Minutes/AMS 02 GDAQ INTERFACE, ETC MEETING 13 APRIL 2000 at CERN
This message available at
URL=http://ams.cern.ch/AMS/Electronics/Minutes/000413.html
Minutes from AMS 02 GDAQ INTERFACE, ETC MEETING
13 April 2000 at CERN
Mike's minutes from meeting on GDAQ Interfaces, 13 Apr 00 at CERN,
with help from Hann and Andre.
* appended ("@" means hardcopy available at Secretariat)
@ Mike's GDAQ interface slides.
Ping Ye's minutes of meeting 8 Apr.
* referenced ("@" means hardcopy available at Secretariat)
R.Becker DAQ Crate mag file drawing
M.Capell crate drawing
@ M.Capell power architecture minutes
@ M.Capell power architecture drawing
X.Cai DAQ Link protocol proposal
@ V.Koutsenko DAQ Proposal
A.Kunin ---"---
@ G.Viertel pSRD Board Production
+ see URL=http://ams.cern.ch/AMS/Electronics, especially /Figs
* attendees:
M.Capell (till 12:30)
L.Baldini
A.Kunin
X.Cai
V.Koutsenko
M.Porciani
F.Petroni
G.Castellini
F.Cindolo
G.Ambrosi
E.Cortina
M.Menichelli
S.Blasko
H.Chang
U.Roser
S.Ting
T.Dai (from 14:00!)
* Mike et al reported on trip to CSIST/Taiwan.
See Ping Ye's minutes of meeting 8 Apr (attached).
* Discussion about responibilites for other electronics
(eg: RICH, ECAL) and additional production at CSIST.
To be continued at meeting next week.
* Discussion on GDAQ Interfaces - see below....
* Discussion about beam test requirements - my understanding:
+ if we have naked chips it is easiest to test them
at ESA cyclotron in Belgium.
+ for packaged chips it is best to use relativistic v.Heavy ions
at GSI (eg 1Gev U-238 gives LET of 40 Mev-g/cm2).
+ Need access to GSI beam for:
- Power PC 750 chip set
- ADSP-218x-yzzz (x=7,y=L ??)
we plan to use as difficult (impossible ?) to get "naked".
- difficult in general to get BGA naked and exposed.
* 12week ARO delivery of PowerPC Eval Kit.
- delay of beam test board ?
* Discussion about possibly foundry access in Taiwan.
+ H.Chang to investgate.
* Can we use "AMS" process for the GA ? (instead of Temic?).
It is cheaper, faster, ...
* Discussion about Guido board:
+ prototypes: 2 working, 4 more in work. 1 reserved for CAEN.
+ Requires ALTERA QUARTUS software,
+ Requires BYTE BLASTER - Geneva has offer.
+ Requires ADSP-218x ICE
Mike is getting about 8, plus 5*sw and 3 eval boards
+ 20 boards need to be ordered asap to maintain price/delivery.
+ Standard Config is 200KGates Altera plus ADSP-2187-L....
+ Guido Board allocations:
Proto Prod ADSP Byte Visual
Group/Contact: type ICE Blast DSP
/Guido 2 ? 2 2 1
Bologna/Cindolo 3 ?
MIT /Koutsenko 1 4+1 2 2 1
Geneva /Cortina 1 ? 1 1
Pisa / 2 ?
CIEMAT / Marín 2 1* 2* 1*
Taiwan /Chang 1 1+1 1+1 2 1
Aachen /Schwering 1 ?
Peruiga/Menichelli ? .+.+2 1 1 1
^US-110VAC
^2185
^400 KGate
note 17 Apr: * Madrid group has ordered by themselves.
NOTE: CERN GROUP REQUESTS "CAEN" PROTOTYPE WHEN THEY ARE DONE WITH IT,
FOR EXAMPLE IN JUNE, SO THEY CAN CONNECT TWO TOGETHER.
* PSRD electronics by U.Roser from G.Viertel's transparencies (which see).
+ 13 different boards are needed. Planning:
- Design/prototyping RWTH, ETH, MIT
- Components delivery design team
- Component screening CSIST
- Board Production CSIST
- Board loading CSIST
- Boards are then delievered to RWTH without any test.
- Board Testing ETH/RWTH
- Conformal coating Contraves
- Vibration test RWTH
- Thermal Vacuum MPI
- EMC test Contraves
+ Xudong and Chang both question the wisdom of deliever the boards
to RWTH without test. Chang requests that the design people should
visit the CSIST at least once before the production. Ting says
we will discuss this after PSRD flight is confirmed (with a
signature) by NASA.
* Discussion of the DC-DC converter.
+ Mauro mentioned a few companies (MDI and ... ) are candidate vendors.
+ The most critical question is the efficiency.
+ Guido suggested a new option that we do customed design ourself.
He believes a 10% gain is feasible.
+ Since Guido (and CAEN) are already doing a design for
"another experiment", it seems a good option to pursue further.
+ Ting asked if CSIST can do it.
+ Chang mention that they don't have experience with magnetic field,
and they don't have good database for space qualified components.
+ It is decided that CSIST MAY do the production, but not the design.
+ I COMPLETELY DISAGREE:
- we are already searching for vendors with known, proven designs
that can be optimized for the 28+-1VDC input in terms of effic.
using full hybrid design for reliability, heat sinking, etc.
- They will be able to exactly model what the mods to their
design will do, not just guess.
- we will have several choices, including mag field problem.
* The July meeting in Taiwan.
+ Guido will organize half of the agenda.
+ Mike with Sam will organize the other half.
+ Participants (updated):
- 4 July - 11 July
Tiesheng Dai,
Xudong Cai,
Andrei Kunin,
Alexei Lebedev,
Vladimir Koutsenko,
Mike Capell,
Luca Baldini,
Guido Castellini,
Sandor Blasko
- 6 July - 11 July
Hanspeter von Gunten,
Giovanni Ambrosi,
Pascal Emonet,
Mauro menichelli,
Massimiliano Porciani,
Franceso Petroni,
Federico Cindolo,
Georg Schwering,
Vassily Plyaskine,
Eduardo Cortina,
Roberto Battiston,
CIEMAT representatives,
PISA representatives
- Sam Ting
+ Note that this will be followed by a meeting on the cryomagnet case.
=============================================================================
Follows copy of Mike's slides on GDAQ interfaces...
=============================================================================
* Discussion of GDAQ Interfaces (Mike):
+ Environment
+ Thermal/Mechanical
+ Power
+ Fiber Optic
+ CAN
+ DAQ
* Environment
+ Static Loads
+ Vibration
+ Pressure Change
+ Atomic Oxygen/Solar UV
+ Zero G
+ Vacuum
+ Thermal Ranges
+ Radiation Dose
+ Single Event Effects
+ Magnet Field
see drawing from R.Becker, at crate corners (assuming GDAQ
crate is on lower rail, centered on y=0) B=540,370,250,270 Gauss.
Field at y=0 is not "appreciably" higher.
+ EMC
Most of these are well specified by approriate NASA docs,
we need to identify the appropriate ones....except as
noted.
* Thermal/Mechanical
+ Volume/Dimensions
+ Mounting Points
+ Weight
+ Power Dissipation
- by Radiation
- by Conduction
For these we have only preliminary ideas, but enough to start with.
See sketch "crate_0" from Mike as a suggestion for discussion.
* Power
+ 2 possible levels to specify
- 28 VDC :: AMS PDB
- 5 & 3.3 V from PSC (Power Supply, Crate)
+ 28 VDC interface per power meeting min 20 Mar 00
and sketch power_arch_0. Initial specs from P.Lipoma:
- Output voltage tolerance: +/- 1.0V
- Line & load regulation: 0.5V TBR
- Transient response to 20% load change: +/- 0.35V, with 1mS recovery
- Total output voltage spread from all sources: +/- 1.5V TBR
- Current limiting: 42 Amps
- Output noise amplitude and spectrum:
Complies with NTS-21000-ISD-ISS,
which is 0.9V p-p 30 Hz to 7kHz,
then decreasing at 10db/decade to 0.28V at 70kHz,
then remaining at this level out to 400MHz.
+ PSC for Crate is ETH/MIT responsibility.
+ Nothing so far for LV.
+ Also need to specify NASA Grounding Spec.
* Fiber Optic
+ In AMS 01 we had 2*transmit
+ In AMS 02, to allow full utilization of inboard coumpting
equipment, we need 1*transmit + 1*receive.
+ Low level interface in NASA docs
+ High level interface as AMS 01, with adjustments.
* CAN
+ This can be/will be copied from AMS 01
+ Need to provide estimate of data rate from MCC to be copied
into high rate data link.
===========================================================================
Except for the details of crate mechanics, I think everything above
here is pretty well known, agreed upon, and just work to pull it
all together and sort the details. Mike to coordinate.
===========================================================================
* DAQ Links
+ Please see X.Cai's proposal on the AMS web page.
+ Figures for alternate proposal by V.Kousenko.
+ Some discussion ensued - of course not enough people
were familiar X.Cai's proposal (including me).
+ There is also some write up by A.Kunin (I haven't seen?)
+ next meeting this topic Tues 18 Apr afternoon.
===========================================================================
Ping Ye's minutes of meeting 8 Apr.
===========================================================================
From pyeh@sinica.edu.tw Fri Apr 14 12:27:36 2000
Date: Mon, 10 Apr 2000 16:46:25 +0800
From: Ping Yeh
To: AMS Electronics Taiwan
Cc: Ping Yeh
Subject: Minutes of AMS Formosa Regent Taipei Meeting 8 April 2000
Dear all,
This is the minutes for the meeting held in Formosa Regent Taipei
on 8 April 2000. I'm using the newly set up mailing list
ams-ele-tw@hepmai.phys.sinica.edu.tw
to distribute this e-mail. Send a mail to
majordomo@hepmail.phys.sinica.edu.tw
with message body
subscribe ams-ele-tw your_email_adress
to subscribe. If you receive this mail directly, you are already
subscribed.
cheers,
Ping
=================================================================
Minutes of AMS Formosa Regent Taipei Meeting
8 April 2000
=================================================================
Time: 10:30am April 8, 2000
Place: Formosa Regent Taipei
Attendees:
MIT: S.C.C. Ting, Mike Capell, Vladimir Koutsenko
CSIST: Te-Sing Wang, Hao Jinchi, Yuan-Tzu Ting,
Yung-Jing Fanchiang, Huann-Ming Chern,
Yin-Kuan Lin, Jan-Wu Hong, Kuang-Shyr Wu
NCU: Yuan-Hann Chang
AS: Shih-Chang Lee, Ping Yeh
1. Responsibilities of CSIST
Mike Capell briefs his understanding of responsibilities
of CSIST regarding AMS-02 electronics.
- Production:
1. Tracker read-out boards (TDR and JDQT),
Tracker power supply boards,
Tracker crates with backplanes.
2. TRD read-out boards (TDR level and JDQ level),
TRD High Voltage control, TRD gas monitor system,
TRD crates with backplanes.
3. Global DAQ boards and crate with backplane.
4. Trigger boards.
5. MCC board.
- Design: Global DAQ system.
- Circuit Design: Level-1 Trigger board.
Yuan-Tzu Ting asked about the production of Tracker Power
Supply (TPS). The understanding is that
A. So far, production of the TPS is a request to be
considered by CSIST.
B. The (Italian) designer is responsible of making the
prototypes and do the environment test on the prototypes.
C. If CSIST agrees to produce TPS, CSIST will receive the
schematics of the board for production from the designer.
The designer is required to go to CSIST to assist in the
production and testing of TPS.
The Level-1 Trigger board is discussed. The understanding
is that the conceptual design based on physics will be
developed in CERN, and the circuit design will be done by
CSIST.
S.C.C. Ting mentioned that there may be requests of
production from other collaborators of AMS-02 in the future.
He emphasized that CSIST's effort should be concentrated on
the "major parts" mentioned above. Additional requests must
go through Mike Capell and Yuan-Hann Chang.
2. Weekly Video Meeting
S.C.C. Ting proposes to set up a weekly video meeting between
CSIST and people in CERN. Ping Yeh agreed to be responsible
for setting it up in Lung-Yuan Park.
3. Procurement
- Test kit of PowerPC 750: Mike Capell and Shih-Chang Lee
will purchase an identical test kit that Mike has ordered
and give it to CSIST.
- Components: In the future all procurement of components
will be done by Shih-Chang Lee and Mike Capell. Shih-Chang
Lee is the point of contact for CSIST people.
4. Point of Contacts
CSIST side:
production: Yung-Jing Fanchiang
design: Huann-Min Chern
AS side (procurement): Shih-Chang Lee
MIT side (procurement): Mike Capell through Shih-Chang Lee
5. E-mail Exchange
Mike proposed to use e-mail to exchange electronic documents.
Dr. Hank in Lung-Yuan Park is assigned as the contact person
who will forward e-mails from internet to CSIST intranet
and vice versa. The address to use is
hank@saturn.yzu.edu.tw
and every e-mail should CC to Yung-Jing Fanchiang's personal
address
fcyj@ms11.hinet.net.
Ping Yeh will set up a mailing list for everyone related.
[minute taker's note: the mailing list is set up in Academia
Sinica with address ams-ele-tw@hepmail.phys.sinica.edu.tw.
Seven addresses are already subscribed:
Mike Capell, Yuan-Hann Chang, Yung-Jing Fanchiang,
Dr. Hank's address for CSIST forwarding,
Vladimir Koutsenko, Shih-Chang Lee, and Ping Yeh.]
6. Test Flight
Yuan-Tzu Ting asked about test flight. S.C.C. Ting's answer
is "not decided yet".
The understanding is that it is possible to have a PSRD
(prototype synchrotron radiation detector) test flight in
February 2001, but still needs final confirmation from NASA.
If the flight is confirmed, ETH may request CSIST to produce
electronic modules for PSRD.
7. Radiation Test
Yuan-Hann Chang describes the planning for radiation test.
Target date: middle to end of July 2000.
Test period: about 1 week.
Target Place: Darmstadt
A test board with a PowerPC 750 and a Motorola MPC107 is
needed for radiation test. CSIST will design and make
this board with components provided by Academia Sinica.
Ping Yeh will provide data sheets of both chips to CSIST.
S.C.C. Ting will request beam time after Vladimir determines
the beam ion type and energy.
8. TRD High Voltage Printed Circuit Board Pre-production
Vladimir Koutsenko had a design of the HV PCB for TRD and he
requested to do pre-production in CSIST. He will give Mike
Capell the component list and Mike Capell will work with
Shih-Chang Lee for procurement. Yung-Jing Fanchiang is
the CSIST contact for pre-production.
9. Minutes of CSIST Meeting on 7 April 2000
Yuan-Tzu Ting went through the minutes of meeting held in
Lung-Yuan on 7 April 2000. After minor changes, the minutes
is agreed upon by all parties. For the record, the minutes
is re-produced here.
1. Two kinds of GDAQ architecture are presented.
- Router-based GDAQ: JL3 CPU conducts all event building.
Need to design router and other glue logic using non-rad
hard FPGA.
- EB-based GDAQ: Using Guido board (same as other detectors
used for JDQT) for event builder. Need to study Analog
Device ADSP-2189 and APEX FPGA for hardware and firmware
design.
CSIST will decide which way is the most efficient approach
to design the GDAQ system.
2. The selection of CPU for GDAQ design
- Using the PowerPC 750 for JL3 computer design.
- Mike Capell will provide the ICD (Interface Control
Documentation) based on CSIST's request by the end of
April.
- Mike Capell will provide PowerPC 750 evaluation kits,
hardware/firmware developing tools to CSIST within
3 months.
- CSIST will propose the detailed architecture of GDAQ
to Mike Capell and Yuan-Hann Chang within one month
after receiving ICD.
- Arrange the rad-hard test for commercial PowerPC 750
chip sets. Mike Capell will propose the test procedure,
test location and performance check for any SEU/latchup.
Schedule and UUT are TBD.
3. The state diagram of using 1355 / SpaceWire protocol
remains open. Dr. Koutsenko's design is under consideration
by CSIST. Need to confirm before starting the GDAQ design.
[minute taker's note: Vladimir Koutsenko gave a copy of
his drawing to CSIST in the meeting.]
4. Academia Sinica will provide the commercial chipsets and
other necessary components for the prototyping design
of GDAQ boards.
5. A Technical Exchange Meeting will be held 7 July - 10 July.
All design-related tasks need to be confirmed or finished
prior to Final Design Review. [minute taker: schedule TBD]
6. The selection and design of memory system is twofold,
memory for event data (4G) and memory for system (program
and data). Only system memory is rad-hard approved, but
memory for event data need to have ECC/CRC to detect the
error with optional correction.
10. Effect of Magnetic Field on the Electronics
S.C.C. Ting asked about the effect of magnetic field on
the electronics. S.C.C. Ting will provide a test facility
to CSIST with magnetic field strength up to 1 kilogauss.
Testing in magnetic field will be included in the environmental
test protocol.
11. Possible Collaboration with Acer
Shih-Chang Lee mentioned that Acer, the biggest computer company
in Taiwan, may contribute manpower in the AMS-02 project on
hardware and/or software. Te-Sing Wang welcomes additional
help. Yuan-Hann Chang emphasized that even with Acer in the
team the center is CSIST.
12. Participants of the Technical Exchange Meeting
4 July - 11 July
Tiesheng Dai, Xudong Cai, Andrei Kunin, Alexei Lebedev,
Vladimir Koutsenko, Mike Capell, Luca Baldini,
Guido Castellini.
6 July - 11 July
Hanspeter von Gunten, Giorvanni Ambrosi, Pascal Emonet,
Sandor Blasko, Mauro Menichelli, Massimiliano Porciani,
Federico Cindolo, Georg Schwering, Vassily Plyaskine,
plus representatives from Pisa and CIEMAT.
END OF MEETING MINUTES
****************** AMS ELECTRONICS TAIWAN ***********************
* To the whole list: *
* ams-ele-tw@hepmail.phys.sinica.edu.tw *
* To the list management program (subscribe/unsubscribe/help): *
* majordomo@hepmail.phys.sinica.edu.tw *
* To the human list manager (comment/suggestion): *
* pyeh@sinica.edu.tw *
*********************************************************************