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

Re: Tracking fADC



Hi Gerard,

I've copied Simon in case he is not on the tracking-hw mailer to answer 
the question about the number of Struck modules.  The plan is to 
assemble the boards by hand, so we will plan for 5 boards.  Keep in mind 
the lab is shutdown from 25 Dec to 3 Jan and I'm not sure how long it 
will take to assemble the boards.  I think it is no problem to work 
around your travel plans to BNL,  and the assembly will begin as soon as 
we receive the boards/parts/documentation.

I agree [ and earlier review summaries agree ] that we need only two 
types of fADCs.  We updated the Gluex detector table in April '06 and it 
reflects the two fADC plan.  Recent discussions/decisions on the number 
of channels per module clearly impact the number of modules/crates for a 
given detector system, so as the detector groups decide we can update 
the table.  I noticed that the table does not list TDCs for the FDC 
anode wires, but I believe this is something that needs to be 
discussed/changed. 

There are new quad channel 100Msps ADC chips out, and you are right, 
these chips will have to be evaluated for power consumption, readout 
modes[serial LVDS, parallel, etc], chip size, and of  course cost.

I am reviewing the cost and schedule for the PED and Construction 
spreadsheets that Elton posted to the portal for an upcoming review.  I 
admit, I do not know when the review is going to happen, but I will post 
my findings to Elton before the holiday break.

Cheers,
Chris
(757)269-5053
~~~~~~~~~~~~~~



Gerard Visser wrote:

> Hi Chris,
>     After further delays mostly due to other projects, I can now 
> finally say the boards went to fab on 12/4. I expect to receive them 
> on 12/15 (elected 7 working day turnaround).
>     I will quickly evaluate one board here and send on to you several 
> bare boards and parts. [And documentation, of course. I am finalizing 
> that now.] Can you remind me, how many total Struck ADC's do we have? 
> I seem to remember 5 200 MHz and 1 100 MHz, but maybe I remember 
> wrong... Please check/clarify. If we have 6 Struck boards that is 6*8 
> channels so I would propose to assemble 4 or 5 of the 16 channel 
> amp/shaper boards. No point to committing parts to more boards than we 
> have ADC's to read, although I am procuring parts for 8 amp/shaper 
> boards per our MOU. Your tech will be assembling by hand, right? 
> [There are no leadless or bottom-soldered parts on this board.]
>     I can come to JLab in January to work with Simon to set up the 
> amp/shaper boards and make measurements. I may have to travel to BNL 
> for STAR setup activities, schedule depends on RHIC run schedule which 
> is in flux still. We'll have to work around this, sorry.
>     Pending a decision to use 24-channel (as opposed to 16-channel) 
> granularity on the cables and the preamp boards, the tracking ADC 
> module will be 72 channels rather than 64. This would be still on a 6U 
> VME module (36 channels on main board and 36 more on a mezzanine). If 
> I understand right, Dan & Simon and Fernando all basically agree to 
> the 24-channel granularity.
>     I don't think it makes much sense to have two different slow ADC's 
> (65 & 125). We should settle on one sample rate. _If_ it turns out 
> then that we use a part which is available also in a pin and 
> functional compatible lower speed version, then we could consider 
> saving some money this way. I think the costing should assume not 
> doing this, at this point.
>     The sample rate of the prototype ADC is not yet finalized. Based 
> on request from Curtis it should be >100 MHz; as you note, that's what 
> the detector electronics summary table shows too. I'll need to 
> investigate the current market offerings and pick a sample rate not 
> less than 100 MHz, based on the best that can be done today for 
> maximum board density and minimum power. I suspect also this exercise 
> will show that significant cost (and power ...--> cost again) savings 
> are possible at some lower sample rates like 80 MHz. That was the case 
> when I looked at this last, but it has been a while. I need time to 
> look at the latest parts.
>     Is the ADC board cost information needed by February per our Nov 
> 16th phone conference, or much sooner?
>     Talk to you later,
>
>     Gerard
>
> C. Cuevas wrote:
>
>> Hi Gerard,
>>
>> A quick message to ask you about the amplifier/shaper boards and to 
>> also ask you about the flash ADC design.
>>
>> We can begin assembly as soon as you send the amp/shaper board.  
>> Please include any detailed drawings needed for SMD reference 
>> designators etc.  Presumably you will also send a bill of materials 
>> with the 'kit'.
>>
>> I am tasked with reviewing the cost/schedule for the Hall D project 
>> in preparations for a review.  The GlueX detector 'table' shows a 12 
>> bit 100Mhz design for the FDC and CDC.  I know you are designing the 
>> 64 channel fADC, but what is the sample rate for the prototype unit? 
>> 65Msps? I have included the detector 'table' and know that there will 
>> be two flash ADC boards.  Some of the cost spreadsheets I am 
>> reviewing show a 65Msps and 125Msps board so I will sort that out 
>> with Elton.
>> Please reply or call at your earliest convenience,
>>
>> Chris
>> (757)269-5053
>> ~~~~~~~~~~~~~~~~
>>