Are you sure thats not a scsi feature and you just jumpered spin up delay on the drive or for that matter didn't change the jumper.

Generally you can jumper the drive so it doesn't spin up until the controller initializes the disc, then in the controller settings you can set delayed start parameters for each drives. That way on scsi arrays all the drives don't spin up the same time or all spin up at power on. It's supposed to be like that.

But you should be able to configure the drive with a jumper and remove delay start and change IDs, or if you have a interposer they usually have jumpers for IDs and delayed start.

Yes could be the power supply though your spec's read Ok being an AT board which will use mostly 5V and 12v. Neither of those PSU are what one would call "heavy duty-high quality" but if they can work up to their maximums should be OK.
If the computer boots but one drive does not spin could be "HD start up delay" or could also be some "sticktion" of the old drive.
Do both drives exhibit the drive not spinning up or just one of them ? And if always same drive that would point to a drive issue not a PSU issue.

Most likely cause of intermittent start operation like this IMO is jumpering or SCSI cable. bad cable will result in the START UNIT command to start the drives not occurring correctly if they are configured for the controller to start them.

Assuming you plan to use these drives with one boot and one secondary, the following is the jumper config you would set on the drive PCB bottom (not the connector next to the SCSI connector, where all jumpers would be open):

boot drive:

CN11-12 and 23-24: closed
all others open

second drive:

CN1-2: closed
CN11-12 and 23-24: closed
all others open

These settings would have the two drives as IDs 0 and 1 and both spin up as soon as power is applied regardless of controller commands or other considerations. The same jumper settings should apply to either drive as more or less all of Fujitsu's MAx LVD 68-pin drives used the same jumpers. You should probably jumper the optical/zip drives to ID 5 and 6 or the like if they occupy one of those two SCSI IDs.

If the drives still suffer intermittent startup issues it's a power supply problem or faulty drives. Either stated power supply is more than adequate for this system, an ATX to AT converter shouldn't affect the amperage sent down the rails to the molex connected peripherals. If the startup issues go away with these settings then the cable is suspect.

Cheers,

Both SCSI drives have unique ID and have jumpers 11-12 closed to force motor start on power up. No other jumpers are set, except for setting ID. Both have been working fine for weeks until today. The drive that started giving me problems is Fujitsu MAN3367MP 36GB, 10K RPM.

I just tried a different cable and the problem still persists.
I just tried another Fujitsu MAP3735NP 70GB , 10K RPM, for a total of two drives, and both work, so maybe the 36GB drive went bad.

Regarding jumper 23-24 (Termination Power), I have these jumpers off, should they be on, what are they for? I do have lvd terminator device at the end of the cable, right behind the last drive.

Termination power injection. I usually set this jumper on devices when troubleshooting. The adapter supplies termination power to the bus under normal circumstances, but termpower is typically also ideally supplied by the last device on the bus on that cable when the cable length exceeds about 3 feet, or when using a lot of devices (I tended to go with 1 termpower enabled device per 3 bus devices).

Over the years I found that when testing or troubleshooting issues with intermittent drive operation/power, enabling that feature was sometimes a make or break in isolating faulty drives.

I just tried setting the jumper 23-24 (Termination Power) on and it didn't help. Also tried the 2nd power supply and it didn't help. I think the drive is bad. All the other drives start up just fine, as soon as I connect power to them they start up. It's disappointing it lasted only few weeks.

could just need a recap of the logic board

maxtherabbit wrote on 2021-08-15, 16:36:

could just need a recap of the logic board

Where would I find the capacitors on it? Looking at the board I see maybe one small capacitor with the letters "100" on the top, is that it? Or would there be some on the back side of the board?

Nah there's probably nothing on the back. I rescind my statement seeing as there are no electrolytics or tantalums on there. SMD ceramics (almost) never fail

Fortunately they're not a particularly rare or expensive drive, Fujitsu's drives were always amongst the best in class for SCSI and still beat the pants off of even modern consumer spin drives for performance.