SIIG SC-MP4A12 ATA Ultra IDE Controller - Set Drives Jumpers Addresses
I've had a helluva time trying to get my SIIG ATA Controller to, well, control. It seemed to enjoy crashing more than anything else. So I undertook the scientific route to find an answer....
Unfortunately it's kinda hard to format the results in a web discussion board (you might do well to copy into a word doc and change the font to courier, or other monospaced font) - Just note that each of the tests follow the same format of results across the top, left to right as noted below. Basically almost every test resulted in a crash, save the very bottom solution. This seems to suggest the IDE card is pretty crappy and intolerant of variations of drive size, jumper positioning on drives, location on the ribbon, position on the IDE bus etc - so....
This is my long list of drive jumper settings, ribbon cable positions and IDE UltraATA Card bus settings
I tested to finally find something that was stable for my SIIG UltraATA 130/100 Pro IDE Controller Card,
Model SC-MP4A12. You see the methodical steps I took to try to find the answer of a stable set-up. In
the end I wasn't wholly satisfied, but not unsatisfied either. You see the TEST number, IDE ADDRESS
(that's the bus address on the SIIG CARD - I had a 3rd boot disk, 40GB attached to the mother board),
DISK name followed by its tested PARTITION number and their SIZES, next are the JUMPER SETTINGS
for the tested drives, START UP (SU) outcome, SHUTDOWN (SD) outcome (where KP is "KERNAL
PANIC" and finally the RIBBON POSITION - where on the drive ribbon was the drive located.
I hope this helps someone (hope you'll please leave feedback if it does...).
Andy
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
1. 1 SEAGT 1/189GB Master OK KP END of RIBBON
2. 1 SEAGT 1/189GB CableSelect OK KP END of RIBBON
3. 1 SEAGT 1/189GB Slave OK KP END of RIBBON
Removed power from third drive (was not on the bus) to check for power consumption problems
Rebooted and in single user mode (hold down Apple+S after the chime) I ran file system check:
/sbin/fsck -fy
and found several errors for Invalid Directory, Volume Bitmap and Invalid Volume Free blocks
Next I moved the Seagate 200 GB (186 usable) disk to the second IDE bus address:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
4. 2 SEAGT 1/189 Slave OK KP END of RIBBON
OK, so the drive was crashing EVERY time at shutdown so I didn't test the other drive jumper settings again.
Instead I decided to partition the drive into 2 partitions, roughly 45GB and 135GB. In order to do this you
need to remove the drive from the UltraATA card bus and connect it to the motherboard bus. You might
need to change the jumpers, depending on what your original boot drive is set at. I then set the partitioned
drive's jumpers back to Master. Here's the tests of that:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
5. 2 SEAGT 2/45&135GB Master OK OK END of RIBBON
FINALLY some success!!! I then tried to set the drive back to an unpartitioned drive again to replicate the
previous crash. I opened DiskUtility and it crashed while trying to partitioning it (PLEASE remember you
LOSE ALL YOUR DATA when partitioning - I was using a blank new disk - BACK UP YOUR DATA!)
So, I learned you can't partition a drive on the UltraATA bus it seems. In order to partition it back again to
a usable disk on the UltaATA bus card I had to hook it back to the motherboard bus. I used Disk Utility
again. But this time I made the partitions more extreme: 16 & 173GB. Here's those results:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
6. 2 SEAGT 2/16&173 Master OK KP END of RIBBON
Did all of the above again to get back to a more reasonable partition as it didn't like that split:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
7. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
8. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
9. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
This was good - startup and shutdowns were fine in each repeated case.
In the 9th test I ran a whole whack of apps from different partitions on the boot drive to see their effect.
All was fine. Next I tried my other Western Digital Drive (this was already partitioned earlier and full of
data and applications. I tried it on the 1st IDE address of the controller - the plan is to have each drive
running off its own IDE bus address:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
10. 1 WD 4/Various Master OK KP END of RIBBON
11. 1 WD 4/Various Master OK KP END of RIBBON
I repeated the above for good measure - seemed to work OK - it seems that having the drive partitioned
helps, so long as the partitions are not too small... Next I started testing both drives on separate buses on
the UltraATA card:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
1. 1 WD 4/Various Master OK KP END of RIBBON
2 SEAGT 2/35&154 Master
2. 1 WD 4/Various Slave OK OK END of RIBBON
2 SEAGT 2/35&154 Master
3. 1 WD 4/Various Slave OK KP END of RIBBON
2 SEAGT 2/35&154 Master crashed after opening Text Edit
4. 1 WD 4/Various Slave OK OK END of RIBBON
2 SEAGT 2/35&154 Master
So I restarted the above and got:
5. 1 WD 4/Various Slave OK KP END of RIBBON
2 SEAGT 2/35&154 Master
I switched the slave master jumpers on the drives:
6. 1 WD 4/Various Master OK KP END of RIBBON
2 SEAGT 2/35&154 Slave
I switched their bus address on the IDE controller card:
7. 2 WD 4/Various Slave OK KP END of RIBBON
1 SEAGT 2/35&154 Master
I then took the SEAGATE drive off the bus all together and put it on the ribbon with the WD drive:
8. 2 WD 4/Various Master OK OK END of RIBBON
2 SEAGT 2/35&154 Slave MID of RIBBON
repeated the test above and got a CRASH. So I tried changing the jumpers on BOTH drives to CABLE SELECT
instead:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
9. 2 WD 4/Various Cable Select OK OK END of RIBBON
2 SEAGT 2/35&154 Cable Select MID of RIBBON
I ran the above three times with success, start up and shutdown. So, just for fun I move the SEAGATE drive
back off the same ribbon to "force" a crash and see if it was "Cable Select" that mattered:
10. 2 WD 4/Various Cable Select OK KP END of RIBBON
1 SEAGT 2/35&154 Cable Select MID of RIBBON
...and it did, indeed, CRASH. So I moved the SEAGATE to the END of the RIBBON:
11. 2 WD 4/Various Cable Select OK OK END of RIBBON
1 SEAGT 2/35&154 Cable Select END of RIBBON
...This was good - END of RIBBON and CABLE SELECT and PARTITIONED DRIVE seemed to be the
pieces. I then tested END and MID again (10.) and got the same results as before, a crash. Putting the drive
back the END again seemed to put things right. I noticed something here - It wasn't restarts that were the
problem - it was SHUTDOWNS - it never crashed on a restart - of course it never wholly shuts down then
either - but this suggests a specific addressing issue I'd think.
So I thought I was done: Each drive, on a separate IDE Card address (1 & 2), partitioned in reasonable sizes,
(Goldilocks style - not too big, not too small) , with each drive located at the end of a ribbon cable. Then I
opened TextEdit and started to write this all out. CRASH. Restart. CRASH. Restart. CRASH. Opening any app,
and on the same drive as the boot disk I might add, still lead to a CRASH.
So, the final set up:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
12. 2 WD 4/Various Cable Select OK OK END of RIBBON
2 SEAGT 2/35&154 Cable Select MID of RIBBON
I've been working a half hour now and no CRASH - I'll save this and restart, shut down and restart again and see
if it continues to be stable.
.....
OK - continues to be stable - UGH - that was a LOT of startups and shutdowns. Happier days ahead - I'd appreciate your feedback in trying to replicate these findings in anyway (systematic assessment of a problem can be much more helpful to vendors to solve it ultimately.
G4 Quicksilver 800Mhz 1.5GB RAM Mac OS X (10.4.8) SIIG UltraATA 133/100 Pro Card, 250GB WD, 200GB Seagate drives
G4 Quicksilver 800Mhz 1.5GB RAM Mac OS X (10.4.8) SIIG UltraATA 133/100 Pro Card, 250GB WD, 200GB Seagate drives
Unfortunately it's kinda hard to format the results in a web discussion board (you might do well to copy into a word doc and change the font to courier, or other monospaced font) - Just note that each of the tests follow the same format of results across the top, left to right as noted below. Basically almost every test resulted in a crash, save the very bottom solution. This seems to suggest the IDE card is pretty crappy and intolerant of variations of drive size, jumper positioning on drives, location on the ribbon, position on the IDE bus etc - so....
This is my long list of drive jumper settings, ribbon cable positions and IDE UltraATA Card bus settings
I tested to finally find something that was stable for my SIIG UltraATA 130/100 Pro IDE Controller Card,
Model SC-MP4A12. You see the methodical steps I took to try to find the answer of a stable set-up. In
the end I wasn't wholly satisfied, but not unsatisfied either. You see the TEST number, IDE ADDRESS
(that's the bus address on the SIIG CARD - I had a 3rd boot disk, 40GB attached to the mother board),
DISK name followed by its tested PARTITION number and their SIZES, next are the JUMPER SETTINGS
for the tested drives, START UP (SU) outcome, SHUTDOWN (SD) outcome (where KP is "KERNAL
PANIC" and finally the RIBBON POSITION - where on the drive ribbon was the drive located.
I hope this helps someone (hope you'll please leave feedback if it does...).
Andy
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
1. 1 SEAGT 1/189GB Master OK KP END of RIBBON
2. 1 SEAGT 1/189GB CableSelect OK KP END of RIBBON
3. 1 SEAGT 1/189GB Slave OK KP END of RIBBON
Removed power from third drive (was not on the bus) to check for power consumption problems
Rebooted and in single user mode (hold down Apple+S after the chime) I ran file system check:
/sbin/fsck -fy
and found several errors for Invalid Directory, Volume Bitmap and Invalid Volume Free blocks
Next I moved the Seagate 200 GB (186 usable) disk to the second IDE bus address:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
4. 2 SEAGT 1/189 Slave OK KP END of RIBBON
OK, so the drive was crashing EVERY time at shutdown so I didn't test the other drive jumper settings again.
Instead I decided to partition the drive into 2 partitions, roughly 45GB and 135GB. In order to do this you
need to remove the drive from the UltraATA card bus and connect it to the motherboard bus. You might
need to change the jumpers, depending on what your original boot drive is set at. I then set the partitioned
drive's jumpers back to Master. Here's the tests of that:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
5. 2 SEAGT 2/45&135GB Master OK OK END of RIBBON
FINALLY some success!!! I then tried to set the drive back to an unpartitioned drive again to replicate the
previous crash. I opened DiskUtility and it crashed while trying to partitioning it (PLEASE remember you
LOSE ALL YOUR DATA when partitioning - I was using a blank new disk - BACK UP YOUR DATA!)
So, I learned you can't partition a drive on the UltraATA bus it seems. In order to partition it back again to
a usable disk on the UltaATA bus card I had to hook it back to the motherboard bus. I used Disk Utility
again. But this time I made the partitions more extreme: 16 & 173GB. Here's those results:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
6. 2 SEAGT 2/16&173 Master OK KP END of RIBBON
Did all of the above again to get back to a more reasonable partition as it didn't like that split:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
7. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
8. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
9. 2 SEAGT 2/35&154 Master OK KP END of RIBBON
This was good - startup and shutdowns were fine in each repeated case.
In the 9th test I ran a whole whack of apps from different partitions on the boot drive to see their effect.
All was fine. Next I tried my other Western Digital Drive (this was already partitioned earlier and full of
data and applications. I tried it on the 1st IDE address of the controller - the plan is to have each drive
running off its own IDE bus address:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
10. 1 WD 4/Various Master OK KP END of RIBBON
11. 1 WD 4/Various Master OK KP END of RIBBON
I repeated the above for good measure - seemed to work OK - it seems that having the drive partitioned
helps, so long as the partitions are not too small... Next I started testing both drives on separate buses on
the UltraATA card:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
1. 1 WD 4/Various Master OK KP END of RIBBON
2 SEAGT 2/35&154 Master
2. 1 WD 4/Various Slave OK OK END of RIBBON
2 SEAGT 2/35&154 Master
3. 1 WD 4/Various Slave OK KP END of RIBBON
2 SEAGT 2/35&154 Master crashed after opening Text Edit
4. 1 WD 4/Various Slave OK OK END of RIBBON
2 SEAGT 2/35&154 Master
So I restarted the above and got:
5. 1 WD 4/Various Slave OK KP END of RIBBON
2 SEAGT 2/35&154 Master
I switched the slave master jumpers on the drives:
6. 1 WD 4/Various Master OK KP END of RIBBON
2 SEAGT 2/35&154 Slave
I switched their bus address on the IDE controller card:
7. 2 WD 4/Various Slave OK KP END of RIBBON
1 SEAGT 2/35&154 Master
I then took the SEAGATE drive off the bus all together and put it on the ribbon with the WD drive:
8. 2 WD 4/Various Master OK OK END of RIBBON
2 SEAGT 2/35&154 Slave MID of RIBBON
repeated the test above and got a CRASH. So I tried changing the jumpers on BOTH drives to CABLE SELECT
instead:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
9. 2 WD 4/Various Cable Select OK OK END of RIBBON
2 SEAGT 2/35&154 Cable Select MID of RIBBON
I ran the above three times with success, start up and shutdown. So, just for fun I move the SEAGATE drive
back off the same ribbon to "force" a crash and see if it was "Cable Select" that mattered:
10. 2 WD 4/Various Cable Select OK KP END of RIBBON
1 SEAGT 2/35&154 Cable Select MID of RIBBON
...and it did, indeed, CRASH. So I moved the SEAGATE to the END of the RIBBON:
11. 2 WD 4/Various Cable Select OK OK END of RIBBON
1 SEAGT 2/35&154 Cable Select END of RIBBON
...This was good - END of RIBBON and CABLE SELECT and PARTITIONED DRIVE seemed to be the
pieces. I then tested END and MID again (10.) and got the same results as before, a crash. Putting the drive
back the END again seemed to put things right. I noticed something here - It wasn't restarts that were the
problem - it was SHUTDOWNS - it never crashed on a restart - of course it never wholly shuts down then
either - but this suggests a specific addressing issue I'd think.
So I thought I was done: Each drive, on a separate IDE Card address (1 & 2), partitioned in reasonable sizes,
(Goldilocks style - not too big, not too small) , with each drive located at the end of a ribbon cable. Then I
opened TextEdit and started to write this all out. CRASH. Restart. CRASH. Restart. CRASH. Opening any app,
and on the same drive as the boot disk I might add, still lead to a CRASH.
So, the final set up:
TEST IDE ADDRESS DISK PART/SIZE DRV JMPR SU SD RIBBON POSTN
12. 2 WD 4/Various Cable Select OK OK END of RIBBON
2 SEAGT 2/35&154 Cable Select MID of RIBBON
I've been working a half hour now and no CRASH - I'll save this and restart, shut down and restart again and see
if it continues to be stable.
.....
OK - continues to be stable - UGH - that was a LOT of startups and shutdowns. Happier days ahead - I'd appreciate your feedback in trying to replicate these findings in anyway (systematic assessment of a problem can be much more helpful to vendors to solve it ultimately.
G4 Quicksilver 800Mhz 1.5GB RAM Mac OS X (10.4.8) SIIG UltraATA 133/100 Pro Card, 250GB WD, 200GB Seagate drives
G4 Quicksilver 800Mhz 1.5GB RAM Mac OS X (10.4.8) SIIG UltraATA 133/100 Pro Card, 250GB WD, 200GB Seagate drives
