DSSD Primer


Silicon strip detectors have been used behind the focal plane of the FMA in a variety of applications. The highly segmented dssd's for implantation/decay studies have been used the most. Fusion cross section measurements also benefit from the segmentation since it allows you to measure the detection efficiency after the focal plane. It is impossible to completely describe all the intricacies of the strip detectors used behind the focal plane of the FMA, but this document should answer some basic questions. To know more you should talk to Cary Davids.

All Shapes and Sizes


The DSSD detectors have a dedicated set of electronics which includes hybrid preamplifiers and amplifiers/discriminators built by Rutherford-Appelton Labs in England. There are 2 sets of amps, one has full-scale gain of 200 MeV and is used to record implantation events in a set of Phillips 16-channel peak-sensing ADC's. The other set has full-scale gain of 20 MeV and is used to record decay events in a set of Silena 8-channel peak-sensing ADC's.


The amplfiers do not have adjustable gain, but are within 10% of each other. The gain-matching is done in software. A Pu/Cm alpha source in front of the dssd provides a calibration for external alpha particles. A reaction with known proton and alpha emitters is also used on occasion for a more accurate internal calibraiton. See the section below on performing an on-line calibration with the Pu/Cm source.



Removal of Other Detectors Behind Focal Plane

Depending on what is behind the PPAC, varying amounts of removal and installation are required. If the ion chamber is there, then it and any spools must be removed. If the flange which has the gate valve is already in place, then there is much less lifting to do.

Preparation of PPAC

Preparation of DSSD Chamber

If it is already mounted on the flange/gate valve assembly, the crane must be used to lift it all in a safe and controlled manner. Use 2 long slings, one on each side of the gate valve. The gate valve points up, but is not exactly vertical. Insert a brass pin in the top bolt hole on the back of the PPAC housing. Slide the flange/valve assembly onto the pin. Insert the rest of the bolts. Remove to remove the pin before tightening bolts too much in case it is stuck. Tighten all the bolts while pushing up on the big flange to center it. When you are done, it looks like this.

Preparation of DSSD Electronics

The two racks (amplifiers and CAMAC crates/NIM bins) are usually bolted together so they can be rolled easily without damaging the cables which go between them. They should sit just to the right of the PPAC gas-handling system. The VME crate rack for the data acquisition system sits to the right of everything else. Roll the racks into place. Mount the preamps on the back flange of the DSSD chamber like this.

Making Connections

On-Line Calibration Procedure (DSSDFIT)

Crude peak-finding routine for one 2d histogram which contains the energy spectrum for each strip on one face. There are up to 5 such spectra to deal with: ERF (1k), ERB (1k), EDF (2k), EDB (2k), EBOX (2k). To calibrate them collect a spectrum with ICAL=0 and no gates so any previous gain corrections are turned off. Once you have a high statistics spectrum, you can run the routine which computes the gains and offsets. Then apply them, turn ICAL on, reapply the gates, zap, collect again and see if they are gain-matched.
OUTLINE:          $FIX ICAL 0 (for 1st pass only)
                  $G2H/COND=(ONED,EDIFFD,DEVETO)/DEL EDF (for 1st pass)
                  $G2H/COND=(ONED,EDIFFD,DEVETO)/DEL EDB (for 1st pass) 
                  $G2H/COND=(ONER,EDIFFR)/DEL ERF (for 1st pass) 
                  $G2H/COND=(ONER,EDIFFR)/DEL ERB (for 1st pass) 
                  $ZAP/2D EDF,EDB,ERF,ERB
                  $UH DSSDFIT EDF/2D  (answer the questions)
                  $@DF_FLT  (loads the gains)
                  $@DF_FIX  (loads the offsets)
                  $UH DSSDFIT EDB/2D
                  -to skip recoil part, go to "$FIX ICAL 1" step.
                  -plug the pulser in to gain match the recoil electronics.
                  $FIX PULSER 1  (takes into account that all chan fire)
                  -collect a spectrum which puts the pulser peak at half of
                  fullscale (512 on 1024 is roughly 100 MeV). Flip
                  a 2x attenuator switch on and collect some more. This will
                  be roughly 50 MeV (ch 256). Remember that the 2D spectrum
                  is used for matching and it has only 1/48 as many counts
                  in each peak as the 1D spectrum so count for a while.
                  $D1 ERF (or ERB)
                  -use SUM to find the centroid of each peak . It will be a 
                  mess since it contains 48 unmatched strips. But taking
                  the centroid of this blob to compute the "energy" being
                  simulated by the pulser allows you to make the smallest
                  correction overall.
                  -compute the "energy" of the 2 peaks assuming that 200 MeV
                  is fullscale. On 1024, Epeak=CHpeak*200/2048.
                  $UH DSSDFIT ERF/2D 
                  -for energies enter the two values computed above.
                  $UH DSSDFIT ERB/2D

                  $@FIX ICAL 1 (turns on gain matching)
                  $G2H/COND=(ONED,EDIFFD,DEVETO) EDF (for 1st pass)
                  $G2H/COND=(ONED,EDIFFD,DEVETO) EDB (for 1st pass) 
                  $G2H/COND=(ONER,EDIFFR) ERF (for 1st pass) 
                  $G2H/COND=(ONER,EDIFFR) ERB (for 1st pass) 
                  $ZAP/2D EDF,EDB,ERF,ERB
                  look at recoil pulser spectrum
                  $FIX PULSER 0
                  unplug pulser to check decay stuff
                  look at gain matching and decide if it is good enough
                  You can do a second iteration by answering with a 2
                  when the program asks you which iteration. 
NOTES on the answers to some of the questions for DSSDFIT:
Last Updated: June 3, 1996 (D.J. Blumenthal, djb@sun0.phy.anl.gov)