Thanks Joe,
I'd focus on proving the ADC IC to FPGA to readout
link is completely working before thinking at all
about the analog signals.
Perhaps adc16_init is failing ?
what does it say in verbose mode ?
How big of valid sampling windows are found during the calibrate step ?
Have you put the ADC IC's into their builtin pattern generating modes
and seen those patterns ?
see for example
0x25 : "en_ramp", "dual_custom_pat", "single_custom_pat"
0x26, 027 for the pattern value(s).
To set these modes I typically go into ruby and interactively run
setreg commands.
Have the ADC ICs been left in a pattern generating mode
so they are ignoring the analog inputs ?
Have you tried the known working sample BOF file ?
Does it have any trouble seeing any of the analog inputs ?
I doubt this is what we are fighting but I'd bet there have
been software upgrades that might make things work snazzier.
Matt
On Tue, 25 Mar 2014, Kujawski, Joseph wrote:
Date: Tue, 25 Mar 2014 11:55:58 -0400
From: "Kujawski, Joseph" <[email protected]>
To: Matt Dexter <[email protected]>
Cc: casper <[email protected]>
Subject: Re: [casper] ADC16-8 conversion problem
Matt,
1) As far as I can tell, I comply with everything in your documentation.
2) I have injected signals from 20mV p-p to 2V p-p into the ADC-16 inputs.
More than one channel
has been tried and the results are always the same, I get oscillation of one
bit in the reported
output at a rate that corresponds to the injected signal.
3) When injecting into channel D4, I have confirmed that pins 37 and 38 of U8
have a differential
signal of amplitude 1Vp-p on each leg.
4) I am using the Ruby script (adc16_init.rb.bat) that you gave me last year to
initialize the
ROACH-2 board.
What should I look at next?
-Joe Kujawski
