TDC board discussion

[Elton Smith <elton@jlab.org>]

Hall D Electronics:


Attending: Chris, Ed, Dave A., Fernando, Elton, Elke, Simon, Dan (did I
forget anyone?)
Notes by Elton and Fernando.

On Monday, Aug 6, we had a discussion about the number of new TDC boards
that would be required and what format they would take. Below is a
consensus proposal that emerged from the meeting.

1. There will be a single TDC board design.
2. The TDC board will have LVDS inputs
3. The board will be able to support 8 F1 TDC chips.
4. This design will apply to all new users at Jlab. The old design will
still be supported and can be fabricated at any time, if needed.
5. The board holds 8 F1 chips = 64 ch low resolution (60 ps)/ = 32 ch high
resolution (120 ps) [55 ps/115 ps, actual].
6. External RAM added to board to assist with buffering required for high
rate capability.

Note: VXS may also be implemented if cost/convenience can be justified.
Presently the signal distribution (clock, sync, trigger) is implemented
with a board and cables on the back of the backplane and this is ok.

For use with GlueX, the board will be configured in two ways:
1. PMT TDC
a. TDC in high resolution option of 60 ps
b. 32- channels per board
c. Input connectors will the common 34-pin 0.050" pitch 17-pair ribbon
cables
d. This TDC will be feed by a splitter/disc board which will be needed
for PMT systems, in particular the Bcal. This splitter/disc will
have analog inputs --LVDS Discriminator Output to TDCs + Analog Output to
fADC-250. Note that the analog output must be conditioned to meet the
dynamic range of the fADC.

2. DC TDC
a. TDC in low resolution option of 120 ps
b. 48- channels per board
c. Input connectors wil be the fine pitch 0.025" 24-pair ribbon cables.
d. Additional line output which can be used to pulse the preamp boards
(24/48 channels at a time?). The requirement for this pulser need to be
detailed. Although the details have not been worked out yet, the pulser
can be easily generated by the FPGA. A driver will provide the required current
and levels.
e. The board will be feed directly from the DC preamp boards.

Note: The PMT TDC (F1TDC-PMT?) and the DC TDC (F1TDC-DC?) are identical
except for the front panel and input connectors. We may decide not to
install 2 F1 chips on those boards destined to be used as F1TDC-DCs.

Comment from Fernando: I believe this to be an excellent decision because
it minimizes the impact on the physics due to multiple scattering, uses standard
commercially available and robust connectors, matches the detector
design constraints and minimizes the use of engineering resources.

Comments and suggestions are welcome.

Cheers, Elton.