Notes about problems with MCM 101

MCM-101 was originally tested 8-Jul-99. It failed because it had "32 bad channels (33-64)." On 6-Aug-99 it was tested again (see description of problems below) with the same result. Around then, Jehanne discovered a short in a surface mount capacitor associated with the 2nd TGV chip. This was fixed and on 19 and 20-Aug-99, it was tested again. It now works.

Additional comments from 19/20-Aug-99 tests
Unlike the previous tests (see below), the MUX's work for all 8 AMU/ADC chip. This MCM seems a little noiser than usual, but I did not try to understand it.

New conclusion: accept this MCM.

Comments from (now out of date) tests:
On 6-Aug-99, after getting Xilinx programs which we think work correctly, this MCM was retested. The results were the same -- the 2nd block of 32 channels does not work. The ADC value was not overflow, but it did not respond to Vcal. It also stayed constant when I turned off the internal preamp reset and issued level-1's. The other channels give different ADC results in that case.

The preamp spy lines for the 2nd chip showed no response to bench-cal, it always stayed around 1.5 V. The AMU spy line also showed no response to Vcal. All other AMU spy lines worked. I did not test all other preamp spy lines, but I checked the 1st and 3rd chips and they were OK. I seems that the input voltage to the ADC is always the same. Therefore, the ADC value is always the same. It seemed clear the the DAC's used to adjust the ADC ramp (Iref and Vref) worked correctly. I was able to change ADC values in the expected directions by changing Iref and Vref, but I was never able to get a response to Vcal input.

Conclusion: 7/8 of this MCM works. I think that either the 2nd TGV chip does not work, or else the connections (all 32) between it and the AMU/ADC are bad. MCM serial number 109 had a problem with the same symptoms.

updated 20-Aug-1999
John Sullivan (sullivan@lanl.gov)