PXIe-4163 running voltage sequence occur "Internal Hardware Error"

[Onezhenji]

Hi,

Background:

I was doing some MOSFET DC curve sweeping using PXIe-4163(24 channels), 4 channels are needed for each test, and I have lots of MOSFETs to test, so I need to switch different channels to satisfy my testing requirements. I connect nothing to PXIe-4163 to check whether it works.

Step Description:

1. I use python and "nidcpower" library to control my PXIe-4163;

2. For each test, I set one channel to be master and the 3 others to be slaves and set the input
terminal of the Measure Trigger of the slaves to be the master device's Source Complete event.

On the slave device(s), the source trigger is the exported source trigger from the master device. (same settings as nidcpower example)

3. After each sequence is complete, I fetch measurements, reset the session, and close the session. (I also tried to add a "session.abort()" after the completion)

4. construct a new session with different channels and repeat step 1,2,3.

phenomena:

It works well for dozens to several hundred sequences, but it would definitely stop and occur an error.

The error message has been attached to this post, please kindly check it, thanks.

[santo_13]

It looks like it is time to send your 4163 for repair/replacement. You need to open a support ticket with NI to evaluate the issue and decide to RMA.

[Onezhenji]

Hi Santo, thank u for replying my post again.

I have 2 same PXIe-1092 chassis and 2 pieces of PXIe-4163 smu and both of them could occur this error.

Today, an AE of NI's distributor told me that constructing(or "initialize" more precisely speaking) a session frequently could be 

unfriendly to my hardware although i close that session everytime. For example, i have several thousands of voltage sequences to run, i initialize a session for my first sequence, then reset and close, and do the same work for the second sequence....

Is it true? if so, why?

[santo_13]

It may be true, but I don't have enough evidence to back up that claim. At initialize and reset, several routines are run on the hardware, but I don't think they might be slowly degrading the instrument.

Typically, I would initialize my hardware once and run all my tests and then close it. In your case, I would not initialize the instrument and close it for every single voltage sequence, it is like restarting your computer every time you need to check an email. You might need to rethink your architecture better if you need to use instruments reliably for a very long time.