Probably a poor designed SCA adapter. I remember using a few different ones and had good luck with some versus other but only on SCA to 68 pin..
As far as ISA, why ? SCA are a SCSI wide and cannot be easily downed to 50 pin ISA, that is why there is 68pin wide and 50 pin narrow.
sorry being stubborn but most SCA drives typically cannot be downed to narrow 50 where some 68 pin wides could....

Horun wrote on 2022-03-20, 02:10:

Probably a poor designed SCA adapter. I remember using a few different ones and had good luck with some versus other but only on SCA to 68 pin..
As far as ISA, why ? SCA are a SCSI wide and cannot be easily downed to 50 pin ISA, that is why there is 68pin wide and 50 pin narrow.
sorry being stubborn but most SCA drives typically cannot be downed to narrow 50 where some 68 pin wides could....

This drive is in fact, narrow (that's what the 'N' in the name indicates) and according to the manual it should be pretty much identical to the ST5660N 50-pin version except for the connector. Also, on the AHA-2940W it works using the same adapter with a 50 pin cable, I'm not using the wide port at all on that one. So I don't think the adapter is to blame in this case.

I'd like to use it with an ISA controller due to the low size (545MB), for a PCI system I'd prefer larger drives.

EDIT: attached pictures of the adapter

Does your drive spin up when it just has the power connection and no SCSI cable / adapter at all?
But I guess you need the SCA adapter for power supply.
What are the jumpers at your SCA adapter for, but for the IDs? Perhaps they control the spin behavoiur.
What happens when connecting your disk to another SCA adapter without SCSI cable?

It looks like the jumpers on your adapter are for RMT_START (pin 38, remote start) and DLYD_START (pin 78, delayed start).

Disruptor wrote on 2022-03-20, 05:40:

Does your drive spin up when it just has the power connection and no SCSI cable / adapter at all?

Nope, with just the adapter connected to the power supply and no SCSI cable there's no spin up, only the drive LED lights up.

Disruptor wrote on 2022-03-20, 05:40:

What are the jumpers at your SCA adapter for, but for the IDs? Perhaps they control the spin behavoiur.

You mean the ID1,ID2,ID3,ID4 jumpers on the adapter? Those are all open (i.e. ID set to 0).

Disruptor wrote on 2022-03-20, 05:40:

What happens when connecting your disk to another SCA adapter without SCSI cable?

I don't have any other adapter right now, but I may try to get another one.

Disruptor wrote on 2022-03-20, 05:40:

It looks like the jumpers on your adapter are for RMT_START (pin 38, remote start) and DLYD_START (pin 78, delayed start).

That is correct, I've traced those jumpers in the adapter to those exact pins in the connector. And in fact RMT_START is also connected to pin 15 in the drive jumper block, which according to the manual should get the drive to wait for the "Start Unit" command from the controller if you close pins 15-16 (connect RMT_START to GND), while leaving it open should make the drive spin up as soon as it's powered on. However, that jumper seems to have no effect regardless whether it's open or closed on the adapter, the drive or both. In all cases the only way to get the drive to spin up is during the SCSI BIOS init, as long as I set the "Send Start Command" to on for drive ID 0 in the controller config.

Now that I think about it, I think I have a broken ST5660N (50-pin connector version of this drive) in my box of dead parts. If that's indeed the case then maybe I could attempt a PCB swap and avoid the use of the adapter, but I will probably have to transfer the firmware from the working drive due to calibration data specific to the drive. I'll see what I can do.

TheMobRules wrote on 2022-03-20, 08:30:

Disruptor wrote on 2022-03-20, 05:40:

What are the jumpers at your SCA adapter for, but for the IDs? Perhaps they control the spin behavoiur.

You mean the ID1,ID2,ID3,ID4 jumpers on the adapter? Those are all open (i.e. ID set to 0).

No, I exactly mean the other ones. Play with them.

Oops I looked up the wrong drive... did not see your manual. my bad. according to the manual you should have an option block of jumper pins on the drive opposite side from the 80 pin connector.
page 5, 18 and page21. "Start/stop option-When a jumper is installed on pins 15 and 16 of the options jumper block, the drive waits for a Start/Stop Unit command from the host beforestarting or stopping the spindle motor."
ok not sure why if no jumper it does not spin up right away...

You could try the delay start jumper, DS on the adapter. The drive should wait a few seconds then start up, regardless of it receiving a start command. Did you say the jumper block on the front of the drive is empty?

The drive is working as intended.

This is one of the quirks of SCSI. SCA was designed to take away all the jumper bullshit from earlier SCSI and make it "plug and play" for server people. Staggered startup is HIGHLY desirable for these people as the power infrastructure wasn't nearly as robust as it is today. So having fifteen 10k-rpm drive spin up at once is a BAD thing, and only gets worse if you have multiple banks of drives. So having the controller do the start up sequence is the way everyone went.

However older controllers dont do this, as you have noticed. And the solution is to use the motor start jumper. SCA drives dont have these. The cheap Chineseium adapters don't help, as they often wont have this jumper on the card.

You have a few options.

1- get a better scsi controller. This is likely not applicable if you are stuck with ISA as the 1542 is probably *THE BEST* ISA SCSI adapter out there. This isn't a likely choice.

2- get proper 50-pin drives. (this is the best option imho) Also doing a board swap is a HORRIBLE idea. There is memory on the board that has memorized the "bad" sectors on the physical disk, and unless you have the software to reprogram these flash chips and the requisite experience, swapping the boards between 2 different drives will likely result in sadness.

3- get a proper SCA adapter. This is the option I often go with, but its because I have a vast experience with SCSI.

Adapters are nice, but 9/10 are fucking garbage and you are honestly just better off getting the correct drive for the system you have. The 1542 is a great controller, but you are going to have trouble with hard drives >1gb anyway, so a small 50-pin drive is the best choice imho.

Well, well. Not to be a contrarian, but I found my broken ST5660N (50-pin version of this drive) and after verifying that the PCBs are mostly identical (except for the connector section) and the screw positions match perfectly, I went ahead and attempted a swap. It was successful (mostly) and confirmed some of my initial suspicions about something being baked into the firmware regarding the motor start option. So here's what I did:

First of all, as expected just swapping the PCB from the broken 50-pin drive to the working one didn't work, there is drive-specific firmware with bad sector data (and probably other stuff) that needs to be transferred first.

So, with that in mind, I examined the PCBs in search of a flash chip, and sure enough, it was the only chip with a label on it: an AM29F010, which can be flashed with the TL866 as long as you have the PLCC32 adapter! Nice!

I used hot air to remove the chip from the PCB of the working 80-pin drive and dumped it with the programmer (attached below). Then I soldered it back to the PCB and verified that the drive is working fine and I didn't mess up anything. Good

Finally, I removed the chip from the PCB of the broken 50-pin drive, flashed it with the firmware I had dumped and soldered it back. Then, I attached the 50-pin PCB to the working drive and... it works perfectly!! But it has the same problem as when I use the adapter, it completely ignores the "Motor Start" jumper setting and the drive needs to be told by the controller to spin up in all cases. So I guess that's definite proof that there is something in the firmware causing this behavior.

So that's it, it was nice to experiment even if the results were not ideal, at least now I don't have to use the adapter anymore, heh. Maybe someone with the proper experience could take a look at the firmware and see what's going on, I tried to disassemble it but I don't even know where to begin, I don't think it's x86 code.