AnsweredAssumed Answered

8753ES a 16kbit, 16kbyte, 32kbit, or 32kbyte memory partition?

Question asked by cheenu on Jan 31, 2013
Latest reply on Feb 11, 2013 by cheenu
previously posted a question  about issues I was having with a test setup involving the 8753ES network analyzer.  I noted that the unit would intermittently freeze up when testing thousands of dies.

After some test runs, I figured out that the 8753ES would fail after 16,384 iterations of a loop.  In some cases exactly that much, in some a hundred or so less than that.

The loop consists of 4 subvi’s:

1)     a) A GPIB Write of MARK1;CAL1;CORR OFF;
        1b) A GPIB Write of POIN __;POWE __;CENT __;SPAN__;
        1c) A GPIB Write of S21;LOGM;IFBW __;TRACK ON;CORR ON;

2)    a) A GPIB Write of *OPC?;SING;
       2b) A 15msec wait,
       2c) A GPIB Read with 10 bytes and a long timeout (10 minutes – so that my NI Spy capture isn’t filled with failed reads),

3)    a) A GPIB Write of CORR ON;MARKCENT;SPAN __;IFBW __;POIN __;

4)    a) A GPIB Write of *OPC?;SING;
       4b) A 15msec wait,
       4c) A GPIB Read with 10 bytes and a long timeout (10 minutes – so that my NI Spy capture isn’t filled with failed reads),

Here’s what happens:

1)     When I run the loop with these 4 subvi’s, starting it off with an *RST command and recalling the calibration state: 
        •        It will fail, like clockwork, after 16,200 or 16,300 iterations.  
        •        One time, it was exactly 16,384 iterations.  
        •        It will be unresponsive to any GPIB commands, and 
                   any front panel controls except for the power button, or Local then Preset hardkeys. 

2)     When I go the network analyzer and manually preset, recall the calibration, and set the IFBW, number of points, power, center, 
        and span and then run only the 2nd (or 4th, they are exactly the same) subvi – the one which writes *OPC;SING; and 
        waits to read the “1” -- in a while loop: 
        •           It runs without failing for more than 40,000 iterations.
        •           The network analyzer is not locked up or unresponsive in any way.

So here are my questions:

1)     Is there some memory unit or partition in the 8753ES network analyzer which is exactly 
        16kbits, 32kbits, 16kbytes, or 32kbytes large that’s getting filled up by the commands I’m sending?
2)     Is there some part of this code which is written incorrectly which would cause this periodic failure?
3)     Periodically sending the PRES; command (and then recalling the calibration file) doesn’t work 
        to prevent this failure, but would sending the *RST command do the trick?

Thanks, and I would appreciate any insight any of you might have about this problem, even if you don’t know the answer.

Edited by: Cheenu on Jan 31, 2013 6:29 PM

Edited by: Cheenu on Jan 31, 2013 6:30 PM