DØ Calorimeter Electronics Upgrade Meeting Minutes

November 4, 1999

present:

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

= ACTION ITEM

Pulser System

UB reported that testing of pulser system was going fine. They are looking at comments from D. Schamberger about the pulser interface card but nothing to report yet.
A request was made again for a pulse generator at FNAL for testing the new system. They were assured that people at FNAL will provide one.
There was a question of having access to the preamp boxes. LG reported on the Instillation meeting that is now being held on Mondays each week. There is some activity planned on the walkways leading up to the boxes - the floor muon chambers will be placed in these positions at some point, restricting access while this work goes on. This probably won't happen in the next week or two. They can work on one side at a time, giving us access to the other side. The other possibility is the resurrection of a "skywalk" or bridge that connects the East and West side platforms that runs between the CC and ECN. There is some solenoid surveying equipment that must be moved out of the way first.
The Instillation meeting should help us avoid any major conflicts.

DH has received a large wooden crate at FNAL with pulser fanouts, daughtercards and tools in anticipation of the Orsay group arriving next week. UB will be here Nov 8 as well.

BLS Systems No update - Dean away Power Supplies DH and PL reported that we are still waiting on wire. This is holding up production. The wire ordered from Newark seems to have got lost. Investigating.... The diodes also have not arrived yet.
Six monitor boards came in from Chipco - there are a few problems that are being fixed. Three are complete and are ready to be mounted.
Front panels are complete and in.
Once we have boards + wire etc., Don can probably help with the final assembly.

DH has a BLS PS on the bench but not opened up yet. He is trying to recover 10-year old CAD files to get the schematics.

DH also working on bracket for control power and solid-state relay for the pulser PS. He needs a connector from the French to connect to the remote sensing unit. At the moment the sensors are connected to the other side of the control board so do not measure the slight drop across the board going to the output.

SCAs DH has done some follow up on the 4-inch bonding pad issue. He spoke to his contact at SuperTech to get the composition, metalization and stacking information. From what he got it seems that there are two layers of metalization, both Al-Si-Cu with 6000 Angstroms for the first and 11500 A for the second. However, there appears to be no barrier metal under the first layer of metalization, which was added to the 6-inch production bonding pads. DH looked at the original design files from Kleinfelder and there does not seem to be a barrier metal layer. He will try and check the actual masks used to make sure this is the case. Accurel in CA can probably tell us the exact cross-section of each layer if it was felt that this was necessary.
DH and PL have examined the chips under a microscope and they do look different to the 6-inch.
In DH discussions with Supertech they did say that the cratering problem we are seeing when bonding can be related to the bonder and its setup, however ASAT has tried numerous settings to no avail. The general conclusion seems to be that this is a weak metal pad. SuperTech did mention a company in Thailand (CEI) that might be able to package - PL had contacted them in the original bidding process.

We are half-way through testing the 8000 6-inch wafers. Yields still about 60% on the large lot (Lot 02-03: 6156 total).

PL has re-examined the estimates of what is needed and what we have on hand. The table below summarizes the results. The ratio of UP/DOWN is 3/2.

SCA type	Needed (For BLS systems only with no spares)	4"	6"	Total	Deficit
UP	13824	7314	1888 + 3694 expected** = 5582	12896 (expected)	928
DOWN	9216	6196	5223	11419	2203 spare
Total	23040	13510	10805	24315

** Lot 02-03 has 6156 devices with an expected yield of 60% after testing about half.

The conclusion after a long discussion was that it was felt that pursuing the packaging of the 4-inch wafers appears to be a precarious road with no hint of success at the other end. We should actively pursue obtaining the 8 wafers that SuperTech has on hand which should yield 1300-1900 good devices. The packaging for these would be included in the original ASAT PO. Packaging these all UP will give us the total needed for the BLS systems with hopefully some spare.

Preamps LB has one of the H-test jigs just about done. She and Shua need to clone this to send to MicroElectronics which should take about a week. LB had to open up 4 of the windows slightly to accomodate her hand-made Hx preamps.

LaTech replied to the query why the output rise time on their tests seemed to be faster than the input. There was also some ripple on their test point 1 which might be due to an impedance mismatch. They are investigating.
LaTech did send a new layout for the traces on their motherboard which LB has forwarded to DS.

KTY still has not sent the 10 I samples. These will probably arrive with the shipment of D, E and F which shipped Oct 29. The only piece left then is C which has a delivery date of November 12 (next week!!! - there was much rejoicing and fatted calves were slaughtered in celebration :-) ).
A,B and G are all here in DAB. LB has looked at the test data for B and G (A data is missing and might have been sent to UTY). They seem to fall inside the specified windows. The suggestion was that after the C's are shipped, the two test jigs should be shipped back by KTY so that we can run a few of these preamps ourselves to ensure all is OK. After that we can sign off on them.
We also need to check the part count which CH and LG will look into next week.

LB did note that in making up the rest of the Hx test pieces that the R3 1% resistor on the He pieces from KTY preproduction were 167 Ohm instead of the specified 171 Ohm. We need to check R3's on the production quantity coming in.

No word yet from MicroEngineering. Expect the prototypes back about end of Nov (4-6 week ship date).

LG reported that all preamp motherboard parts were at Chipco, except for the connectors and some boards. They have 65 v3.2b red boards and 290 v3.2a green boards to start with. We have another 175 that MG is finishing measuring which will be sent next week. Samtec has started work on the connectors - 4000 should be shipped by the end of the week with another 10k for the next couple of weeks. We need to send them the remaining 25k connectors.

There was a brief discussion about the procedure for stuffing and testing preamps and boards once the production is in. This should start end November or early December. We will probably stuff one crates worth (96 boards) at a time and test them in the 5k test stand for 1-2 weeks before stuffing them in the detector. The CC will be done first as we still need to wait on the Hx preamps for the EC. LG will draw up the procedures needed. We need to find places for the stuffing and temporary storage.

LG spent a few hours with DS remotely getting the 5k test stand DAQ running again. There was a loose ground connection from the timing control box power supply.

LG and DH have been looking at the temperatures in the 5k test stand. Don Emory will move the temperature probes away from the metal side walls with ceramic spacers on the right hand side (RHS) of the test crate, similar to what was done on the LHS. This ensures we are measuring the air temperature rather than the box. LG has tried placing a baffle diverting the air intake of the fan to the left side of the box, and has put small 12V PC fans blowing across the top and bottom parts of the LHS box. None of these seem to have made much difference. With both sides of the box populated (96 boards) the temp can rise to 53 deg C in some places. Some boards are too hot to touch for more than a few seconds. LG will try out a similar test on the RHS as to the touch, it seemed this side might be a little cooler.

Timing and Control DE has basically finished debugging the timout command control block. The system runs at 132 nsec. The commands now need to be integrated into the complete system which might cause some problems with addressing. DE should still be able to work on this for another couple of weeks. The rest of the system was being relaid out by the SUNY-SB EE student so DE needs to check how to integrate his pieces.

This page updated 5-November-1999
Send comments/corrections to Leslie Groer groer@fnal.gov