Experimental Physics and
Industrial Control System

Dear Dr. Rivers,

Thank you so much for your kind suggestions!

We will try your suggestions and return back soon.

Sincerely yours,

Yujong Kim

2012. 12. 7. 오전 7:45에 "Mark Rivers" <[email protected]>님이 작성:

Here are a couple of ways to track down the problem:

- Run an areaDetector IOC on another computer but use medm to display it on the system you are running medm on for the scope.  Do the cartesian plots in the statistics plugin look OK?  If not then the problem is with medm on the scope machine.

- Run medm for the scope display on one of the system you know works with medm cartesian plots, leaving the scope IOC on the computer it is currently running on.  Are the plots OK?  If not then the problem is the scope IOC.

Mark

________________________________
From: Yujong Kim [[email protected]]
Sent: Friday, December 07, 2012 8:35 AM
To: Mark Rivers
Cc: Trevor Downer; EPICS Tech Talk; Nicholas P. DiMonte; Anthony Andrews; Kukhee Kim
Subject: RE: Tektronix 3014 Tracing Problem


Dear Dr. Rivers and Dr. DiMonte,

Thank you so much for your valuable suggestions!

Yes, our operating system is Fedora 14.

We did not have any big problem when we use other EPICS extensions or modules (AreaDetector, MEDM, MCA, Burt) with Prosilca GigE cameras, TDK Lambda power supplies, and we can make good Cartisian plots with MEDM well.

We already tried the CA Max Array Byte issue in both IOC and MEDM sides. But failed.

Does this related to lesstif or OpenMotif issue for MEDM?

At the moment, in this computer for scope, MEDM is complaining font issues when we launch it.

We could solve the font issue by installing OpenMotif and fonts aliases at other computers for Lambda power supplies (not this for scope).

Sincerely yours,

Yujong Kim

2012. 12. 7. 오전 7:17에 "Mark Rivers" <[email protected]<mailto:[email protected]>>님이 작성:
His screen shots don't look like a Windows machine, they look like a Fedora machine I think.

Mark

________________________________
From: [email protected]<mailto:[email protected]> [[email protected]<mailto:[email protected]>] on behalf of Nicholas P. DiMonte [[email protected]<mailto:[email protected]>]
Sent: Friday, December 07, 2012 8:11 AM
To: Trevor Downer
Cc: Yoojong Kim; [email protected]<mailto:[email protected]>
Subject: Re: Tektronix 3014 Tracing Problem

Trevor,

There are two separate issues that you are dealing with. The SDDS script needs the SDDS package installed to display the waveforms and is not needed for the MEDM display. I added these SDDS scripts to produce a cleaner and more detail display of the plots for our technicians here at the APS, I included them as an example only. Sorry for the confusion.

The MEDM display issues may be that EPICS_CA_MAX_ARRAY_BYTES environment variable is not set high enough. This variable needs to be set on both the IOC and on the computer your using. For the TDS3000 at 10k record size, you will needed to set this variable to 80000 to accommodate the waveform being a "float" data type. The updated release for this support will have the option of selecting between 10k and 500 data points, which means this variable could be set as low as 20000.

I'm not an expert when it comes to running EPICS on a Windows machine, but the blue section in the MEDM display looks like it might be a EPICS_CA_MAX_ARRAY_BYTES environment variable issue. On a UNIX/Linux system I would see four black squares, one for each waveform, but they would be all blank if this variable is set to low. The blue section I see in your post is the back ground color I chose for this section, so I'm puzzled as to why the "Cartesian Plot" objects are not visible, maybe it a Windows thing. Window experts please respond.

On a side note, I will be releasing a long over due update to this support by February, which has improvements to reliability issues with the TDS3000 series and enhanced waveform handling for the TDS5000 series. Also, a clarification on the use of the SDDS utilities has been updated.

Keep me updated so I can add this issue to the new release.
Thanks,
Nick D.


On 12/6/2012 6:00 PM, Trevor Downer wrote:

Greetings and salutations!

I am a student at Idaho State University working on a project to
control a Tektronix 3014 oscilloscope, using EPICS, for the Idaho
Accelerator Center.
We have successfully installed EPICS and the necessary extensions to
communicate with the TDS3014 oscilloscope. Using medm we cannot see
the waveform tracings, we can however see the tracings from a web
browser, however its refresh rate is quite slow.
There seems to be a problem when I click on the SDDS Plot button, medm
outputs a message in the terminal reading:

"medm/trace10k: line 8: [: T: integer _expression_ expected
 Usage: trace [1-4][HV or LL]<CR>\n"

 Attached is a screen shot of the traceless oscilloscope in medm and
the trace10k file where there seems to be a problem.
I would greatly appreciate any help to fixing the tracing problem,
thank you very much!

-Trevor Downer