Re: Bcal read out options.

["Fernando J. Barbosa" <barbosa@jlab.org>]

Hi Elton and George,

My comments follow:

Elton Smith wrote:
> HI George,
>
> Thanks for your comments. Elton.
>
>
>
>
> Elton Smith
> Jefferson Lab MS 12H5
> 12000 Jefferson Ave
> Suite # 16
> Newport News, VA 23606
> elton@jlab.org
> (757) 269-7625
> (757) 269-6331 fax
>
> On Fri, 17 Jul 2009, Georgios Voulgaris wrote:
>
>   
>> Hi Elton ,
>>
>> ��� I had a first reading of the document 1300 and I would like to comment about
>> timing , paragraph 4.3 .
>> ��� �� When we have signals with high dynamic rage, timing resolution is limited
>> because of "slewing" i.e. a big pulse reaches threshold earlier then a small pulse,
>> assuming that both of the pulses have the same rise time. There are several remedies
>> to this:
>> ��� a) Offline correction .
>> ��� ��� ��� i) We can have a parametrization of time vs. pulse high and use tis to
>> correct the timing.
>>     
Yes.
>> ��� �� �� � ii) If we use Flash ADC we can fit the pulse data with a curve and then
>> use the 50% point for timing.
>>     
The timing resolution is not good enough.
>> ��� For both cases the correction deteriorates as P.H. decreases. Unfortunately I
>> can't give an estimate for this correction because its been a long time since I have
>> used them.
>>     
>> ��� a) Hardware:
>> ��� �� �� �� �� i) Modern leading edge discriminators can go down to 3 or 5 mV
>> threshold so there is no need for extra amplification. ( Caen N844 ). One has to
>> make sure that noise level from all sources is vary small.
>>     
Yes but noise is the limitation.
>> ��� ��� ��� ��� ii) Constant fraction discriminator, ( internal pulse forming and
>> discrimination). These devices have rate limitation, they are used� more often in�
>> Nuclear� physics,� but� it may be worth the trouble� to test� them. (Caen N842 or
>> similar).
>>     
Yes, CFDs have rate limitations and are not good at the low end. Paul 
Eugenio is testing them with the TOF prototype. Most likely all 
discriminators will be LE in GlueX.
>> ��� �� �� �� �� iii) Amplification� plus pulse shaping so that a short pulse with 1
>> ns rise time is created. One has to look for the electronics or develop them. Mostly
>> "home brew" electronics.
>>     
This is the plan with the amplification at the detector end because of 
the wide dynamic range and the low amplitude signal levels, especially 
for the case of the FMPMTs. The shaping amplifier will be of the 
differentiator/tail cancellation type.
>> ��� ��� Of course there are other reasons that affect the timing:
>> ��� �� Size of the shower, hadrons or delta rays that can confuse the analysis. I
>> believe that a FADC is needed to resolve� this� issue.
>> ��� �� �� I suppose that several of these issues have been investigated already.
>>     
We have investigated some and continue to do so but there have been no 
definite tests with the SiPMs. We need to plan a set of tests this Fall 
with the SiPM prototypes from both SensL and Hamamatsu which will 
include the new custom electronics.

Best regards,
Fernando


>> ��� �� �� I will try to connect to the meeting, but usually we end up with a "broken
>> telephone", so I'd rather use e-mail.
>>
>> Best Regards,
>> George Voulgaris.
>>
>> ��� ��
>>
>> ���
>>
>>
>>     
>
>
>   
begin:vcard
fn:Fernando J. Barbosa
n:Barbosa;Fernando J.
org:Jefferson Lab
adr:Suite #10, 12B3;;12000 Jefferson Ave.;Newport News;VA;23606;USA
tel;work:757-269-7433
version:2.1
end:vcard