[Halld-cal] Draft f250 data words for comments

[Elton Smith]

Dear colleagues,

We have been discussing various considerations for the next f250 
firmware upgrade. Based on the size of the spring data with mode 7, we 
see that the data volume is dominated by header words and not the 
"content" per se. Therefore, it appears that we should be able to handle 
the GlueX rate with 3 words for the calorimeter "content". However, it 
would be desirable to separate the pedestal into one word, which could 
be dropped if size becomes an issue. Attached is a draft scheme for 
discussion that achieves these goals.

Feedback is welcome.

Cheers, Elton.

-- 
Elton Smith
Jefferson Lab MS 12H3
12000 Jefferson Ave STE 4
Newport News, VA 23606
(757) 269-7625
(757) 269-6331 fax

-------------- next part --------------

Current Output:

Mode 7: Outputs Data types 7, 8 and 10

Data Type  7: Pulse Integral
(31) = 1
(30-27) = 7
(26-23) = channel number (0-15)
(22-21) = pulse number (0-3)
(20-19) = quality factor (0-3)
(18-0) = pulse integral

Data Type  8: Pulse Time
(31) = 1
(30-27) = 8
(26-23) = channel number (0-15)
(22-21) = pulse number (0-3)
(20-19) = quality factor (0-3)
(18-15) = reserved (read as 0)
(14-6) = coarse time (9 bits)
(5-0) = fine time (6 bits)

Data Type 10: Pedestal and Peak
(31) = 1
(30-27) = 10
(26-23) = channel number (0-15)
(22-21) = pulse number (0-3)
(20-12) = pedestal, average of first 4 samples of window (9 bits)
(11-0) = pulse peak value (12 bits)


New Proposed:

Mode X: Output data type X

Data Type  X: Pulse  Integral, Peak, Time, Pedestal
(31) = 1
(30-27) = X
(26-23) = channel number (0-15)
(22-21) = pulse number (0-3)
(20-19) = quality factor (0-3)
(18-0) = pulse integral

(31) = 0
(30-22) = coarse time (9 bits)
(21-16) = fine time (6 bits)
(15-13) = quality factor (3 bits)
(12-0) = pulse peak, includes overflow (13 bits)

(31) = 0
(30-15) = pedestal sum (16 bits)
(14-0) = unused