AnsweredAssumed Answered

vrf GPIB slowdown

Question asked by Dr_joel Employee on Jun 16, 2003
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"><HTML><HEAD><META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"><META content="MSHTML 6.00.2600.0" name=GENERATOR><STYLE></STYLE></HEAD><BODY bgColor=#ffffff><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>HelloVic,</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003>    <FONT face=Arialcolor=#0000ff size=2>I did some of the original speed testing on the 8753D and8753ES, so I have some of the information you need.  It can actually befound in the user and programming manuals.</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>I havean 8720D manual, should be the same as the 8753 for the most part.  For2-port calibration on, reading 1 data array, 201 points, narrow band, IFBW=3700,the data read time is 1.54 seconds. (remember, this is for a forward and reversesweep, applying 2 port cal, and reading out 1 parameter).  If you arereading out only 1 parameter with 1-port cal, I would expect it to be muchfaster (2-3x) so your time of .5 sec is reasonable.</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>The8753ES (or 8720E or ES) manual shows for the same set-up, the total time is0.151 secs (vs 1.54 above).  So, we see that the ES is much faster (muchfaster CP and DSP in the ES). Probably more in line with the 8714ES.  Inthis case, GP-IB is typically not the bottle neck, but rather the instrumentmeasurement time can be.</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>Datatransfer time only is also included in the specs section, with form 2 (32 bit,recommended...that's about what the analyzer uses internally)</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2>8753D=85 msec, 8753ES= 15 msec.  You can add to that the sweeptime:</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2>8753D=200 msec, 8753ES=77msec.  Then there might be some otheroverhead as well (markers on, time domain, it's slower for wide bandmeasurements...).</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>Togive you an idea of LAN  vs GP-IB, the new PNA network analyzer familysupports both, with the details that 1601 points over GP-IB is 43 msec, and overLAN is 2 msec.</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>(201points, the times are 7 ms and 1 msec, but the resolution is 1 msec, so I'm notsure if that's a valid time).</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ff size=2>Asalways, it's all in the manuals...if you can find it.</FONT></SPAN></DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2></FONT></SPAN> </DIV><DIV><SPAN class=725302117-17062003><FONT face=Arial color=#0000ffsize=2>Joel</FONT></SPAN></DIV><BLOCKQUOTE>  <DIV class=OutlookMessageHeader dir=ltr align=left><FONT face=Tahoma  size=2>-----Original Message-----<BR><B>From:</B> vicr@hq.lindsayelec.com  [mailto:vicr@hq.lindsayelec.com]<BR><B>Sent:</B> Monday, June 16, 2003 10:40  AM<BR><B>To:</B> VEE vrf<BR><B>Subject:</B> [vrf] GPIB  slowdown<BR><BR></FONT></DIV>  <DIV><FONT face=Arial>Thank-you to all that responded to my first e-mail. I  tried all the tips given without any real success.</FONT></DIV>  <DIV><FONT face=Arial></FONT> </DIV>  <DIV><FONT face=Arial>Turns out my old NI GPIB card was installed such  that the IRQ feature was not being used. I disabled one of comm ports and  re-installed the card. The free IRQ was found and used.</FONT></DIV>  <DIV><FONT face=Arial></FONT> </DIV>  <DIV><FONT face=Arial>Using the 8753D I can now transfer 201 complex data  points in a little under half a second (instead of 12 or 13 seconds). I use  OPC? to synchronize the transfer with the sweep. Hartmut's code, using the  status register, yielded similar performance.</FONT></DIV>  <DIV><FONT face=Arial></FONT> </DIV>  <DIV><FONT face=Arial>I can work with this, but am a little disappointed. An  8714ES with a LAN connection will transfer data much quicker than  this.</FONT></DIV>  <DIV><FONT face=Arial></FONT> </DIV>  <DIV><FONT face=Arial>What is a reasonable performance expectation for an old  GPIB card?</FONT></DIV>  <DIV><FONT face=Arial></FONT> </DIV>  <DIV><FONT face=Arial>Best Regards,</FONT></DIV>  <DIV> </DIV>  <DIV><FONT face=Arial>Vic Race</FONT></DIV>  <DIV> </DIV>  <DIV><FONT face=Arial>Lindsay Electronics<BR>50 Mary St. W.<BR>Lindsay, ON,  K9V 4S7<BR>Voice: 705-324-2196 Ext. 241<BR>Fax: 705-324-5474<BR><A  href="http://www.lindsayelec.com">www.lindsayelec.com</A><BR></FONT></DIV>---<BR>You  are currently subscribed to vrf as: joel_dunsmore@agilent.com<BR>To subscribe  send a blank email to "join-vrf@it.lists.it.agilent.com".<BR>To unsubscribe  send a blank email to "leave-vrf@it.lists.it.agilent.com".<BR>To send messages  to this mailing list, email "vrf@it.lists.it.agilent.com". <BR>If you need  help with the mailing list send a message to  "owner-vrf@it.lists.it.agilent.com". </BLOCKQUOTE></BODY></HTML>  

Outcomes