Switch Software management

[sakthi11_4]

Hi,

      I am in the process of developing a System to monitor the components on the Printed circuit boards in oven,It s a basically a failure analysis of the components such as C's,Resistors and Capacitors. 

The following is an overview of my test system 

The test system should monitor the components on the PCB 1 year.

I would like to measure the Resistances and Capacitances using an LCR meter to identify the failure. Failure of the Resistor is identified with a Open circuit and a failure of the Capacitor is identifed with a short circuit/open circuit.I have 420 capacitances to be measured and I would like to automate the measurement using NI Multiplexers.

I have to monitor the first Capacitance (C1) for 3 mins ( which would be connected to the first input of the multiplexer) , followed by second for 3 mins and so on. Once all the 420 capacitances are monitored , the loop should start again from Capacitance C1.  I would like to know how to program the NI 's switches. Could someone please suggest me how to go ahead with this.? Is Labview good enough to automate the switches or should I go ahead with a switch management software ? Can you please explain in detail as am a newbie to Labview ?  Looking forward to your response. Thanks in advance 

[Now_With_Underscores · Accepted solution]

Hello switcher,

Yes, LabVIEW is an ideal programming environment! :). 

I assume you have no NI switch Hardware at present. 

What range of capacitance measurements are you monitoring?

What voltage are the caps at?  It sounds like this is just a test board, but I assume the capacitors are held at some constant DC bias?  (Otherwise, the test data you collect won't represent real world capacitance).

Without any specific input except the number of capacitors, I recommend 5x PXI-2575 multiplexers configured as a 2-wire 98x1 mux.  You could use just 3x PXI-2575 modules in 196x1 1-wire mode if your DUJT capacitance is large enough to overcome the total parasitic capacitance of 420 caps (all the negative leads would be tied to the LCT's minus input), but I recommend the 2-wire approach, since it isolates each cap into a differential measurement.

The PXI-4072 has a capacitance measurement function. 

As far as software, I recommend using NI-DAQmx switching API, but you could also use the NI-Switch or NI Switch Executive APIs as well.  However, I think NI Switch Executive is overkill for this application and NI-Switch will require a unique session to each switch module (NI-Switch is IVI compliant, and IVI has no concept of tying multiple modules together).

...ask away if you have questions.  I dont' expect I answered them all in this first reply.

[sakthi11_4]

Hi John

              Thanks for your response. The capacitor that needs to be measured are 2 nF and they are held under constant dc bias. I am just thinking about the one wire mode using Multiplexers. Based on my research I have understood that there must be synchronization between multiplexers and DMM.  Can you please throw some light on the softwares I need for Programming the multiplexers. Labview + NI switch ??? I would like to have an automated switching after a particular time.? Please let me know how this could be achieved .? 

[Now_With_Underscores]

Hello there, 2nF approaches the territory where we need to consider the parasitic capacitance in the cabling/modules/etc.  For example, the multi-conductor LFH-200 cable that plugs into the PXI-2575 specifies 25pF/ft nominal parasitic capacitance with a single-ended signal, versus 15pF/ft nominal with a 2-wire signal. There's also internal capacitance from the switch, your test harness (anything that plugs into the LFH-200 cabling), etc.  With a 1-wire system, all DUT capacitor minus terminals are connected together all the time, which means you have a bunch of additional coupling locations for additional stray capacitance... wherever you have a stray path from any of the minus leads back to the particular channel under test.  For example, you could have coupling back through each capacitor's power supply back through the DUT's power supply and then into your measurement, etc.  With a 2-wire system, both the plus and minus are switched, which greatly reduces the number of venues for stray parasitic capacitance.  You might be able to get away with a 1-wire system, but my concern is with 420 channels your parasitic capacitance could approach your DUT capacitance.  If the parasitics are constant, you could compensate out the error, but I still recommend a 2-wire solution. 

NI HW definitely allows synchronization between DMM/Switch using external triggering, but in your case I recommend just using SW API calls: "switch: connect CHn" "dmm: take measurement y times" "switch: connect CHn+1" etc.  Synchronization is a method to measure as fast as possible, whereas you're taking measurements over a period of hours with a required period for each DUT... with HW synchronization, you can't implement a "wait t seconds" (unless you used additional external hardware, e.g. 555 timer to add delay, but this isn't practical when there's a perfectly good SW API). 

Note that if you're not comfortable with LabVIEW programming, but do know text-based programming, NI-Switch has a full-featured C++ API.  NI-Switch includes numerous examples in both LabVIEW and C++, and of course the community here can offer tips/hints/advice if you run into coding trouble. 

[sakthi11_4]

Hi John,

          The only constraint I have implementing this system is the number of measurement points. We have to reduce the number of measurement points as much as we can. So,as you suggested the better thing would be to compensate out the error due to parasitic capacitance. I am quite comfortable with text based programming langauge than Labview, I hope that I can implement the software in C++.I have just downloaded the NI switch software manual and am going through it. I ll post here in case if I run into any trouble with the Software development. Thanks again for your valuable inputs.

Regards,

Sakthivel.K 

[DavidDaniella]

Then it could be used in C++, seems to be very exciting.

[Janeswordy]

"I have purchased a used 3750G-24TS switch, it is of the older generation it is slightly larger than the current models as the front fo the switch has some vents.The switch does not have an internal power supply and I am confused as to which power supply this switch needs to make it work." "Secondly I understand this model is not POE, if I purchase a power supply that supports poe can I make this into a poe switch or is there naything i can do to make it into a poe switch?" To solve this problem, you can easy use PoE patchpanel to work this switch as PoE switch. Cisco Catalyst 3750G-24TS switch is an innovative switch that improves LAN operating efficiency by combining industry-leading ease of use and the highest resiliency available for stackable switches. Related Articles: Powerful WS-C3750X-24T-S Catalyst 3750-X Switch Easy-to-use enterprise-class lines of stackable and standalone Cisco WS-C3750X-48P-S switch

[Now_With_Underscores]

Hey there,

Your questions regard an ethernet switch.  This forum addresses NI-specific general purpose signal routing boards.  Neither NI nor our forum members will be able to assist you with these questions.  Have a great day!