Test System Speed vs. test equipment

[Troy]

Have any of you run into a situation where your test takes longer with one particular piece of test equipment over another.  I have an ATE system that uses either HP 8594A or Agilent E4403B spectrum analyzers.  I have 10 stations with a mix of these analyzers.  For a particular test, all but 2 of the test stations take about 50 seconds to complete the test.  Some are with E4403Bs and some are with 8594As.  However, I have 2 stations that take an additional 10 seconds every time the test is run.  Both of these stations are equiped with E4403Bs.  We tried swapping analyzers around and the slow test seems to follow these two particular analyzers.  I am controling them with GPIB and I am preseting the instrument at the start of the ATE application, but not at the start of each test.  We have also performed a manual preset of the equipment with no effect on the test speed.  The PC is more than adequate to handle the load and there are no other applications running the background.  There is only one other instrument on the GPIB bus, but I am not communicating with it while communicating with the spec an.

My code is written in LV8.2 and I am using VISA controls in the code as opposed to the low level GPIB.  I have gone through the code and other than a general rough quality to the code (not the best LabView style), I see nothing that should cause this type of delay.  As I said above, it appears to be focused to just the two analyzers, but others of same make and model work fine.

The real question here is am I potentially missing something in my initial VISA setup that could be causing this.  For the most part, I wrote my own drivers as at the time we developed the code, the IVI driver was the only one available and it was not the most stable driver.  Each time I want to interact with the analyzer, I use VISA open, read, write, and then close once I am done, so I am not leaving a reference open.

Any thoughts to  what could be going on here?

[BerkO]

Hello Troy,

If you are getting the expected behavior and appropriate readings from the instruments you most probably aren't missing any VISA communication settings.

Now I worked with the E4404B for quite a bit and one thing that I noticed is that, the more you work the instrument, the more the performance declined.  A power cycle usually resolved this issue. Especially switching frequently between measurement personalities was eventually slowing the general operation. I unfortuantely couldn't update the firmware on my instrument because the front panel curcuitry was not functional but I had two before the most recent revision.

Also for the units that are identical, are you running the same firmware revisions? This is pretty important because I know for sure agilent did a lot of improvements for those instruments.  Also having extra modules (expanded memory, extra measurement personalities) etc. will effect the speed of the operation as well.

Another suggestion I can give you is to try the driver that we developed for the E4403 that you are using.

You can find the driver here.

Let us know how it goes.

Message Edited by BerkO on 05-02-2008 03:09 PM

[Troy]

Berk,

Thanks.  I didn't think about the firmware being a potenital factor.  I will have the guys report to me what the firmware is for each of the analyzers and see if there is a difference. 

As for the driver, I would like to use yours, but it would require a pretty significant change to the code (like I said, not the best LabView style in use).  If the firmware doesn't look to be the problem, I will resort to redoing the code to include the NI driver.  It could use general clean up anyway, but that's time I really don't have, unfortunately.

I'll let everyone know what I find. 

Troy.

[BerkO]

Hello Troy,

If it is not the hardware or firmware difference between instruments, and if the exact same code is running on each of them, most probably agilent would be able to guide you better as to why only some of the instruments are behaving slowly.

I hope it works out.

Please let us know if you have any more questions!

[Troy]

Berk,

You were correct, it is a difference in firmware.  However the new firmware is the slower of the two versions.  I have two with A.14.01, and two with A.14.03.  I did a bit of digging on Agilent's web site and the defect changed in A.14.03 was to change the timing of the GPIB Aux_Timing register. 

I am not sure exactly what this register does or is, as this is the first I have heard of it.  I am still researching, but do you know if this would be a user adjustable register?  I am hoping so, as this could become a big deal as time goes on and the 8594s I have in the system begin to drop.

[BerkO]

Hello Troy,

I looked at the manuals that relate to the instrument but that register doesn't seem like one that is even mapped in the user's manual or calibration manual.

I believe downgrading the firmware is possible and if the newer version is disrupting your current application you might want to try downgrading the firmware to the same version as the problem free instruments. 

I haven't heard of that register before either.

Please let us know of the progress.

Good luck!

[Troy]

Berk

Just a quick update.  I talked with Agilent's local Tech support guys (their office is right across the business park) and they stated they were not going to look into the speed issue at this time.  They did give me instructions on how to back rev the firmware in the analyzer.  I will have my guys do that in the coming days and let you know how it turns out.