[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: TOF constant fraction



Hall D PID Mail List:

Hi Sasha,
indeed with the low/high coincidence you will lose all signals that do not
tiggger the high threshold. The timing is determined by the low threshold
but at the same time you require a certain minimal signal amplitude defined
by the high threshold that is quite larger than the low threshold. This will
minimize the walk because the time when the signal goes above the low 
threshold
for signals of sufficient amplitude is basically always the same. If you 
think
the signals are too small for the discriminator because usually the 
lowest threshold
setting for most discriminators is around 30mV you might use an 
amplifier. However
an amplifier will always also amplify the noise so one has to be careful.
If you use a linear fan in/out to multiplex the signal this will not 
introduce much jitter,
however care has to be taken about the DC offset. Usually the DC offset 
can be adjusted
with a front panel potentiometer and should be set such that the base 
line of the output
signal is at zero. Note also that this DC offset can be rate dependent 
but usually only if
the rates are very high like hundreds of kHz or MHz. However these 
linear fan in/out units
usually have the bad side effect that they introduce noise, in 
particular 60Hz line noise.
So keep an eye on that.
hope that helps,
cheers,
Beni


> Hall D PID Mail List:
>
>   Benni,
>
> Thanks for your comments. Paul is already looking
> if we can get hold of other CFD types for the test. 
>
> Regarding V812: the delay is indeed set at 20ns. Per 
> your suggestion, we'll do a measurement at 4ns to see
> if this makes any difference. 
>
> Doing ADC "time-walk corrections" to CFD is not
> trivial for us at the moment because the simple
> trigger logic we use does not allow us to match
> a TDC event in a multi-hit F1TDC with an ADC event
> in a single-hit CAEN ADC we have at the moment.
> Our plan is to wait for the flash ADC to arrive, and
> then build a more complex "synchronized" trigger logic
> for this specific TDC/ADC pair.
>
> I'm still trying to digest your suggestion to split a signal
> into low- and high-threshold discriminator channels and
> then combine them again. Does this eliminate all the
> small signals between low and high thresholds which
> we want to keep? Does an addition of a fan-in/fan-out 
> unit introduce its own time-walk?
>
> Sasha
>