[Halld-tracking-hw] ADC125 as a TDC...

[Gerard Visser]

Hi Beni,
	I'm not sure what to make of that tail either. One question that might be 
related: are you putting a point on the histogram for every strip? What would it 
look like if you just used the earliest strip hit? Certainly the side strips 
which see smaller signal will not cross the timing threshold as soon as the 
central strip(s) do. That effect must produce a tail to the left in this plot, 
but whether it would be such an extreme tail, I don't have a clue.
	Have you had an opportunity to begin to optimize the LE threshold? (I'm still 
guessing this would mainly be done with a goal of overall drift coordinate 
resolution, so maybe that has to wait a bit. But on the other hand also would 
worry that all results are threshold dependent and is it set roughly where we 
will want it?)
	Sincerely,

		Gerard

On 5/17/2011 8:20 AM, Beni Zihlmann wrote:
> Hi All,
> attached are 3 plots. The first is a 2d plot for FDC wire 11 on cosmic
> data. The vertical axis is the strip number and the horizontal axis
> is the time difference between the strip and wire 11. The times are
> determined using Gerard's algorithm with interpolation on the FDC
> data for both the wire and the strip. Note that the time difference
> is calculated as "t_wire - t_strip" so the long tail to the negative times
> is due to the strip while we have a rather sharp edge on the right side.
> this can be seen more clearly when looking at the projection of the 2d
> plot to the horizontal axis. This is shown in the next to plots individually
> for the channel 0-23 and 24-47 which represents the two preamp cards
> for the top and bottom cathode planes. I try to extract a width by fitting
> mostly only the upper edge. As you can see the width is between 1 and
> 1.5 ns. So the individual resolution may be about sqrt(2) better. You can
> also see that the timing resolution depends on the average signal amplitude
> which is smaller for the second preamp (channel 24-47). I do not know what
> causes the shift in time between the two preamp card. This shift is certainly
> much larger than any cable length miss match etc. But Lubomir observed
> this behavior already earlier.
> What causes the long tail to negative time difference is not completely clear
> to me. One reason might be that there are considerable gain variations between
> strip channels and I plot all hits not just the one with the largest amplitude
> in a hit.
> Let me know if you have additional ideas what else one could tests.
>
> cheers,
> Beni