Update TOA and TOT scans#434
Conversation
I am pretty sure I have left their files untouched. I did delete some files related to the old tot scan, but I don't think I deleted any files related to the toa scan. Which files are you referring to? |
|
Oh yea, you're right, I misread My question remains though - it looks like you have implemented |
|
Oh yeah, the purpose of the function is the same. The plan was to get a working version. Once Pierce and Will finish their functions and we see how effective they are, we can pick one going forward. |
|
Jack, Pierce and Will summarized progress here and shared it with CMS folks: https://docs.google.com/presentation/d/1PuzthVSWL0KrQL0XiZCgzaWRW_tTlfmDLGLtzZWkPWk/edit?usp=sharing I'll note that Jack's procedure for the TOA vref and trim tuning is not the same as the one from CMS, whereas Will's + Pierce's follows the CMS procedure. However, Jack's procedure seems to work fine. I suggest merging his algorithms for now.
|
tomeichlersmith
left a comment
tomeichlersmith
left a comment
There was a problem hiding this comment.
A second, more thoughtful read of the code (still haven't tried to run it!) gives me more confidence. It looks like a thoughtful algorithm, well done!
You can also add doxygen-style comments within the functions themselves, e.g.
/**
* First, we incrementally lower toa_vref from its maximum of 255 until XXX
*/This would then end up in the online doxygen manual: https://ldmx-software.github.io/pflib/
(although many of the pages there have gone stale and I need to update them)
|
@knh3dg are you waiting for anything from me? I've been waiting for more documentation on your updated algorithms. I'm willing to merge this and have both algorithms available in one version of pflib so they are easier to compare. |
|
I think Jack edited the PR description: If you need more, maybe you can make some explicit requests? |
tomeichlersmith
left a comment
tomeichlersmith
left a comment
There was a problem hiding this comment.
Yea, that's probably fine, but as I looked through I noticed one more bug. The toa_csv_scan as written does not "unshuffle" the links appropriately. It needs to use the "mapping" stored within the target to map a roc/channel into a i_erx/i_ch.
3 participants
After setting up the ECON/ROC/DAQ, ROC calibration should go as follows:
TASKS.LOCAL_PEDESTAL_LEVEL
TASKS.GLOBAL_PEDESTAL_LEVEL
TASKS.TOA_SCAN
TASKS.TOT_SCAN
If TOA_SCAN or TOT_SCAN is having trouble, e.g. many trim values are 63, try setting the target calib to a higher value. Note that TOA_SCAN aligns in the low range while TOT_SCAN aligns in the high range. Using more events, e.g. 100 instead of 10, should also improve consistency, but from my testing the difference is often quite small.