Kay just fixed the CSS adl to opi converter so that XY graphs (medm Cartesian plots) are converted correctly. I just tested that the following are now converted correctly and the BOY displays are correct:
XY plots with only a Y axis (no X axis)
XY plots with both a Y axis PV and an X axis PV
This is a screen shot of BOY configuration for an autoconverted XY plot with both X and Y axis PVs (look at Trace 0).
This is a screen shot of the live BOY display. The X axis is the correct waveform record values.
Thanks for doing this Kay!
From: firstname.lastname@example.org [mailto:email@example.com] On Behalf Of Jörn Dreyer
Sent: Wednesday, April 18, 2018 6:58 AM
Subject: Re: CSS xy graph question
yes the NORD field is set correctly:
epics> dbpr XXX:cam1:WaveLengthArray_RBV
ASG: BUSY: 0 DESC: Array of wavelength per bin
DISA: 0 DISP: 0 DISV: 1
NAME: XXX:cam1:WaveLengthArray_RBV NORD: 4000
SEVR: NO_ALARM STAT: NO_ALARM TPRO: 0 VAL: (nil)
epics> dbpr XXX:image1:ArrayData
ASG: BUSY: 0 DESC: DISA: 0
DISP: 0 DISV: 1 NAME: XXX:image1:ArrayData
NORD: 2048 SEVR: NO_ALARM STAT: NO_ALARM TPRO: 0
If I do a caget it returns:
$> caget HIBEF:XFEL:cam1:WaveLengthArray_RBV
HIBEF:XFEL:cam1:WaveLengthArray_RBV 4000 340.32 340.702 341.084 341.467
341.849 342.231 342.614 342.996 343.378 343.76 344.142 344.524 344.906 345.288
So that part works. If I add this PV as "X PV" to the XY graph widget, CSS has a problem and disconnects from the IOC nand reconnects to it aftera short moment.
Am Mittwoch, 18. April 2018, 13:35:55 CEST schrieb Mark Rivers:
> Has the NORD field of that waveform record been set correctly? What
> happens if you do caget on it?
> From: firstname.lastname@example.org <email@example.com>
> behalf of Jörn Dreyer <firstname.lastname@example.org> Sent: Wednesday, April 18,
> 4:31 AM
> To: email@example.com
> Subject: CSS xy graph question
> I have a question regarding the XY graph widget in CS Studio.
> I want to define the x axis labels based on an array read from my
> areadetector driver. The PV is of type waveform with 4000 elements of
> type double and contains one value per bin. If I put this record as
> "X PV", the widget shows an error and the message that the value of this PV is nil.
> The reason for this is, that the driver is universal for a class of
> spectrometers where every pixel in the sensor stands for a certain
> wavelength. Different devices have different ranges of wavelength they
> Of course I could hard code the range, but then one needs to generate
> a separate OPI file for each type of device.
> What is the correct type for the PV that defines the x axis?
> Jörn Dreyer
Dr. Joern Dreyer
HIBEF DAQ + Controls
Institut für Strahlenphysik
Helmholtz-Zentrum Dresden - Rossendorf e.V
Abt. FWKX Tel: +49 351 260 3263
FAX: +49 351 260 3700
01314 Dresden eMail:
Helmholtz-Zentrum Dresden - Rossendorf e.V (HZDR) Bautzner Landstraße 400, 01328 Dresden
Vorstand: Prof. Dr. Dr. h. c. Roland Sauerbrey, Dr. Ulrich Breuer
Vereinsregister: VR 1693 beim Amtsgericht Dresden