Sphenix-tpc-l mailing list

[Takao Sakaguchi <takao AT bnl.gov>]

Hi Martin and all,

 Just during the presentation of ALICE's TPC upgrade at QM,

I estimated the rate for our case. If we readout out all channels

without zero-suppression at 10MHz, it will result in

25.6GBits/sec/board, 640Gbits/sec/sector, and 15.4TBits/sec

from the whole TPC.

 I thought the idea was to let the FPGA on FEE takes care of the

undershoot/cross-talk if necessary. Definitely, we should discuss

about it soon.

Takao

On Wed, Feb 8, 2017 at 9:28 AM, Martin Purschke <purschke AT bnl.gov> wrote:

All,

Johanna visited my QM poster, which had a section about the streaming
readout. I had this cartoon where I used our existing minidrift TPC data
(I showed this in a previous meeting) where the above-threshold samples
are preserved (you can look at my poster here -
https://indico.bnl.gov/getFile.py/access?contribId=11&resId=1&materialId=slides&confId=2794).

You need this because the SAMPA chip doesn't have enough connectivity to
the world to offload the entire waveform, you need to trim down.

Johanna (whose student is working on the code for the CRU) was adamant
that they tried that and failed - they need a larger than the 32-channel
peep-hole view to do common noise subtraction etc properly. They take
all the data up to the CRU and do a more sophisticated dynamic threshold
analysis on the fly. She didn't answer my question how they would deal
with the link over-subscription, but later I learned from other ALICErs
that they apparently want to dial down the sampling frequency to make up
the shortfall. (I wasn't aware that this is possible).

Off the top of my head, I don't think this is an option for us. But if
true, and this is indeed needed, it would have far-reaching consequences
for our design. I will try to get some more info, but the people who
would know more technical details tend not to hang out at the QM.

After one night's sleep over it, I think we should see how low we can
take the threshold to drive the links as hard as we can, and see if this
is adequate to run at 10MHz.

Just FYI ..

        Martin

--
Martin L. Purschke, Ph.D.        ;   purschke AT bnl.gov
                                 ;   http://www.phenix.bnl.gov/~purschke
                                 ;
Brookhaven National Laboratory   ;   phone: +1-631-344-5244
Physics Department Bldg 510 C    ;   fax:   +1-631-344-3253
Upton, NY 11973-5000             ;   skype: mpurschke
-----------------------------------------------------------------------
_______________________________________________
Sphenix-tpc-l mailing list
Sphenix-tpc-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-tpc-l

--

Takao Sakaguchi
Brookhaven National Laboratory
Physics Department, Bldg. 510C
takao AT bnl.gov
Office: 1-631-344-3345