search

Teststandees / hcal_teststand_scripts

HCAL teststand scripts
0 stars 11 forks source link

QIE mapping #2

Open awhitbeck opened 9 years ago

awhitbeck commented 9 years ago

QIE mapping is too slow... need to do some research on how to speed it up. Some ideas:

Use the uhtr histograms to read out data - which is very fast.

Use the link pattern readout to first figure out where groups of four QIEs then set those groups of four to either fixed range mode 0,1,2,3.

-the combination of the two should be very fast

@Teststandees/citizens

awhitbeck commented 9 years ago

did we ever debug the faster mapping code?

elliot-hughes commented 9 years ago

No, but now's a good time to do this, since the map changed. I'll do it.

elliot-hughes commented 9 years ago

This is more annoying than I anticipated. When I use the slow version, I get results like

{"slot": 2, "uhtr_slot": 11, "fiber": 1, "link": 16, "half": 0, "qie": 17, "crate": 1, "id": "0x00000BAD 0x00000000", "channel": 3},

The "id" field is odd. I'm going to go to lunch and look into this after.

awhitbeck commented 9 years ago

How are you reading the unique ID? From the device itself or from the IGLOO2 register?

elliot-hughes commented 9 years ago

The "0x00000BAD 0x00000000" unique ID was due to a bug introduced by removing quotes from the returned results of ngccm.send_commands_parsed. I fixed the bug. I updated the map in the repository.

awhitbeck commented 9 years ago

nice!

awhitbeck commented 9 years ago

can this be closed?