HP 7936 Support Manual page 129

8-19.
SERVO TERROR/DERROR
CONTROLLER (hex/dec)
ERROR
(hex/dec)
1/1
AC/172
2/2 72/114
3/3 73/115
6/6 76/118
7/7 78/120
9/9 B9/185
A/10
79/121
B/11 7A/122
C/12
7B/123
E/14 70/125
10/16 7F/127
13/19
80/128
16/22 AO/173
17/23 1C/28
18/24 66/102
19/25 BA/186
1A/26
BB/187
1B/27 BC/188
1C/28 BO/189
DATA ERROR
TROUBLESHOOTING
To date, there is little empirical data regarding
detailed troubleshooting hints for data errors.
Refer to the H P 7936 and H P 7937 Disc Drive C E
Handbook, part no. 07937-90905, for the latest
troubleshooting information as it becomes
available.
The first step in the determination of the cause of
data errors is to identify the nature of the data er-
ror. This is done most effectively by reading the
ERT log and/or the run time log and analyzing the
information provided. The formats for the ERT
log and the run time log are described in para-
graphs 8-12 and 8-13, respectively. The bits in the
error code bytes for the logs are shown in figures
8-1 and 8-2. Care must be taken to remember
which of the logs is being read as the definition of
the data errors is different in each case.
The next step is to determine whether the error is
"hard" or "soft".
A
hard error is repeatable at a
certain address and indicates that the sector was
written improperly by the electronics, or the media
is defective. Soft errors are much less severe (they
Troubleshooting and Diagnostics
7936 and 7937
do not cause the loss of data) and may in fact be
totally ignored if they occur infrequently.
After noting the error entries, a Read Only Error
Rate Test (RO ERT) should be executed on (and
possibly around) the offending sector(s). The
results of this test should then be compared with
the first log entries to determine if the error is
hard, soft, or gone.
If the error is hard, there may be a media defect
that should be spared. The only way to determine
this is to perform a a Write-Then-Read Error Rate
Test (WTR ERT) on the suspect sector(s). If the
sector(s) being tested contains important data, then
care should be taken to back up the data (if pos-
sible) before any write operation is performed. If
after performing a WTR ERT on the suspect area,
there is still an error in this sector, then the media
is most likely at fault and a sparing operation is in
order. However, before making a decision to
spare, other areas should also be tested in order to
prove that this is not just a random occurrence.
If the hard error disappears after being rewritten,
the media is unlikely to be at fault and the blame
shifts to the electronics.
If the error is soft to begin with, the electronics
will initially be suspect. There is one test that can
be used to differentiate between errors caused by
the servo system and the read/write system. First,
execute a full volume RO ERT and see if there are
any soft errors. Next, execute several random RO
ER T passes (10 or more) and see if the error rate
changes for the worse. If it does, the servo system
may be at fault. If there is no change, then the
read/write system is more suspect.
Another source of soft errors is problems with the
drive environment. Do not overlook the "cleanli-
ness" of the primary power input, and RFI and
ESD levels.
8-20.
CHECKING POWER SUPPLY
VOLTAGES
There are no test points on power supply PCA-A4;
however, the power supply output voltages can be
measured at connector P2 on power jumper cable
W6. See figure 8-3. To reach W6, it is necessary to
remove the rear cover from the drive. Cable W6
8-7