From breuer@enp.umd.edu Thu Nov 7 00:34:52 2002 Date: Thu, 7 Nov 2002 00:47:55 -0400 From: breuer@enp.umd.edu To: nakahara@jlab.org Cc: breuer@enp.umd.edu Subject: HV shutdown board manual status: 06-Nov-2002 (preliminary) MANUAL FOR G0 HV-PROTECTION CIRCUIT =================================== Question: direct to Herbert Breuer (breuer@enp.umd.edu, Tel. 301-405-6108) short specific version: ----------------------- - The GREEN HV PROTECTION CIRCUIT BOARDs are located on a 19-inch wide aluminum plate mounted in one of the racks between the HV power supplies at approximately eye hight. The circuit boards have the components visible and are not protected from direct contact. - Of the two boards only the right board is currently in use. - The six LEMO connectors in the bottom row are for input from the photomultipliers, channels 1 to 6; details are in the table below. - At the center of the board above the respective input is a blue potentiometer with an adjustment screw on top; this potentiometer sets the threshold level: to the right to smaller thresholds, to the left to larger ones; the smallest threshold value is about +100 mV, which corresponds to -4 mV on the 50 Ohm imput for the NA inputs and -100 mV for the French. - On the right side of each potentiometer is a golden-top test point sticking up where the threshold can be measured with a volt meter; note that the reading is positive. - If the GREEN LIGHT is on, the channel has NOT tripped; thus any missing green light identifies that a specific channel has tripped. If any channel trips then the board's interlock output will trip and the RED LIGHT goes on. If the output "RELAY OUT" is connected to the "INTERLOCK" input of the CAEN HV power supply with the TOGGLE SHWITCH in the "H" position, the HV powersupply will be disabled and the HV will go down immediately and cannot be ramped up again without a reset. - The RESET BUTTON is a small black button on the left upper corner of the board; pressing the button will restore all green lights, turn off the red light and removes the interlock from the HV supply. The HV then can be turned on in the normal fashion (e.g. via the HV GUI). - On the right side of the board are 4 lemo connectors; two TTL inputs labelled A and B on top and two outputs on the bottom (one TTL out and one RELAY OUT). - One of the TTL inputs on top receives the G0 SMS MAGNET INTERLOCK cable. The G0 magnet interlock prevents the HV power supplies to be used below 1000 A of magnet current. Below 1000 A the red light will be on (with most likely all green lights on) and the board's output will be in the "inhibit" or "interlock" state. - When the G0 SMS magnet current has rampred up to above 1000 A the magnet operator has to RESET THE MAGNET INTERLOCK via the magnet GUI interface. This will remove the interlock state from the board and from HV without further manual intervention. = DO NOT REMOVE THE G0 MAGNET INTERLOCK CABLE; if necessary ask the magnet operator to lower the current trip value. - The CAEN HV INTERLOCK CHAIN via Lemo cables starts at the RELAY OUTPUT (lemo connector) on the right side of the interlock board (lower left of the set of 4 lemo connectors). The lemo cable is daisychained via lemo-Ts to all CAEN HV power supplies to the "INTERLOCK" input; the switch on the HV supply must be in the "H" position. If the light on the HV supply next to the input is red, then the HV supply has tripped due to the input from the interlock board. The specific cabling and values are given below; the anode current trip points are set to 100 micro-A (i.e. a net value of 1250 mV for NA and 500 mV for French detectors). channel patchp. connecting detector zero trigger input cable cable offset threshold 1 529 28 01-14FL +120mV +1370mV 2 530 27 03-01FL +120 +1370 3 531 26 05-06FL +60 +1310 4 532 25 07-10FL +120 +1370 5 FR ampl. lemo 02-16FL +1mV +500mV 6 FR ampl. lemo 06-16FL -3 +500 TTL 39* 20/19 G0-magnet -- short to ground * cable 39 is on the electronics side of the racks and has a label "MAGNET POWER INTERLOCK". More detailed information for the interested person --------------------------------------------------- [this is currently just a list of items, with comments about what to change for a final board to be produced later; the present board is a functioning prototype] inputs are switchable between 1/1 and 1/25 attenuatation, via the white 3/4 turn microswitch (1.25 mV/uA input or 0.05 mV/uA; an amplifier before input recommended) the input is filtered (integrated) with a time constant of about 60 milliseconds to avoid trips on short current spikes amplifier times 25 with positive output for negative input discriminator with setpoint as described above range about 100 mV to about 11.5 V discriminator output triggers flip-flop, which keeps output state and green light state unfortunately most channels (except of channel 1 of board 1) will trigger again on transition to below threshold [need all channels permanent trip unless safe to reset] unfortunately also: you can reset the board while the input is above threshold; this should not currently be a problem since presumably all HV's have tripped the amplifier output can be checked on foot number 6 (bottom IC with 8 feet, top row of feet, second from right) [need test point] [the discriminator should have a zero offset to allow for both polarity input offsets; eg 1V offset, with effective threshold -900 to +10500 mV] the two TTL inputs can be set to default "true" or default "false" via white micro switch (3/4 turn); since we do not use TTL input from the magnet, but rather a relay to ground at the magnet power supply, the default "true" must be chosen. On magnet interlock set (short) the TTL will fall to "false" triggering the interlock output. For daisychaining of boards the TTL input should be set to default "false" so that a "true" input from the TTL output of the previous board is required. The "relay" output should be used to trigger the HV interlock; the relay will go to closed (a "short circuit") if the board detects an interlock condition (red light on). Note: the current magnet interlock is not the preferred fail-safe type. Currently, a "short" signals an interlock condition; this can be cleared by removing the cable, which is not a fail-safe setup. The logic needs to be inverted in the future, but the board currently does not possess the required input logic. The required logic can be simulated with one of the analogue inputs and a -5V level put in parallel via a lemo-T at the input of the board. Board #1 has a lemo cable with -5V through 500 Ohm for such a purpose. Set the threshold level to about 5V in this mode of operation. ________________________________________________________________________________ Dr. Herbert Breuer, Dept. of Physics, Univ. of Maryland, College Park, MD 20742 Tel.: USA-301-405-6108; FAX: USA-301-405-8558; Department-FAX: USA-301-314-9525 E-mail: breuer@enp.umd.edu ________________________________________________________________________________