LAr Pix ASIC Testing at Caltech Leon Mualem

LAr. Pix ASIC Testing at Caltech Leon Mualem Caltech

LAr. Pix. V 1 Socket Testing interface issues • Small changes to scripts for new board - Single chip instead of 4 (for now) requires different config file • Cable offset on socket boards go the opposite from the 4 -chip board (Doesn’t use 45 -50, instead of 1 -6) - Crimp a new cable, and away we go • Chip ID dip switch needs smaller resistor - 10 k seems to be too high, there is a big enough drop that it does not reliably read the address. (Data packets report chip_ID with one or 2 different bits. ) Leon Mualem Jan 9 2020 LAr. Pix Mtg 2

LAr. Pix. V 1 Socket Testing performance issues 1. 2. 3. - - I can change chips and still have the system work. • Nice little sockets these are. I ran through 2 chips and found very similar performance Swapped back and forth One of them now has some non-functional channels (oscillating) Very low humidity now, may be damaged Channels 16 -23 looked different than the other 24. I looked at them with periodic reset off, and a reasonable global threshold of 50. Plots and more details on next Peter's traces are MUCH quieter than mine on the analog monitor • • I have lots of clock and 60 Hz. Shielding helped 60 Hz a lot (just closed my box) If I turn on BW limit on my scope at 20 MHz they are similar Peter usually uses BW limit, so I will too, make it nicer I tried a shorter cable, cutting out unused grounds Clock noise still exists even without a cable/board, so much of it is on the DAQ board itself Leon Mualem Jan 9 2020 LAr. Pix Mtg 3

LAr. Pix. V 1 Socket Testing Chan 15 • Channels 16 -23 looked different - I looked at them with periodic reset off, and a reasonable global threshold of 50. - Most showed consistent leakage/offsets (first plot) - Channels 16 -23 had similar leakage current, (slope) but they started off 200 m. V lower after reset, not about 300 m. V like the others. They have more wiggle as well. - Lots of theories: • Connections • Biasing • Pickup • Tried some shielding Leon Mualem Jan 9 2020 LAr. Pix Mtg Chan 16 4

V 1 Socket board shielding • Examined layout from the. brd files in cadence allegro viewer to look for candidates - Suspected some traces from the new charge injection connector, but grounding them did not change anything - Suspected MISO_A line, from broadcast tower (loopback jumper) - Creating a foil party hat for that, and the socket helped several of the channels (20 -23), but the problem remains on 16 -19 - Cover for the socket eliminated 60 Hz noise and sensitivity to hand waving scientist even with the box open Leon Mualem Jan 9 2020 LAr. Pix Mtg 5

V 1 Socket board shielding • Examined layout from the. brd files in cadence allegro viewer to look for candidates - Suspected some traces from the new charge injection connector, but grounding them did not change anything - Suspected MISO_A line, from broadcast tower (loopback jumper) - Creating a foil party hat for that, and the socket helped several of the channels (20 -23), but the problem remains on 16 -19 - Cover for the socket eliminated 60 Hz noise and sensitivity to hand waving scientist even with the box open Leon Mualem Jan 9 2020 LAr. Pix Mtg 6

V 1 Socket board shielding • Examined layout from the. brd files in cadence allegro viewer to look for candidates - Suspected some traces from the new charge injection connector, but grounding them did not change anything - Suspected MISO_A line, from broadcast tower (loopback jumper) - Creating a foil party hat for that, and the socket helped several of the channels (20 -23), but the problem remains on 16 -19 - Cover for the socket eliminated 60 Hz noise and sensitivity to hand waving scientist even with the box open Leon Mualem Jan 9 2020 LAr. Pix Mtg 7

Ch 16 droops at readout (normal/zoomed) • It appears that the noise is related to readout (MISO_A signals) • After reset from readout, channel droops • Zoomed image shows the readout burst and monitor line drooping Leon Mualem Jan 9 2020 LAr. Pix Mtg 8

Ch 15 no readout droop • It appears that the noise is related to readout (MISO_A signals) • Normal channel has readout at each peak, but no droop • Zoomed signal shows data bits and noise, but no droop Leon Mualem Jan 9 2020 LAr. Pix Mtg 9

Ch 16 -19 droop source • It appears that the noise is related to readout (MISO_A signals) • Running with long periodic reset and high threshold (no data) shows normal signals • Doing a verify_config to drive data makes ch 16 droop to near ground for that period • Other channel shows similar noise without dropping baseline Leon Mualem Jan 9 2020 LAr. Pix Mtg 10

Ch 15 and 16 interference • Evidence of readout on ch 15 drooping ch 16 • It looks like the periodic resets would fix droop problem, so maybe with default reset will be better (4096 instead of 1 M, 250 x more frequent) Leon Mualem Jan 9 2020 LAr. Pix Mtg 11

Pathological channels • Took a quick look at “oscillating” channels • Appears to be very high leakage current on these • Operation appears to be normal, but slope is very high, 50 MICROseconds instead of 100 MILLIseconds Leon Mualem Jan 9 2020 LAr. Pix Mtg 12

Conclusions on progress • V 1 socket board is working well for ~90% of channels • Not covered here, but some preliminary charge injection tests look ok also • Signal lines are VERY sensitive to signals - Less concern for v 2 with differential signals, but still • Chip. ID pins need different config (lower resistor) - Less concerned for v 2 which does not use external config • Droopy channels will be influenced by other channels, but periodic reset will reduce the effect. • High leakage channels may be created in the testing process • Slow testing for now, looking at signals on analog monitor, once measurements are identified, will automate processing Leon Mualem Jan 9 2020 LAr. Pix Mtg 13