search

msr-consulting / exscalabar_server

Repository for the EXSCALABAR server.
http://www.msrconsults.com/ukmet-gh/exscalabar
0 stars 1 forks source link

Speaker enable signals #118

Closed JustinLangridge closed 8 years ago

JustinLangridge commented 8 years ago

These still aren't working. Referring back to email a email I sent on 6 May 2016, the problem appears to be the following:

Speaker enable is coming out of the 7842R card on DIO 10 of connector 0. Hardware needs is to come out of DIO10 on connector 1.

Similarly iSpk is coming out of DIO 0-4 on connector 0 but needs to be coming out of DIO 0-4 on connector 1.

Could this be changed o the FPGA? Ideally asap as I would like to start testing the speaker/mic functionality

lo-co commented 8 years ago

This was changed the last time you all requested this (the build was dated May 16th. Isn't this what you were looking for? Have you double-checked to see what is coming out of the cable itself? I am rebuilding (again).

fpga_io

lo-co commented 8 years ago

Set the build as default...

JustinLangridge commented 8 years ago

Had to interrupt the FPGA build earlier today - still hadn't finished 6 hours in. Will redo hardware tests to check we don;t have cabling issue.

JustinLangridge commented 8 years ago

ok am getting out a signal on DIO 10 for the speaker enable that cycles with the speaker button on the UI. It is 3.2V though not 5V - could be part of the issue?

iSpk outputs are fixed at 3.2V (and don't cycle with the UI speaker button)

lo-co commented 8 years ago

OK - they iSpk shouldn't cycle with the speaker button.

3.2 V is TTL. I wouldn't think that is the problem, but input needs 5 V we will have to swap to an AO. This sounds like a question for Rob.

JustinLangridge commented 8 years ago

Ok we have a routing issue through the main electroics box. Another thing for us when we pull the box out (likely Monday morning)

JustinLangridge commented 8 years ago

As a test we've changed the code such that the UI speaker button now changes both the ispk and speaker enable signals. This makes things compatible with the hardware which is configured to use the ispk signals to turn the speakers on and off.

There may be other places in the code that this needs to be changed so that the speaker cycling operates correctly?

Note that despite these changes the speakers are not yet operating properly. This now looks likely to be a hardware issue

lo-co commented 8 years ago

Can you give me an update on this?

JustinLangridge commented 8 years ago

With the above change such that we now link the ispk outputs to the UI speaker button we can get the speakers on and functioning.

We played with setting the speaker voltage settings to get a reasonable level that does not saturate the mics. This was ok but a bit of an exercise in random guessing - what are these two controls actually setting in terms of the waveform and what are their bounds?

We have looked at the resonance curves on all cells (but not simultaneously as we are still fixing three of our cell top electronics boxes). These look ok apart from PAS 4 where the time domain mic signal with speaker operating looks to have a big offset which none of the other channels do. This seems to be a feature of the PAS4 and we don't know what is causing it yet.Is the digitization on this channel definitely set up the same as the rest (i.e. differential)?

It would also be useful if you could summarize briefly the processing and scaling that is done on the mic signals before they get to the UI plots (e.g. what are reasonable signal levels on the arbitrary units scale that we should be looking for). If you have plots of expected time and freq domain signals that would be useful

JustinLangridge commented 8 years ago

Speaker enables are now working. Time and frequency domain speaker signals are not great though - will open separate issues for these. Closing