Problem with CVI6/GPIB/HP E3631A Power Supply

[Gregoire Diehl]

Hello Group,

i have post this messages yesterday already in labwindows-cvi.general,
but i think
perhaps it will fit here better.

Since i have upgrade to Labwindows/CVI 6.0, i get strange result from
my HP Power Supply (HP E3631A), using the GPIB interface.

For example:
If i have a device connected to the Power Supply i can see the
following result on the instruments display.

1. Channel Output 5V 0.300 mA current
2. Channel Output 12V 0.124 mA current

If i want to read this result over the gpib interface (using the
hpe363xa_Measure function) i get difference results:

1. Channel Output 5V 3.000 mA current
2. Channel Output 12V 1.24 mA current

or

1. Channel Output 0.5V 30.00 mA current
2. Channel Output
0.012V 12.4 mA current

or

1. Channel Output 5V 3.000 mA current
2. Channel Output 1.2V 0.124 mA current

If i do the measurement twice in follow i get the correct result with
the second function call.

Example:

1.Call of hpe363xa_Measure give this results:

1. Channel Output 0.5V 30.00 mA current
2. Channel Output 0.012V 12.4 mA current

and the 2. Call give this (correct) results:

1. Channel Output 5.0V 0.300 mA current
2. Channel Output 12.0V 0.124 mA current

I have downloaded the latest driver for the HP Power Supply from the
Nation Instruments Website (old is dated 05/17/99 and the newest dated
06/14/2000), but without improvement. Both driver Version give the
faulty results with CVI 6. With CVI 5.5 i have no problems reading the
correct values over the GPIB-interface.
Remark: A second call of the same function would return the correct
values, but the first call does not.

Can someone please point an assertion for the described behavior?
I would be very grat
eful for each assistance

kind regards
Gregoire Diehl

[Zdenek Rykala]

If you want detect problem, try use NI-Spy. Configure NI-SPy to spying NI-VISA. This configuration you can find in menu Spy>>Options>>View Selections. Check only NI-VISA. When configuration is done, enable capture (F8). Now you can launch your test program. NI-SPY displays all VISA calls and their results. There should be call viQueryf with ":MEAS:CURR?" or (":MEAS?" - voltage) commands. You can detect what instrument returns. If all values will be correct in NI-SPY, driver does not parse values correctly. If you do not solve this problem, attach file from NI-Spy.

[Gregoire Diehl]

Hello,

thanks for the proposal, how to get closer to the error.
I will test it out and then post the result.

Thanks again
Gregoire

[Randy Solomonson]

A spy capture is a great idea. There are two versions of the driver, an IVI driver and a Legacy driver. Try the other type of driver. Let me know how it goes.

Randy Solomonson
Application Engineer
National Instruments

[Gregoire Diehl]

Hello Randy and Zdenek,

> A spy capture is a great idea. There are two versions of thedriver,
> an IVI driver and a Legacy driver. Try the other type of driver.Let
> me know how it goes.

I have tested 2 different drivers (IVI and Legacy), without any
improvement. But it seems to me, that the latest Version of the Driver
(IVI) is the best, cause i get the correct result always after the
second call. With the other version i can get faulty result up to 10
times and than the correct values will be returned.
I have attached a spy-file where you can see the returning results.
Please have a look at line number 3 and 6. The function return 49,99V
first and then 4,99V (correct).
The function returns two different exponent. The same effect at line
num
ber 9 and 12, which return 3.287A first and than 0.330A (correct).
I have tested the driver on another machine with CVI5.5 running and
the results were always correct. On my system, where i have CVi5.5 and
6 installed, i get the faulty result under CVI5.5, too. So i think
that
the problem is in a direct context with the new Version of
LabView/CVI.

Any ideas?

thanks & best regards
Gregoire

PS: I have switch several HP Power Supply?s to be sure there is no
hardware defect.











[Attachment Capture.spy, see below]

[Zdenek Rykala]

Hi,
Driver scans values correctly. In the second step, try found difference between two PC, which you are using. PC with CVI 6.0 should be use NI-VISA 2.52 and your second PC with CVI 5.5 may be using VISA 2.01. I do not believe that problem will be in CVI or VISA.

Which kind of hardware are you using? May be its timing problem if your computer with CVI 6.0 is 2 times faster as second one. I do not know all your sequence of functions, which you are calling. You can add before first call of Measure function delay (for example 100 ms). This is my last idea now.

Regards,
Zdenek

[Gregoire Diehl]

Hello Zdenek,

> Which kind of hardware are you using? May be its timing problem if
> your computer with CVI 6.0 is 2 times faster as second one.
Exactly in reverse :-(. The system, which is working correctly is a
PentiumIII 1GHz machine, the other system is a PIII-700Mhz. Nearly
same configuration. But i'm using an older AT-GPIB/TNT card for ISA,
during the correct system used a PNP PCI GPIB-Card.

> know all your sequence of functions, which you are calling. You can
> add before first call of Measure function delay (for example 100
ms).
But what i find out, was that the errors occurs even, if i am in
interactiv mode and running the function panel one by one manualy and
there i think there could not be a timing problem.

> This is my
last idea now.
I am also rather helpless

best regards
Gregoire