DØ Calorimeter Electronics Upgrade Meeting Minutes

March 16, 2000

present:

Meeting time is 9:00 am, Thursdays in DAB1. Video connection available.
Contact Leslie Groer groer@fnal.gov.

= ACTION ITEM

BLS Systems

DS reported that he understands some of the previous problems with the T&C board. One of the signals was taken on the wrong edge of the 132ns pulse.
He is having 20 more prototype BLS motherboards made with some modifications. One of these is a trace that was too long. He expects boards back by end of March and then another 2 weeks (???) for assembly by Chipco - he still needs to get parts together for the assembly, but all should be in stock.
There are some changes as well to accomodate the high eta regions where there are actually only 44 channels and not 48.
These boards will have jumpers.
DS has received 10% of the BLS daughterboards (that hold the SCAs) - 500 boards. These also will be sent to Chipco for assembly and expect 2-4 weeks for these.
DS has enough right-angle limo connectors for the BLS crate controller cards that take the trigger signals for the 5k test area. Six are needed per quadrant - we need 80 eventually which he needs to order.
A glitch in the T&C was fixed where a signal needed resynching to the clock. This *might* have fixed the difference in the pulse height measurements that were seen between December and January.
The fanout module is going well. The layout is done and the boards are being etched at BNL. There are 3 boards with lots of connectors on each. Two test boards are built and tested and one more is being made. This then needs to be tested with the fanout power supply. Once this is working we can drive a whole quadrant with the T&C instead of just one BLS crate.
There have been a fair number of changes to the T&C layout. DS needs to get his CAD system to autoroute the new signals but the system has trouble with the tighter rules and 6 mil pads and 12 mil pitch near the FPGAs and some of this needs to be done by hand, which takes a while. These are 11"x14" boards.
DS will need 1000 down and 1500 up SCAs to populate the 500 SCA boards when they get to him in 2-4 weeks. PL and LG will get these together, along with the labels that were made by Shaohua Fu last summer. Power Supplies PREAMP PS: DH has measured the power load on the preamp supply connected to box 10 in situ and it looks like we can get away with minimal rework. The 5A breaker switches for the supplies are being relaced with 10A breakers but we can get by with two AC distribution boxes and do not need to replace the plugs or receptacles.
A fuse block has been ordered to try and do the rework of the preamp DC power distribution after the power supplies. The 10 gauge wire has also been ordered directly from Dearborn. Anderson will send a sample of the 45A pin connector to try out with the existing 30A receptacle. We would like to avoid having to shave any strands off the 10 gauge wire to get it into the pins.

BLS PS: A second supply is under construction. This should be ready by next week when DS is here to test with a full BLS crate in the 5k. DS still has to add the trigger summers to the BLS cards which put a load on the +13V line. There are 8 adders + 2 drivers per board. These can be emulated with resistors.
The water cooling in the 5k needs to be fixed up to allow the test supply to run. This is not of concern in the real detector as DS claims there is plenty of flow available. We also have extra pipes available to get separate cooling for the PS and the crates if need be.
Mike Cherry has pulled some of the pipes out that need rework and has had the ends sweated off. We will replace these as we need them.
DH and DS believe we are still a couple of weeks away from being able to sign off on the BLS transformer order. The order has been placed and is waiting on us to say "Go!". We should probably check what the lead time will be once we give the go-ahead.
PL has been modifying the monitor boards and writing up procedures. He estimates about a day a piece which might get up to 1.5/day as one becomes familiar with the procedure. He will check with Bob Jones group if it is possible if they can do any of these (we have about 80 to do).

PULSER PS: No movement on these supplies as awaiting the control cable and connector.

ADC PS: The -5,2V rail in the ADC PS in the MCH for boxes 10 and 1 keeps tripping. This supply needs to pulled out and swapped which Mike Cherry and Victor Minor will do tomorrow. Also the supply for 11 and 0 occassionaly loses the +5V interlock which is fixed by a small modification that can be done in place. PL will look into these.

PL reported that the remaining monitor boards should be back from Chipco assembly at the end of March.

Preamp Testing DH measured the new cooling fans' speeds in a B-field. For the small fan there is about a 5% drop in the speed (nominal speed is ~2865 rpm) going from the horizontal position to the vertical for a field of about 275 Gauss. For the large fan the drop is less than 2% (3440 rpm nominal). The B-field vector in the position of the preamp box is closer to the horizontal than the vertical. For the test the small fan was left attached to the heat exchanger. DH had to jury-rig the readout as the token ring readout of the 1553's does not work anymore. We need to migrate pieces over to the new monitoring and control system soon. Unfortunately, data-logging is still missing.

LG got the B3run system on the 3rd floor working again after DH added an extra #10 cable to the +5V line from the ADC supply. There was about a 0.1V drop on both the power and return line and the crate backplane was sitting at about 4.7-4.8V instead of 5.0V which made the Bit3 board very unhappy.

MG will continue testing stuffed preamp boards. All the CC boards are done and over 100 EC boards are done. Chipco had a delay in production after "losing" a reel of resistors but is now in production again and we expect shipments to start again next week. We have about 60-70 of the v3.1 motherboards which can be used in the meantime.

LG will test some of the Hx's in the test setup now that it is working again. We have received almost all the H's from MicroEngineering and are now in "spares" production as there originally delivered quantities fell short about 10%. We will also begin using these in board stuffing next week.

DS is concerned that he saw only about 5.3V full scale from the preamps on average during his tests last week. We had designed for closer to 6V for the full dynamic range (this was the original incentive for going to +12V on the output power rail). This means we only have about 0.3V of pile up before things become non-linear for large signals. DS is worried that the bias for the transistors is dependent on the temperature, which will make cooling the preamps more imperative. DS recalled that some measurements were taken by a Columbia summer student a few years ago (Abdrashitov?) and a D0 note was put out, which we need to try and relocate.

PP wondered about the tests with the grounding of the pulser command cables that had been discussed at our previous meeting. DS did not have time to do these last visit as is still not totally sure how the tests were conducted in France. The grounding issues in the 5k test system need to be resolved to finalize the tests.
DS also claimed that he saw a good ECL signal with the correct rise time but it looked like there was 50 Ohm termination at the pulser fanout end, instead of the required 75 Ohm to match the yellow command cables. If this is the case, then about 1/2 the signal could be reflecting.

SCAs PL reported that we have received all ~5400 UP SCAs from the final production of the 4" wafers (six lots). Annie has been testing, but she was away most of this week. Results so far:

**estimated
Lot no.	Total	Tested	Good	Yield(%)	Expected number
171	811	171	76	44.4	360
282A	227	227	72	31.7	72
284	956
285	1140
317	1308
318	824	824	526	63.8	526
Total	5266	1222	674	55.2**	2904**

From the previous productions and testing, this is what we have in hand:

SCA type	Needed for 13 BLS systems (includes spares)	4"	6"	Total	Deficit
UP	14976	7314	5232	12546	2430
DOWN	9984	6196	5223	11419	1435 spare
Total	24960	13510	10455	23965

. If the yields hold out for the rest of the testing on the new batch, we might be OK! (fingers crossed). Control & Monitoring We will need a few dozen short sections of blue 1553 cables to connect the 1553 rack monitors for the pulser power supplies + other miscellaneous. PL said that Annie and Victor have terminated these before. He has ordered connectors and barrels.
LG has got the information to the monitoring group to get all the ADC PSs and crates into the database and monitored and will test these today. DS, LG, Russell Gilmartin and Alan Jonckheere have been working on getting the download scheme working for the ADCs and VBDs. Pieces exist to test out next week when DS is here.
We have various monitoring tools available but the device names have been changed (once again!) so some of these need to be modified slightly. We are nearing the end of defining devices for all the existing pieces of the hardware. Only the new pieces remain (T&C, PIB etc). Some of these tools are being used during the DAQ test runs and need to be documented which LG is working on with the online controls group.
LG is consolidating the 1553 cables and swapping out bad 1553 devices as he finds them. Cabling AS reported that the ICD group have tested their boxes with a 3 drawer system and 6 preamps and seen a maximum output of 3 Watts. They are not really concerned about cooling but will put monitoring temperature probes on the outside of their boxes and run extra cabling through the winders to read these out.
All the tiles have been manufactured and are in testing at UTA. They expect to start installation after April 17 once the forward preshower is done.
AS is working on understanding the extra space needed for the proposed chutes going to the EC winders. There are some RG58 black cables which DS suspects (ed note: and conversation with on Friday Richard Partridge confirms) are Level 0 cables that sit inside the iron bar that supports the chute going to the winders. It appears the Level 0 cables will not be used in Run 2, but this needs to be checked next week when DS is here. This space also houses some of the caloimeter HV cables (red) and on the ECN, the pulser current cables. The plan is to install a chute that would bury these cables if they stay in their present location. It is not clear if it is even possible to move these cables so burying might be the only option. One option might be to rerun the pulser current cables in these new chutes, but this may be tricky as they are terminated and connected on both ends already. Burying all these cables is probably not of concern, but space should be left available in the new chutes in case any new cables need to be pulled.
AS also reported the ICD signal cables have been pulled from under the floor in the CC platform and stowed on the sides in holes in the iron support beams.

This page updated 19-March-2000
Send comments/corrections to Leslie Groer groer@fnal.gov