June 27, 1997 All That Fits is News to Print Vol.10,No.2

Contents of Vol.10,No.2

  1. Klystron Software Changes
  2. Knobbing a Steering Correction
  3. Running a SCP on a MAC or PC
  4. Modems for Running SCPs from Home
  5. Multiknob Enhancements
  6. MCC Whiteboard
  7. Fast Feedback Panel for Changing Exclusion and Inclusion Masks
  8. MPSFAULT Display Enhancement
  9. Control Of PPS Devices
TeX source

Page contact and owner at end of this issue.

Klystron Software Changes

May 21, 1996

Author: Keith Jobe Subsystem: Accelerator User Impact: Small
Panel Changes: None Documentation: No Help File: None

The klystron triggering and software support has remained relatively stable and received very little maintenance within the SLC control system for almost 10 years. The incorporation of PEP-II with the SLC program has necessitated some changes in the manner phase values are sampled and reported. Along with these changes, a number of improvements have been brought to the klystron system.

The software residing in the PIOP is responsible for the sampling, calculations, and reporting of most of the essential klystron and subbooster parameters. This code determines if the pulse is Accelerate or Standby, and ignores the standby-based data when the accelerate trigger rate is sufficient. This algorithm allowed the database to display a creditable value for klystron stability and performance independent of the operational mode of the station. (Since Standby and Accelerate pulses have different-enough phase and amplitudes, combining their data will give meaningless phase and amplitude jitter values).

The linac operation for PEP-II requires that some sectors and certain klystrons operate with a beam-code based pulse-backphased or pulse-amplitude-modulated drive. Even more so than for standby and accelerate pulses, combining data from these modes of operation will result in meaningless database information.

The feature refined for PEP-II operation is the use of the HSTA_1BEAM bit in the klystron/subboosters database. When set, the station compares the most recently broadcast PP value with a bitmask built up of beam code 8 and the related substitute beamcodes. Data collected for these pulses is not used in the determination of the klystrons performance. This algorithm will need a bit of refinement for LER/HER operation.

New Features

Beam-Code Filtering for Database Values

Klystron software now respects the HSTA_1BEAM identifier. When set, the PIOP is configured with the PP and substitute beam codes for the PEP-II operation. This is needed for stations which are part of a back-phased sector, which are individually backphased, of have PAU drive or phase control used to control the beams for the PEP energy requirements. This status may be seen from the DECODE STATUS BITS display from the diagnostics panel and is identified at the bottom of the PIOP Data Display .

This feature is currently enabled in all subboosters and klystrons in sectors 3, 4, 8-10.

Sample Rate Counter

The database data is usually taken on all accelerate pulses, or on all standby pulses if the accelerator rate is too small. When beam-code filtering is enabled, the klystron is sampled for the database on non-PEP pulses. This rate is often less than the accelerator rate, and can in some cases be 0. A line appears on the PIOP Data Display indicating the sample rate for the station (which is normally hidden when the value is the full accelerate rate).

ACCelerate Trigger Rate 30 30 1
Sample Rate 20 30 1 -- tols
MODulator Trigger Rate 60 60 1

This reads as modulator firing at 60 PPS, beam-related triggers at 30 PPS, but database values sampled at 20 PPS. Presumably beam 8 is active at 10 PPS. Klystron sample-rate errors

If the sample rate is too low to support a reasonable database value, the status message NO Sample Rate appears on the Klystron CUD and on the usual klystron displays indicating a sample rate problem. This is similar to the existing No Accelerate Rate display text.

Additionally, the micro refuses to phase trim stations with this error.

SLED Phase-jump Monitoring

A new analysis of the SLED phase waveform measures the phase jump near the PSK (phase shift keying) modulation that discharges the SLED cavity. In a perfect world, this value would read nearly zero when the cavity is properly tuned, but due to the relatively slow modulator beam voltage pulse rise time and phase detector non-linearity, this value is often as large as 50 degrees.

The RF PHASE FTP indicates the two sample values and times with short green lines, and the numeric value can be seen on the lower left of the display ( ``SLDP: 32.1" ). This analysis is performed every 5 minutes by the MDL slow feedback loop, and is recorded in the history buffer in an attempt to help identify mistuned cavities.

Substitute Beam Code Display

A new display is available from the KLYS DIAGnostic SQUARED panel, the SUBSTitute PP BITMAP display shows the actual bitmaps used by the PIOP in determining the FTP or sampling beam code match conditions. This is a difficult to read HEX display of a 256 bit value, with the LSB on the lower right. For example:

KLYS 31 data-filter: 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0004 0000 0100

reads as veto PP=8 (the``1") and PP=34 (the ``4").

TOC Counters

Every klystrons thyratron has a trigger circuit that pulses the switch tube. If the thyratron is incorrectly set up or is near end-of-life, the grid of the thyratron arcs and a protection timer generates a one second pause in the klystron triggers (and a corresponding interruption in the beams acceleration). The Klystron Error Display indicates the TOC fault counts, corresponding to events in the error log.

These faults are also indicated on the usual displays, but due to their duration and frequency are extremely difficult to observe.

Enhancements to Standard Features

Separate Errors for Klystron and Waveguide Vacuum Trips

The PIOP now generates separate errors in the error log for the two vacuum error sources: PIOP-E-WG_VACUUM and PIOP-E-KLYS_VACUUM . The error accounting for the klystron-errors display still sums the two errors.

Polynomial Correction of Correlation-plot Phase Data Sampling

The BPM support for klystron phases has been taught to unfold the klystron phase data and apply the database polynomial to linearize the data. Prior support used the literal ``straight-line" approximation, which was sufficient to identify jitter sources but of limited utility in phasing klystrons with the BPM buffered data system.

Display of SBST Glasman Voltages of PIOP DATA Display

Throughout the prior run, the aging subbooster power supplies were replaced with modern power supplies from Glasman. These supplies have been monitored by the Analog Status system (available from INDEX -> ANALOG SYSTEM INDEX -> LINAC ANALOG STATUS -> SBST HV MONITOR ). This output is now displayed on the PIOP DATA display for each subbooster. History plots and custom displays should continue to be obtained from the analog status set of panels.

DAC Readback Errors

The Subboosters and some klystrons use a DAC rather than a stepping motor to control their phase shifter. When the DAC is set and has an immediate readback error, the error message generated now contains the expected and actual bit patterns.

Software Bug Fixes

Serial Number on FTP Display

The serial number for the klystron ``fell off" the FTP displays when we moved to the X-window system. This has been moved a bit and is visible again.

Correct Display and Error Reporting for Multi-unit FTPs

Changes in the code for the DEC-Alpha resulted in the many klystron displays failing to display if there was a single micro or klystron station timing out or reporting an error. This has been fixed, and the traditional error text now appears on the display.

Correct Handling of Substitute Beam Code Translation for FTPs

A bug was introduced in the downloading of the FTP bitmap for data filtering for beam-code only displays and certain FTP displays. This can be seen by changing the FTP Data Pattern followed directly by an RF PHASE FTP . Following the data pattern by an amplitude FTP sets the PIOP bitmask correctly. This bug has not been fixed, but I expect it will disappear soon.

Phase Trim of **** Phase Stations

Klystrons with no rate (20-4A) report phase values of **** (-1.0e4) and would ``trim" every 5 minutes. This would result in the fox phase shifter changing by (10,000 - 0) mod 360 = -80 degrees each time. Klystrons which are not reporting a valid phase value will no longer trim, and must wait until a valid phase measurement completes prior to any changes.

Trimming of a klystron is disabled if any of the following conditions exist:

  1. CAMAC or Database errors with PIOP
  2. BAD Cable status
  3. Low RF Power
  4. No Sample Rate
  5. Failed last FOX Home test
  6. Requires IPL
  7. Phase reads all ****'s

Knobbing a Steering Correction

April 4th, 1997

Author: Greg White Subsystem: Steering User Impact: Small
Panel Changes: Few Documentation: No Help File: Yes

A facility to ``knob in'' a correction has been added to steering. The buttons to help do this are located on a new panel named ``STEER OPTION PANEL'' which is accessible from the ``STEER DIAGNOSTIC'' panel. Some minor re-arrangement has been made on that panel to accommodate switching to this new panel.

The new buttons are




The Knob Plane buttons allow the user to specify which knob will be assigned to the correctors of the corresponding plane, and to deassign the knob. When one of them is pressed, they prompt for a knob number, entering a -1 will deassign the knob.

When the ``Knob Steer Calc'' is ``off'' the steering

 x or y 

buttons for x and y have their regular meaning of ``trim or perturb the magnets directly to the values calculated last by the steering method''. If the ``Knob Steer Calc'' button is toggled to highlighted, then subsequently, whenever a ``trim'' button is pressed, a multi-knob file of correctors in the corresponding plane will be written instead, and assigned to the knob specified in the ``Knob Plane'' buttons.

Then when the ``Knob Steer Calc'' is toggled back to off, the ``trim'' buttons revert to their usual meaning, and any knobs assigned for steering are deassigned (so there are two ways to deassign a knob).

The multi-knobs implement the ``percentage correction'', and are labelled in the case of x correctors ``X corrs %''. That is, when the knob reads 100, the correction is 100% implemented. Its limits are -100% to +200%.

Some additional points: if ``Knob Steer Calc'' is highlighted then both planes must be steered through knobs; its not possible to, say, enter a knob number only on ``Knob Plane X'' and then hit TRIM Y. That will result in the message ``No Knob assigned for Y Correctors''. You can assign a knob for only one plane, you just can't knob one plane and regularly trim the other. Also, power steering doesn't use the knob interface and proceeds as it always has regardless of ``Knob Steer Calc''. That includes iterative steering. The dampening factor also does not apply to knobed corrections.

Running a SCP on a MAC or PC

April 16, 1997

Author: Terri Lahey Subsystem: SCP User Impact: Small
Panel Changes: None Documentation: Yes Help File: No

To run a SCP on a MAC or PC, you must run an XServer (program) on your computer. The supported XServers (ie on which the SCP will run correctly) are:

eXodus on a MAC
XWin32 on a PC, running NT or windows95

SCS has arranged discounted licenses for SLAC business. You will need a license (for each node) to run this Xserver software. Check the web or contact SCS Help Desk to see if one or more licenses have been assigned to you. If you need a license, contact your ACS or the person listed below for your department:

PEPII: Andrea Chan
Accelerator: Keith Jobe
Controls: Samanta Annuzzi

The software is available via SCS central distribution by using your SCS SLAC NT account. Contact SCS Help Desk to get a SLAC NT account.

For help on installing and running the Xservers, see VMS documents:


If you have trouble installing this software or running a SCP, contact Controls Software.

Modems for Running SCPs from Home

April 17, 1997

Author: Terri Lahey Subsystem: SCP User Impact: None
Panel Changes: None Documentation: Yes Help File: None

A new group of phone numbers and modems are available at MCC for running SCPs over modem lines (from home). For non-control system functions or ASCII-terminal connections, you can continue to use SCS modems. The MCC modems require user accounts and passwords. Contact Samanta Annuzzi for an account request form.

PC users who want to run a SCP or other control system functions are recommended to use MCC US Robotics for Xsessions.

MAC users are recommended to use the SCS ARA SERVER. This ARA server will continue to support connecting to MCC and running SCPs. There is also an alternate ARA SERVER at MCC. Contact SCS Help Desk to request a DIALBACK account in the ARA Server.

Phone Numbers for running SCPs

MCC US Robotics for PCs dialing PPP connections:

(650) 926-5440 --- 926-5443
(650) 926-5487
(408) 873-1099

SCS ARA Server for MACs using dialback appletalk connections

(650) 926-5030 --- 926-5045 ARA v2.0 & up
(650) 926-5046 --- 926-5049 ARA v1.0

MCC ARA Server for MACs using dialback appletalk connections

(650) 233-1986
(650) 854-5491
(408) 255-7653

SCS modems for WEB access & non-control system functions

(650) 926-5000

Multiknob Enhancements

April 18, 1997

Author: Ron Chestnut Subsystem: Multiknob User Impact: Small
Panel Changes: None Documentation: Yes Help File: None

Two new multiknob options have been added to meet some PEP-II Chromaticity requirements. They may also be of general interest. The Chromaticity package calculates corrector coefficients to knob the X- or Y- Chromaticity. The current settings of the correctors imply an actual value for the Chromaticity and the coefficients age quickly, i. e. the knob becomes stale fairly quickly.

Formerly only an ABSOLUTE knob could have a START value, since RELATIVE knobs always start at zero. Now if you specify a START value for a RELATIVE knob, this value is added to the initial 0 to give a meaningful display value.

A new qualifier, TIMEOUT, can now be specified to indicate a time, in minutes, after which the knob becomes stale. At this point any subsequent attempt to turn the knob will result in a timeout message and the knob is deassigned. This TIMEOUT time is measured from the time of knob assignment, NOT the last movement of the knob.

An example of a file using these two features is:

! The next two devices are ones I was allowed to play with
! while debugging - they have nothing to do with chromaticity!

MCC Whiteboard

4 June, 1997

Author: Keith Jobe Subsystem: Operations User Impact: Varies
Panel Changes: Few Documentation: No Help File: No

The ``WHITEBOARD" for current operations status is now operational. This contains the current operations program plus information relevant to physicists and machine maintenance personnel.

This display may be seen on video monitors in MCC, from the World Wide Web off our departments' home page at http://www.slac.stanford.edu/grp/ad/ad.html , and on monitors connected to the SLAC CATV system.

This display information is derived from a text file which contains the displayed information plus some formatting information. The display cycles the screens of information updating the display every 10 seconds. Modifying the file results in a new display within a few minutes. Lines in the text file reading `` *screen " force the following text to start on a new screen, otherwise the text is formatted onto multiple screens as required. Lines starting with an `` * "'s are ignored.

The display is a CUD display, assignable through the Next CUD Monitors. The normal destination for this display is a rather low-resolution (640x480) NCD X-terminal connected directly to a CATV video modulator. This display is listed as MCCX42 CATV.

Fast Feedback Panel for Changing Exclusion and Inclusion Masks

June 4, 1997

Author: Grossberg, Hendrickson Subsystem: Fast Feedback User Impact: Large
Panel Changes: One Documentation: No Help File: Yes

Fast feedback loops use beamcode modifiers in order to synchronize BPM acquisition on appropriate beam pulses. Scheduling of these modifiers for diverse operating modes has been a maintenance problem, requiring significant database work for each run. New software makes it easier to reconfigure feedback masks from the SCP without requiring DBEDITs.

Since feedback scheduling can be complicated and have subtle considerations, users must communicate any changes to Mike Stanek (MWS) and Linda Hendrickson (LJH). Novice users are encouraged to consult an expert before making any changes.

The new FBCKMASK panel, accessible from the Fast Feedback CALB/DIAG panel, allows the user

  1. to change exclusion and inclusion masks for fast feedback loops and for the measurement micros;
  2. to view masks and beam rates for the changed user masks as well as those from the database;
  3. to activate the changed masks.

Both exclusion and inclusion masks have the effect of limiting the pulses on which feedback loops receive measurements. Exclusion masks prevent measurement on any pulse in which the associated bits are scheduled; typically they indicate when the beam is not there. Inclusion masks prevent measurement on any pulse in which the associated bits are not scheduled. They are usually used to synchronize measurements between multiple micros or to select a particular type of pulse.

There are two rows of Add, Remove, Zero buttons for changing the feedback loop exclusion and inclusion masks by adding a beam code modifier to an existing mask, removing a modifier from an existing mask, or zeroing an entire mask. These are followed by three rows of Add, Remove, Zero buttons for changing the exclusion and two inclusion masks of the measurement micros of the loop. A restoration button is available to negate all changes at any time.

There are buttons for displaying all beam code modifier names, all masks for the selected feedback loop, and beam rates and patterns resulting from the masks. A toggle button associated with the display buttons allows the user to see displays based on masks currently in the database or based on the user-changed masks.

An activate button causes fast feedback to use the new masks. It writes them to the database, creates new bpm buffer rings for all associated measurement micros, coldstarts the feedback loop, and records the mask changes in the error log.

Configuration buttons exist on the panel, mainly for safety and documentation reasons. Currently these are generic buttons which create and load masks for all feedback loops and all micros. Eventually, loops may be divided into meaningful groups and, with the incorporation of an activate-config button and button macros, the mask configuration function would be a particularly useful tool. Meanwhile, it can be used as is to maintain a history of mask changes and to allow restoration of previous masks if needed.

MPSFAULT Display Enhancement

June 12, 1997

Author: Ron Chestnut Subsystem: MPS User Impact: Small
Panel Changes: None Documentation: No Help File: None

The MPSFAULTS CUD and SCP display have been modified to include information on the last four rate limit changes. This extra data is extracted from the MGRP History data, which itself is available from the Beam Control panel.

Only the time since the rate change and the new MGRP name associated with the change are displayed. The hope is that this name will always be selected so that it is meaningful to operators. Comments and suggestions will be gladly collected and collated.

Control Of PPS Devices

May 22, 1997

Author: Kenneth Underwood Subsystem: Control System User Impact: Small
Panel Changes: None Documentation: No Help File: No

Much of the Personnel Protection System (PPS) can be monitored and controlled through the SLC Control System as conventional digital devices. However, the PPS digital controls are unique in that a hardware permit located in the MCC control room is required for the PPS control function to actually occur. Unfortunately, an inadvertent PPS control function from a remote SCP coincident with the hardware permit caused significant disruption to PEP-II activities.

At the request of Marc Ross any attempted control of a PPS, STPR, or DOOR digital device will check for the presence of a new account privilege called SLC_PPS_CONTROL. If the account that attempts to control one of these privileged digital devices does not hold this right, the attempt is refused with appropriate error messages.

At present only the control room COW, MCCSCP, EOIC_DESK, and SLCSHR accounts have the right to control PPS, STPR, or DOOR devices. A very few individual accounts may be added as the need arises. Also, there are probably a few digital controls that are not properly defined. As these exceptions are identified, database changes will be made to correct these problems.

Back to top of this issue

May 22, 1997 Index Panel indx102

Translated from original PlainTeX by index2html.pl.

*Links followed by an asterisk are limited to SLAC clients only.
  Last modified on Wednesday, July 26, 2006 Webmaster