Contents of Vol.11,No.1

1. Buffered BPM Sampler
2. Changes for EPICS DM Print
3. Intensity Scaling for FB31 Matrox
4. Knobs Panel
5. Avoiding Simultaneous IPLs

Page contact and owner at end of this issue.

Buffered BPM Sampler

December 15,1997

In order to improve beam jitter diagnostics, a system has been developed to acquire and correlate buffered acquisition for BPMs and BPM-like devices such as toroids and gated ADCs.

BPM Acquisition and Correlation

A new BSMP standalone process is created to regularly acquire and save data, similar to the BPM sampler. Correlations may be plotted between any two devices, or averages taken for a single device. Fit parameters are saved into history buffers, and data files are available for user displays. The fit types supported include all of those available in the Correlation Plots system. The standalone is fully database and file driven.

The correlation of suspect devices is specified by the BSMP database primary, in micro VX30. Each unit of that primary specifies a correlation of one or two BPM like devices to be performed by the BPM sampler, and the fit to be performed on the acquired data. The BSMP standalone runs periodically to acquire those correlations and store the results in Correlation Plots compatible files and to store the fit parameters in the database from which they're history buffered. A new SCP panel is also available to help browse this data.

Data acquisition is driven by the control file called BSMP_CRR_DIR:BSMP_CRR.CTRL. Each line in the file drives the buffered acquisition for a single BPMD by indicating BPMD number, number of pulses to acquire, delay between acquisitions, filtering parameters, and data error/gap handling parameters. Comments at the top of that file provide more information on setting it up. For the time being, only BPMD 15 (ALLBEAMS) is supported. TMIT data values are automatically normalized by 10**10.

For each successful acquisition pass, a Correlations Plots compatible file is stored for each online unit in the BSMP database into the directory BSMP_CRR_DIR. The file names are of the form DATA_0nnn.DAT where nnn is the BSMP database unit.

SCP Interface

The new SCP panel is of the ``dynamic buttons'' kind. Once a BSMP correlation is selected the history of its fit parameters; its latest data acquisition correlation, or a nominated ``gold'' correlation can all be displayed.

The ``gold'' acquisition is intended only as a method to mark a particularly interesting correlation. One ``gold'' can be saved for each BSMP unit, and once saved it replaces the previous gold acquisition.

Changes for EPICS DM Print

October 23, 1997

The Epics Display Manager (DM) has traditionally taken a long time, as much as three to five minutes, to print a display. Tom Himel has researched some alternative methods of printing and as a result print times for Black and White are now down to about 30 seconds on the laser printers.

Other important changes are:

• The information in the DM Printer Select menu is used for both Quikprint and Print. Quikprint is activated from the display pulldown menu using the right mouse button key (MB1). Print is activated from the pulldown menu or the "Print" button at the top left of most displays.
• The generated file has a ".ps" appended, so that printing via VMS and the Imagen printers will work.
• The Quikprint option now waits 5 seconds, then allows ANY window to be selected for printing. For example, the user may select a StripTool strip chart or Matlab plot on the same X-terminal. It is important to note that the utility now used is a screen grabber, not a window grabber, which means that the window you wish to print must be fully visible, i.e. the user must manually bring the screen to the foreground before using Print or Quikprint. The actual contents of the screen defined by the outline of your special window are the contents dumped to the printer.

The three options on the DM printer selection menu are translated as follows:

1. B&W - This is fastest.
2. Greyscale - Gives similar output somewhat more slowly.
3. Color - Very slow, but colorful.

The user may also opt to ``Print to File" to get a postscript file instead of hardcopy.

Currently the only print queues available for DM printout include the MCC imagen and laserwriter on the first floor (ima1 and lwmcc1_ps) and on the second floor (ima2 and lwmcc2_ps). Requests for new print queues must be sent to Jingchen@slac.stanford.edu.

To get color hardcopies or transparencies, the user must ``Print to File", using the "Color" option. The resulting postscript files must then be transferred (FTP'ed) to MCC and then output to the color printer (PRINT/QUEUE=MCC$HP_COLOR ).

Intensity Scaling for FB31 Matrox

January 20, 1998

For the FB31 Matrox board, users may now control beam intensity scaling by entering parameters on a touch panel. From the main feedback panel to the feedback magnet panel, the user may select the FB31 PHAS & AMPL panel, and from there to the FB31 FFWD panel. The user may enter the average beam current (TMIT) for the electron and positron beams, or enter the Width of the intensity on the Matrox board. The new buttons are

and the units are in particles, so typical entries are 3.E10, for example.

Knobs Panel

November 20, 1997

This is a note concerning the use of the sliders on the knobs panel.

Dragging a timing device can produce an incremental change which is too large a change to be acceptable by the timing software.

This happens when the range of the scale is too big as when you have just assigned a knob to a device. In the future we plan to try to be able to initially set the range such that there is no initial problem but for now there is.

So before attempting to drag or set the value of a timing knob, set the magnification to at least 1000. This will set the range such that you should not get timing error messages.

If you forget to do this, simply restore the knob to the value it was when assigned using the restore drop down list and then increase the magnification.

For setpoints, you will also get an error message and no change if you enter a new value which produces too large an incremental change. At the moment there is no work-around.

Note: This discussion is included in the help for the slider. It is also presented in the knob panel error text window if an error occurs.

Avoiding Simultaneous IPLs

December 15, 1997

A small change has been made to the single-micro IPL function to protect against starting an IPL when an IPL is already in progress for that micro.

Henceforth, if a request is made for an IPL of a single micro and the software determines that the micro is currently being IPL'd, the requester will be presented with the following information and query:

"This micro already being IPL'd; IPL anyway?"

IN ALMOST ALL CASES, THE RESPONSE SHOULD BE THE DEFAULT ``NO."

The only reason for responding ``YES" might be if the previously-IPLing SCP died or got hung.

This new functionality is for single-micro IPL's only and does not affect IPL ALL*.

Back to top of this issue

Translated from original PlainTeX by index2html.pl.