7.25.2014 Meeting

Webex meeting for those not at SLAC: https://confluence.slac.stanford.edu/display/hpsg/Webex+connection+details

Agenda

• FEB testing
• AOB

Minutes

One main discussion topic is that we probably spent time changing to many parameters as there are several results we don't seem to be able to recreate or they are inconsistent. We have for example spent time trying to understand why the external power source run noise performance was not reproducible anymore in Ben's room but it turns out that this was with the other FEB and in group C. The results that indicated that the inductor chokes was a problem is not reproducible either with Ben's setup anymore. 

Final outcome:

We agreed that if we don't learn anything before WEDNESDAY we should pull the trigger and order FEBs in order to continue the assembly, we have yet to see the final performance which may be affected by new issues. This still is based on the assumption that the current noise performance would provide us with a working detector (assuming no new things arise) so nailing down the target level of noise is a key in this decision (see below). New boards might have to come later (with subject to funding). 

Tentative plan:
Mon - get quote and have Lupe place the PO.
Wed evening - pull trigger assuming we don't learn anything drastic.
=>This means we can realistically have the new boards loaded and ready to go when we get back from JLab.

Ryan is out but would keep in touch with Ben to see how tests are going on Mon, Tue & Wed. 

Other notes:

It is slightly unclear as to exactly what our target noise is and we need to nail that down in order to be able to take the decision on the next step with the FEB production next week. Omar will write up the numbers for noise, gains and signal from the test run, bonded and unbounded hybrids. He will coordinate with Ben and Sho. We should circulate that to make sure that we agree this is consistent. It was unclear as to how much more noise this board has compared to the test run for example. 

Ben and Ryan spent this week probing across the board and there is no clear sign of the increased noise. There is a spike from the pre-amp chain into the ADC caused by current-sourcing being inadequate and there seems to be a 1MHz ring to that on the FEB. Note that this was verified to have been there on our previous designs (devboard, test run, what not). The devboard ringing seems to be at 600kHz, not clear why. Ben is experimenting with a cap to be able to source that current and it seems to be working. The noise pattern we originally set out to defeat is still there though but this can help. 

Since the noise seems to be common mode we might need to investigate what ways we have to remove it. Sho has measured it but not tried to remove it yet. Ryan mentioned we could have a filter running to lower this if it sticks. There was also a suggestion to look into an empty ADC channel in order to see if the noise is there. It might tell us where the noise is coming from and possibly it can be used to subtract from other channels in some way.

Ben was gonna check that the op-amp and ADC are properly setup according to specs. 

Ben was gonna move the setup into group C and see how it performs. We were working under the assumption that external power is great but it turns out that was the less tested FEB. So a key thing is to run that board through more tests and see how it performs. 

Tim mentioned (through email) that, as we talked above, we should make sure we understand the size of the noise and if it's common mode, and if it's might be possible to try and shield the path from the hybrid to the FEB to avoid picking up things from the environment.