Decarad System

Procedure to move decarads

LCLSII-4.5-PP-2113-R0, Placement of Radiation Sensors for SRF Cavity Commissioning

CEDOG lock out

Introduction

The Decarad system was developed at JLab - it consists of 10 radiation detecting heads (geiger-mueller (GM) tubes) that connect (signal & HV cables) to a central box. That box has a board inside for changing the HV level and reading out the GM tube signals. The center box is connected to a digi port server via DB9 at box to RJ50 at the digi, which is then connected to the outside world via ethernet. The software uses the epics tool Stream with an asyn connection to the digi to send commands to the center box to read and write. The software is part of the RadiationMonitoring slac epics package.

The plan for commissioning is to attach the 10 heads to a frame on wheels (made from 80/20) and put the digi and central unit on a cart and roll the system next to the CM under test. In the navigation bar to the right, under the decarad heading is a link to the page with the diagrams showing the nearest ethernet, 120V, and 208V outlets to each CM.

Photo of 1 center box plus one radiation monitor with ruler

Help for Users

From the lower left corner of lcls home SC gun - X-ray/MISC+Global - select Decarad 1 or Decarad 2 to get the pydm display. It shows the read backs of the high voltage, the firmware number, a heartbeat, and the individual channel radiation read backs in approximate mr/hr, et al. It also has buttons to open a striptool for that decarad's dose rates and an archive viewer with the readbacks. Also on the pydm display is a place to type in the location of the system ("near CM35" for example) as well as positions for the individual heads ("near cavity 1", "between cavitiies 7 & 8", for example). These typed in values as well as the desired high voltage setting is saved across IOC reboots. All numeric values except for the heartbeat and firmware number are archived.

Showing path to get main decarad display, strip tool, and archive viewer.

Main decarad display.

n.b. The GM tubes saturate with high radiation:

Screen shot of saturation curve from LND 714 datasheet - attached below.

The radiation read back PV names are RADM:SYS0:x00:GAMMA_DOSE_RATE, where x is 1 or 2.

About the software

The decarads are talked to through a DB9 connector connected to a Digi Portserver. The software uses EPICS ASYN and Stream Device to send the commands to set the HV, read back the measurements, etc. The software is part of the RadiationMonitoring slac epics package. The source code can be reached from a development machine (currently afs): cd $EPICS_IOC_TOP/RadiationMonitoring. The database code is in RadiationMonitoringApp/Db/common in three files: decaradChannel.template, decaradCommon.template, decarad.substitutions. The file with the read/write commands is RadiationMonitoring/protocols/decarad.proto. The startup files are in iocBoot/sioc-sys0-rm01. The ioc status & control can be seen from LCLShome, SC gun, Network/Global under RadiationMonitoring sioc-sys0-rm01. The IP addresses are ts-sys0-nw01 and -nw02.

The database files have the PVs divided by per-head and per-system:
decaradChannel.template - records associated with each GM head
decaradCommon.template - records associated with the system
decarad.substitutions - substitution file for templates

For archiving:
RadiationMonitoringApp/srcArchive/common/decaradChannel.tpl-arch and decaradCommon.tpl-arch
RadiationMonitoringApp/srcArchive/lcls/sioc-sys0-rm01.sub-arch

PV names:

PV Name	Purpose	Save/restore	Archive
$(P):DEV_FW	read firmware number
$(P):HEARTBEAT	read heartbeat		10s
$(P):HVSTATUS	read status of high voltage		10s
$(P):CHAS_TEMP	read temperature of chassis		10s
$(P):HVMON	high voltage reading (V)		10s
$(P):HVCTRL	high voltage on/off 0 = on, 1 =off	yes	10s
$(P):LOCATION	String to hold a description of the rack position	yes
$(P):nn:GAMMA_DOSE_RATE	read back of dose from GM tube nn (01-10) in mR/hr (close enough, saturation curve starts to roll over about 10R/hr - see plot below)		1s
$(P):nn:POSN	string to hold a description of the position of GM tube number nn	yes
$(P):SEQ	sequence to readout the PVs at 1 Hz. The SEQ record has a delay of 0.06 seconds between sending 1 to each PV's PROC field so commands don't collide in the cable

There are two sets of decarads (10 heads + central unit) so the PVs listed above are doubled and differentiated by $(P).
Suggested $(P) = RADM:SYS0:100 or RADM:SYS0:200 for the two systems.

How to set the HV

From: Larry King <king@jlab.org>
Sent: Wednesday, March 9, 2022 12:49
To: Nelson, Janice L. <jnelson@slac.stanford.edu>
Subject: RE: decarad hv settings

Hi there, Janice,

No bother at all, I’m happy to tell you what I know about the HV setting.

I’m assuming you’re using the same G-M tube as the original design. The data sheet for the LND714 tube says the operating range is 450-650V, and they’ll work as low as 380V. Setting the HV is really just a balance between longevity and sensitivity. The tubes are only good for a limited number of interactions, so the higher the voltage, the more interactions and the shorter the lifespan.

When we first rolled the DecaRads out in 2010, we picked 650V for max sensitivity. After a few years of replacing tubes, we opted for a lower bias of 500V. While the cost of the tubes is not exorbitant, the downtime for tube replacement can add up. Since the 10 tubes share a common HVPS, when one tube dies it often affects the response of the remaining tubes. Sometimes the dead tubes just stop responding, but other times it may also present a significant load on the small HVPS and drag down the HV for all.

The HV for each chassis is set with a small pot on the front panel, and a DMM on the 1000:1 BNC connector. There’s no feedback in the bias, so when the heads are loaded down by radiation or tube death, the HV droops. Also, be extremely careful tweaking the pot. It’s a very small, surface mounted potentiometer that is easily popped off of the PCB with too much insertion force with the tweaker, OR just by rotating the tweaker too forcefully, if the pot and the hole in the chassis are off-axis. Please pass this message along to whomever might be doing this adjustment. I’d recommend taping up the hole after adjustment, just to discourage the casual tweaker. You know the type.

Good to hear from you. Just let me know if I can help.

Regards,

Larry

To Do:

Order from
TECO - 80/20
Get db9 cables from Doug
Test digis in tunnel
- if cables ok, roll up other cable on reel
Test a system in the tunnel
See what values HVSTAT gets
Test software: displays, save/restore, archiving
Put caps on frames

Cable Connection

Notes on cable connection (from email 8/12/21)

RJ50 cable/breakout (digiport server side) DecaRad Chassis DB9
Pins Pins
1 blue/wht TX- 2
2 org/wht RX- 8
3 org NC
4 green/wht NC
5 green TX+ 7
6 brown/wht RX+ 3
7 brown SGND 6
8 blk/wht NC
9 black NC
10 blue NC

Hardware Components

See LND 714 datasheet linked below for description of rad monitors. Docs below also have drawings of central unit hardware.

Screen shot of saturation curve from LND 714 datasheet.

Using Decarad frame.pdf as the basic design

Thing	source	part number	Qty for 2 frames	total	Ordered	Received	Assembled/ Tested
1.5x1.5" lite smooth 80-20	tecotechnology.com prices from 8020.net	1515-LS-96	10@72" 12@14" 12@24"	$7/foot comes to $700 $658.56	x
flange mount for caster to cross piece I bought 100 3458 bolts	tecotechnology.com	2426 attach kit: 3458	24 flanges 4*24 bolts/nuts	24$23 424*$1 = $650	x
caster I bought 100 3110 bolts	teco tech	2323 attach kit: 3110	24 4*24 bolts/nuts	24$30 424*$0.3 = $750	x x
5 hole tee plate for lower cross piece to vertical piece I bought 190 3320 bolts	teco tech	4340 attach kit: 3320	12 5*12 bolts	12$10 512*$0.77 = $175	x x
7 hole corner between end verticals and first upper horiz	teco tech	4352 attach kit: 3320	4 7*4 bolts	4$10.34 74*$0.77 = $65	x
6 hole T to connect upper horiz with verticals in the middle	teco tech	4310 attach kit: 3320	8 6*8 bolts	8$10 68*$0.77 = $125	x
4 hole corner bracket to tie decarad heads to - only need two bolts per	teco tech	4301 attach kit: 3320	20 20*2	40$5.12 402*$0.77 = $275	x
teco tech total (bob's total was $1k)				$2571.48 quote	o
wheel to hold 150' ethernet cable	https://www.amazon.com/Bayco-KW-130-Heavy-Duty-Capacity/dp/B01MRX24FD/ref=sr_1_6?crid=1DHMSFXA5VJ3W		2	2*$30=$60	o	xx	1 assembled
50' retractable extension cord with 4 plugs	https://www.amazon.com/WEN-PC5043R-14-Gauge-Heavy-Duty-Extension/dp/B07M8Q1MJR/ref=sr_1_2?crid=330APPDZ4XWXX		2	2*$50 = $150		xx	2 work great
150' cat 6 ethernet cable	https://www.amazon.com/Outdoor-Waterproof-Ethernet-Direct-Shielded/dp/B001B6C5H8/ref=sr_1_3?crid=19LWORAQ0ZOR6		2	2*$50=$100	o	xx
cart	mcmaster-carr	2602T24	2	2*$283.23 = $575	o	xx	2 work great
velcro zipties for decarads 15"long by 1/2" wide 5 for 9.73 green	mcmaster-carr	6605K14	40	8*9.73 = $78	o	x	got 'em!
velcro zipties for cables 11" long by 0.18" wide 10 for 8.39 blue	mcmaster-carr	6605K13	50	5*$8.39 = $45	o	x	Need to test with a real decarad
~~power strip with GFCI~~	~~mcmaster-carr~~	~~7693K49~~	2	~~2$87.31 = $175*~~
DB9 male connector	L-com	C&P9M	4	4*$1.10 = $5	o
DB9 pins	L-com	CPM10	4	4*$3.91 = $20	o
DB9 hood	L-com	SDC9AG	4	4*$3.81 = $16	o
L-com total				5+20+16=$41	o