Minutes of FAST-20 Study Group
LJLamers at aol.com
LJLamers at aol.com
Sun Jun 5 11:07:16 PDT 1994
6/4/94
---------------------------------------------------------------------------
Minutes of FAST-20 Study Group - June 2, 1994 X3T10/94-131r0
Accredited Standards Committee*
X3, Information Processing Systems
Doc. No. X3T10/94-131R0
Date: June 2, 1994
Project:
Ref. Doc.:
Reply to: J. Lohmeyer
To: Membership of X3T10
From: Lamers/Lohmeyer
Subject: Minutes of X3T10 FAST-20 Study Group June 2, 1994
Agenda
1. Opening Remarks 2
2. Attendance and Membership, Introductions 2
3. Approval of Agenda 2
4. Document Distribution 2
5. Old Business 3
5.1 Single-ended Update [94-xxx] (Ham) 3
5.2 FAST-20 Specs [94-xxx] () 3
5.2.1 Receiver Voltage Level 3
5.2.2 Node Capacitance 4
5.2.3 Cable Length 4
5.2.4 Number of Devices 4
5.2.5 Minimum Strength of Active Negation 4
6. New Business 4
6.1 Transfer Period [] (Lohmeyer) 4
6.2 Voh Termination/Drivers [94-038r3] (Aloisi) 4
6.3 Differential Update [94-xxx] (Ham) 4
7. Meeting Schedule 4
8. Adjournment 5
Results of Meeting
1. Opening Remarks
John Lohmeyer convened the meeting at 9:30 am. John thanked Jim
McGrath of Quantum for hosting the meeting.
As is customary, the people attending introduced themselves. A
copy of the attendance list was circulated for attendance and
corrections.
It was stated that the meeting had been authorized by X3T10 and
would be conducted under the X3 rules. Ad hoc meetings take no
final actions, but prepare recommendations for approval by the
X3T10 task group. The voting rules for the meeting are those of
the parent committee, X3T10. These rules are: one vote per
company; and any participating company member may vote.
The minutes of this meeting will be posted to the SCSI BBS and
the SCSI Reflector and will be included in the next committee
mailing.
2. Attendance and Membership, Introductions
Attendance at working group meetings does not count toward
minimum attendance requirements for X3T10 membership. Working
group meetings are open to any person or company to attend and to
express their opinion on the subjects being discussed.
The following people attended the meeting.
Attendee Company Email Address
Lawrence J. Lamers Adaptec ljlamers at aol.com
Norm Harris Adaptec nharris at adaptec.com
Wally Bridgewater Adaptec wally at eng.adaptec.com
Mark Knecht AMD mark.knecht at amd.com
David Wang AMD david.wang at amd.com
John Lohmeyer AT&T/NCR john.lohmeyer at ftcollins
Microelectronics co.ncr.com
Joe Stoupa Burr-Brown Corp
Dale Smith Cirrus Logic dale at cirrus.com
Louis Grantham Dallas
Semiconductor
Bill Ham Digital Equipment ham at subsys.enet.dec.com
Corp
Ting Chan Qlogic @qlc.com
Jim McGrath Quantum Corp jmcgrath at qntm.com
Jamse Oliver Silicon Graphics oliver at sgi.com
Vit Novak Sun Microsystems vit.novak at sun.com
Jack Shiao Tandem Computer shiao_jack at tandem.com
Pete Tobias Tandem Computer tobias_pete at tandem.com
Kevin Gingerich Texas Instruements
Paul Aloisi Unitrode aloisi at uicc.com
Integrated Circuit
3. Approval of Agenda
The agenda was approved as proposed. Item 5.2.4, 5.2.5, and 5.3
were added during the meeting.
4. Document Distribution
X3T10/94-103r1 Proposed Implementor's Note for FAST-20 Node
Capacitance
5. Old Business
5.1 Single-ended Update [94-xxx] (Ham)
Bill Ham presented his latest data on FAST-20 single-ended. The
data showed that the propagation delay has a significant impact
on being able to run at 20 MHz. Each load on the cable adds
approximately 0.5 ns delay to the signal. Since single-ended is
relying on reflections to get the amplitude up the propagation
delays affects the ability to get a valid signal.
This assumes non-incident wave switching.
The model looks like an AC transmission line voltage divider.
Reducing the number of loads or reducing the node capacitance
improves the situation.
Number of devices
Node Capacitance
Cable Length
Reducing cable length reduces prop delay so reflections arrive in
time to reduce the shelf.
The impedance mis-match between devices and cable is main
problem.
It looks like a table showing the trade-offs of number of
devices, cable length, and node capacitance.
Suggested to calculate distributed capacitance for adding devices
as a function of load and spacing.
Bill Ham noted that given the test results it is likely that 16
devices on a 3 meter bus for FAST-10 has signal issues.
Wally noted that most of the threshold variation comes from power
supply variations. He suggested a 0.5% tolerance on power
supply. Mixed environments (slow and fast) can be a problem
because of the low hysterisis in noisier environments. The
consensus was 5.0% tolerance on power supply voltage. A
temperature range of 0-70 degrees C.
Bill Ham reiterated the need for changing the thresholds. If the
thresholds don't change some other parameter needs to be
modified.
Kevin Gingerich stated that a minimum current for active negation
also needs to be specified to assure that the first step to 2.0
volts is made in the face of a impedance mismatch at the middle
of the cable. The calculation assumed a 3.24 high voltage, a 0.2
low voltage, a 20 ohm impedance mismatch on an 84 ohm impedance,
maintaining the 0.8 volt step. This resulted in a 62 ma driver
requirement which was then reduced by the current input from the
active terminators. The result is a 22 ma minimum active
negation strength driver is required at 2.0 v D.C.
Kevin Gingerich presented a formulae for calculating the allowed
number of devices based on bus capacitance.
Z0=sq(L/C) assuming cable at 10 pf/ft.
Z0'= sq(L/(C+C'))
Z0'/Z0=sq(1/(1+C/C'))
C'/C=.75
C'=7.5 pf/ft
Kevin took an action item to draft this as an informative annex.
5.2 FAST-20 Specs [94-xxx] ()
5.2.1 Receiver Voltage Level
proposed consensus
vIHH 1.9 1.9
vIHL 1.6 1.6
vILH 1.4 1.3
vILL 1.1 1.0
5.2.2 Node Capacitance
20pf maximum, but lower is better
5.2.3 Cable Length
depends on number of devices - 1.5 meters for 8 devices; 3.0
meters for 3 devices
5.2.4 Number of Devices
depends on cable length - 8 devices at 1.5 meters; 3 devices at
3.0 meters
5.2.5 Minimum Strength of Active Negation
vOH (xx) = 2.0 v D.C> min 22 ma negation
5.3 Implementor's Note for FAST-20 [94-103r1] (Novak)
Accepted as modified.
6. New Business
6.1 Transfer Period [] (Lohmeyer)
How to specify the FAST-20 synchronous transfer rate. Because
the transfer period is modulo 4 it is not possible to specify
exactly 20 Mb/sec. What can be achieved is:
20.83 Mb/sec (48 ns period) with a value of 12
19.23 Mb/sec (52 ns period) with a value of 13
Add implementor's note to advise folks of this issue.
6.2 Voh Termination/Drivers [94-038r3] (Aloisi)
The current SPI document does not specify a maximum voltage in
the terminator nor the driver. The consensus is to put in the
limitation on the terminator sourcing current above 3.24 v D.C.
The max Voh will be specified at 3.7 v D.C. max.
The correct approach is to specify the min Vih and leave the max
Vih to the silicon vendor. The test circuit in the document is
used to measure these voltages.
Leave of the 0.0 on Vil and specify 0.8 Vil max. The min Vil is
specified by the vendor.
6.3 Differential Update [94-xxx] (Ham)
Bill Ham reported on the differential testing he has done. One
recommendation is to specify a min of 2 volts. Drop Vod
specification since RS-485 is more restrictive. Another is to
draft more relaxed configuration rules.
Bill stated that differential works so well but has a power
dissipation problem. He suggested that reduced voltage swings
could allow on chip drivers and still offer headroom for higher
transfer rates. On chip drivers reduce skew management issues.
Kevin Gingerich stated that it is technically feasible but common
mode will be reduced. 200 Mb/sec parallel SCSI is a definite
possibility.
7. Meeting Schedule
No future meetings planned on FAST-20.
8. Adjournment
The meeting adjourned at 4:00 PM.
-----------------------------------------------------------------------------
Regards,
Lawrence J. Lamers
Adaptec TEL: 408-945-8600 ext 3214
691 Milpitas Bvld FAX: 408-957-7193
Milpitas, CA 95035 EMAIL: LJLAMERS at AOL.COM
More information about the T10
mailing list