SPI-2 Letter ballot resolution

Gene_Milligan at notes.seagate.com Gene_Milligan at notes.seagate.com
Mon Feb 16 11:02:08 PST 1998


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* Gene_Milligan at notes.seagate.com
*
Regarding 98-108r1:

<<c) to provide continuity between reserved pins and ground pins between
devices and
terminators.>>

     This wording could lead to grounded reserved lines. I suggest changing
it to:

"c) to provide continuity between all devices and terminators for reserved
pins as well as ground pins."

<<The maximum sine wave signal attenuation shall be 0,095 dB maximum per
meter at 5 MHz,
measured deferentially or a maximum sine wave signal attenuation of 1,41 dB
at 5 MHz for the
entire bus measured deferentially.>>

     Change "deferentially" in two places to "differentially".

<<A good reference would be either figure 45 or table A.2. Accepted>>

     Accepted which?

<<32) The desire for recipients not to check for more recent, or
inadvertent use, of reserved bits has been in place for a long while. Is it
not time to be more blatant than "Recipients may check reserved bits,
bytes, words or fields for zero values and report errors if non-zero values
are received."? Rejected>>

     I reject the rejection. It is wrong to state the requirement backwards
to the decision taken by T10. This should be changed to something like
"Recipients are allowed to check reserved bits, bytes, words or fields for
zero values and report errors if non-zero values are received although this
may impede future enhancements."

<<65) I have repeatedly argued that the voltages specified in Table 22 for
HVD are component specifications and are not appropriate for the SCSI
environment. They merely lead to FUD about an inadvertent inclusion of
another version SCSI device on the HVD bus. In addition to the likely hood
that the cables would burn up with these values, if they did occur they
would also occur with LVD. Rejected: This is part of 485 and there is
nothing we can do about it.>>

     I reject the rejection. It is non-sensical to have the component
standard dictate the system requirements. It should be the reverse. T10
would be derelict to bury their head in the sand on this issue. The sources
of the offset voltage are independent of whether HVD or LVD is being used.
Once again, even if it made a difference which type was being used, the
cable would have burned up prior to the transceivers being allowed to come
out of the high impedance mode to attempt signaling. The specifications for
the insulation are in hundreds of volts, this does not make it necessary to
have the devices accept hundreds of volts. The statement "This is part of
485 and there is nothing we can do about it." is spurious and has no
relationship to the application.

<<For greater reliability I suggest changing the requirement to "A device
shall not change its present signal driver or receiver mode based on the
DIFFSENS voltage level unless a new mode is sensed continuously for at
least 25 micro seconds." or to "A device shall not change its present
signal driver or receiver mode to a different mode other than high
impedance based on the DIFFSENS voltage level unless a new mode is sensed
continuously for at least 100 ms "
Rejected>>

     I request that the working group and the plenary vote to confirm that
delaying the change to high impedance or the new mode for 100 ms while
being impinged by the wrong mode will not reduce component MTBF (i.e.
confirm the rejection) as an individual item.

<<a) RST signal shall asserted continuously by the initiator during removal
or insertion.>>

     Should be "shall be".

Thank you for the changes that have been made. However unless item (65) is
appropriately resolved, my negative will remain an unresolved negative to
be forwarded with the draft standard.

Gene


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com




More information about the T10 mailing list