[Halld-offline] FADC125 upsampling algorithm implemented in emulation mode
davidl at jlab.org
Mon Feb 9 07:53:25 EST 2015
One other thing I forgot to mention. This latest change also fixes an issue with the emulated pulse
pedestal for F250 hits being a sum rather than an average. The reported value is now an average.
This will likely only affect those looking at mode 1 data (of which there is very little) since it has
flash samples but no pulse time/pedestal in the output data file. Thanks to Nathan S. for identifying
this problem a week ago.
On Feb 9, 2015, at 6:35 AM, David Lawrence <davidl at jlab.org> wrote:
> Dear Offliners,
> I have just checked in a change to the DAQ library that implements emulation of the CMU
> timing algoritm for the f125 ADC in emulation mode. This will not affect you unless you specify
> that the f125 pulse times found in the data stream should be ignored, thereby triggering
> emulation of the pulse time hits using window raw data. Here are a couple of notes:
> To ignore firmware generated pulse times and thereby turn on emulation, add the following:
> By default you will now get the CMU up sampling algorithm. To emulate the older f250
> algorithm add the following to the command line:
> n.b. this f250 emulation mode does not seem to reproduce what the firmware reports very well.
> It is, however, the same algorithm used for emulation of the f125 prior to this most recent
> commit so can be used for comparison.
> IMPORTANT: The units reported by the f125 upsampling algorithm are 1/10 of a sample
> time while those of the f250 algorithm are 1/64 of a sample time. The code just committed
> scales the upsampling algorithm results to be in units of 1/64 of a sample in order to
> match the existing downstream code. When this new algorithm is implemented in firmware
> we will need to handle the unit change in some different way (e.g. calibration constants).
> Regardless, the user will need to be keenly aware of what they are doing when comparing
> emulated f125 pulse times to firmware generated pulse times given that there will be
> 2 different firmware algorithms and 2 different emulation algorithms and likely 2 different
> eras of the DAQ library modifications for implementing the current best case. Sorry if this
> is confusing, but that is the situation. I’d rather people be a bit confused than blissfully
> ignorant of the potential for complication here.
> Halld-offline mailing list
> Halld-offline at jlab.org
More information about the Halld-offline