SMU isolation

[azstreak]

Hello,

We need to isolate PXIe-4139 SMUs from +/- 50V signals used to drive the Gate and Source of MOSFETs being tested. The isolation will be  How long does it take to isolate the SMUs? Is there a lifetime specification for the isolation relays? How are the isolation relays replaced, if necessary?

Thank you.

[ZannahBanana]

Hi azstreak,

Thank you for your question! It looks like part of your question got cut off - "The isolation will be -". If you can elaborate on this that would be great! Also, why are you needing to isolate the +/- 50V signals? And where are you looking to put the isolation relays? I've been doing some research on the PXIe-4139 and finding little information about any type of isolation relay being implemented. I did find some information on Isolation Characteristics of the PXIe-4139 on page 17 in the manual linked below. Can you share any references for other sources about isolating SMUs?

http://www.ni.com/pdf/manuals/376115d.pdf

Thanks,

Hannah

National Instruments

[azstreak]

Hello Hannah,

I have another tester that will drive the Gate and Source pins, +/- 50V so I don't want the SMU to affect those pins. I want to just isolate the SMU and run those tests using the other tester.

Thank you for the reference but I've already studied it. I don't any more information on the SMUs which is why I'm asking here. I just want to use the isolation relays already inside the SMUs and not add any more. Thus I need to know how fast they operate, their lifetime, etc..... I already know I'm not going to be operating beyond the spec of

" Isolation voltage, Channel-to-earth ground,150 VDC, CAT I, verified by dielectric withstand test, 5 s, continuous, 150VAC limits "

Any ideas or answers? Thank you.

[ZannahBanana]

Hi azstreak,

Thank you for the additional explanation, much appreciated. From what I have found, the relays in the SMU are used primarily for automatic shutdown and output disconnection should there be overcurrent, overvoltage or overtemperature occurances in the power supply. May I ask why not simply disconnect the PXIe-4139 when running tests with the alternate tester?

Best,

Hannah

National Instruments

[azstreak]

Thank for the question but please just answer my questions. I cannot change the design, nor do I want to add unnecessary relays if I can use the relays inside the SMUs. I would expect them to operate exceptionally well if there is not an overcurrent, overvoltage or overtemperature occurances in the power supply. The relays are in the specifications and I want to use them to put the outputs into an OFF state.

So.... here's what Keithley says about their 2400 series SMUs in the OFF state, "With this output-off state, the output relay opens when the OUTPUT is turned OFF. This disconnects external circuitry from the input/output of the SourceMeter. To prevent excessive wear on the output relay, do not use this output off state for tests that turn the output off and on frequently (such as the Output Auto-Off mode). Note that there is a ~ 15msec settling time for the relay when the output is turned on."

Perhaps the better question is now: "What is the output OFF condition of the 4139?" Thank you for working through this issue. It is important.

Regards,

Steve

[ZannahBanana]

Hi Steve,

I was able to find a bit more information on those relays. It is as follows:

Device = On; Relay = Open

Device = Running; Relay = Closed

Device = Reset/Power Cycle; Relay = Open

There is a property node that can set the Output Connected Property. You can find this in the DC Power palette under Measurement I/O in LabVIEW. Drop the property node on the block diagram and select Source » Output Connected. This property node will probably have the relays disconnect the fastest since it simply disconnects the output.

There is also a VI called niDCPower disable.vi which opens the output relay on devices that have an output relay. This will be a bit slower because it performs a reset and disconnects the output. You may have already discovered these vi's, but I figured they would be helpful to include in my response to your question.

As for how long it will take for the relay to switch, you can probably figure that out by calling Output Connected and then Output Disconnected in a loop, timing the iteration, and then taking an average. I was unable to find specifications for the timing.

I hope this is a bit more helpful!

Cheers,

Hannah

National Instruments

[azstreak]

Hannah,

Thank you. We will check it out with our LV Guys tomorrow. In fact I'll let you know what we find out.

Regards,

Steve

[ZannahBanana]

Sounds good, Steve, I would greatly appreciate that!

The last item I wanted to make you aware of is that frequently disconnecting these output relays can cause premature wear and would require them to be replaced. It appears that this is not able to be done by the user and you would have to send your card in for RMA should these relays fail. These relays are spec'd to last for 10E8 mechanical operations, so based on how you are planning to use the card, you can account for the relay lifetime.

Best,

Hannah