switch is opened, the LC plate doesn't discharge.  Reason?  There is nowhere for the charge to go.

Before building a better circuit that can charge and discharge the LC plate, you will want to verify the switch can handle the potentially large inrush currents (that occur initially) when the LC plate begins to charge.  The circuit you want looks like the following:

To reduce the inrush current, you can add resistance to the circuit.  Hope this helps!

Chad Erickson

Switch Product Support Engineer

NI - USA

Message Edited by Chad PSE on 07-22-2009 12:05 PM

[holoman0888]

Chad,

I tried hooking up 6, 12, 16 plates to see what happened. Each of these experiments gave the same results. It took about 30 seconds for the plates to go to scattering. If you only hook up one it takes about a minute. I thought connecting more plates would be kind of like adding a resistor (more wire, more resistance) but so far it only half-worked. Do you think that if I use a large enough wire I could imitate the effect of a resistor? Or would it just be easier to use a resistor? We have some resistors lying around here, I'm attaching the link to them (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=AVT200-200-ND. Would that be possible to use? Thanks again for all the help, & sorry for all the questions. Being a mech. eng. major we don't go too deep into learning this stuff. Thanks again

holoman0888