Power Supplies and Electronics Procedures

The power supplies and readout electronics associated with the FPP are a mixture of commercially purchased equipment and equipment designed and/or assembled with the Rutgers University Department of Physics & Astronomy Electronics Shop. The reader is directed towards the manuals made available by the manufacturer for the detailed information not provided here for the commercial equipment. For the Rutgers constructed equipment, further documentation is available though the Rutgers web pages:

http://www.physics.rutgers.edu/npx/fppdir/fpp-homepage.html

and through FPP notebooks (try for example contacting R. Gilman for notebooks maintained by Rutgers, CEBAF Center, phone 757.269.7011).

The LeCroy 1458 HV control crate houses the Lecroy 1469P modules which control the HV for the FPP chambers. The 1469P has 3 master HV channels and each master HV channel controls eight slave channels. In slot 7 of the 1458 is the 1469P module which controls chamber 1 and chamber 2. In slot 8 of the 1458 is the 1469P module which controls chamber 3 and chamber 4. The individual slave channels can trip from high current faults or other trip faults, but all eight slave channels must be raised and lowered together by setting the master high voltage. The HV provides +1.8 - 1.9 kV nominal to each of the $\approx$ 5100 wires in the four FPP straw chambers. The power supply is located in the detector stack at the top of crate 6 in the upper electronics level. This unit is controlled through HAC13. Connections from the power supply to the chambers are made using standard SHV connectors mounted on red RG-59/U HV cable good to 5 kV.

The high-current low-voltage supply boxes were assembled by Rutgers University. They are designed to provide a maximum current of about 1.6 / 0.6 A at -5 / +5 V to each of the 318 readout cards on the four chambers. There are 63 / 63 / 90 / 102 cards on chambers 1 / 2 / 3 / 4. Typical operating currents are about two-thirds of this nominal maximum value. The +/-5 V power lines are independently fused to each card. Each of the eight supply boxes contains two or three power supplies, each rated for either 35 or 50 A. There are two power boxes for each chamber. Six boxes are located at the lower rear end of the detector stack. The second boxes for chambers 3 and 4 are located at the top of the detector stack, on an aluminum plate just off the upper electronics level. These power boxes are monitored through EPICS, but turned on/off though front panel switches.

Hewlett-Packard 6111A power supplies are used to provide typically 2 - 3 mA current per readout card. Each of the front and rear chambers have their own power supply. The front chambers thresholds are fused, to limit current drawn in case of a short on the board. The rear chamber cards use a 1.5 k$\Omega$ resistor external to the board to limit current drawn, in case of a short on the board. Board threshold circuitry also has a 1.5 k$\Omega$ to ground which with the external 1.5 k$\Omega$ makes a voltage divider. Therefore, the rear threshold supplies are typically set to a voltage which is a factor of two larger than the front threshold supplies to give the same threshold voltage at the readout board. Initial tests indicate that at least a 1.5 V threshold must be applied to the cards to prevent oscillations - this level will stop oscillations that arise when the voltage applied is reduced to about 1.0 V. In practice it has been found that the front chambers should be operated at 4 V and the rear at 7 V. Efficiency studies show that the chamber threshold could be raised by 50% with minor loss in efficiency. The HP supplies are also mounted in the hadron arm detector stack, on an aluminum panel located beneath the two upper high current supplies.

Each straw wire contains a 25$\mu$m $\phi$, Au-plated tungsten-rhenium wire. The number of wires per plane varies from 176 to 272. Wires are multiplexed 8 wires into one electronics channel, leading to a required 636 TDC channels. In practice a few extra channels are used, so that each 34 wire (16 differential signal channels plus one ground pair) twisted pair cable contains only signals from one of the four chambers. LeCroy 1877 multihit FASTBUS TDCs are used to measure the leading edge time and width of the pulses, to demultiplex the wire hit. Within each group of eight wires, the widths are set to about 25, 45, 35, 55, 90, 65, 105, and 75 ns. The TDCs are located in the upper FASTBUS crate located on the lower electronics level of the spectrometer space frame in the detector hut. The FPP rack, containing level shifter cards, is located opposite the FASTBUS crates on the lower electronics level. It shifts signals sizes from the reduced $\pm$50 mV readout card output levels to ECL standard levels, for input to the TDCs. The connections between the readout cards and the level shifter cards, as well as between the level shifter cards and the TDCs, are made with 16-conductor twisted-pair cables. A wiremap, detailing the cabling, is posted on the side of the FPP rack.

Power-up Procedure

1. Ensure that gas flow has been established in the chambers as outlined in the previous section. If it has not, STOP RIGHT HERE! Gas flow must be well-established and steady-state BEFORE the HV may be enabled.
2. Ensure that all power supplies as well as the FASTBUS crate are off and the LV, HV, and TDC cables are connected.
3. Turn on the threshold and LV power supplies. Use EPICS to turn the threshold voltages up to correct values, about 4.0 V for front chambers 1 and 2, and 7 V for rear chambers 3 and 4.
4. Use HAC13 to turn up the chamber voltages. Standard values are 1875 V for front and rear chambers. It is probably best to raise the HV in 300V steps. After each step wait for the current to settle below 1 $\mu$A, then go up to the next level until 1875V is reached. Peak currents during turn-on should not exceed about 40 $\mu$A. A 10 V/s ramp rate leads to a leakage current of several $\mu$A. Trip levels should be set to 110 $\mu$A both for turning on HV and for normal operation, so that bad spills do not trip the chambers. Current should settle to about a $\mu$A or less within a few minutes. If the power supply trips during the ramping procedure, it is possible that you are moving too fast, or that some problem has developed with a chamber. Rezero things and begin the procedure again. NEVER USE THE AUTO-RESET FUNCTION. If the power supply trips again, STOP IMMEDIATELY AND INVESTIGATE. There is probably a problem and expert advice may be needed. Some detailed information, intended for experts debugging hardware problems, is available in the Rutgers web pages.
5. Check for poor signal connections evidenced by hot wires (wires counting extremely fast) or dead wires (wires with no counts) using the histogramming software and cosmic rays. Be careful: apparent problems may result from bad demultiplexing rather than from poor signal connections. Remake any connections as necessary by first powering down the FASTBUS crate.

If at all possible, the HV and LV power supplies should be left on continuously if and only if gas is available to the chamber. This avoids time loss to reconditioning and maintains the desirable steady-state operating condition.