Seagate ST-238 interleave

[Casey Tompkins]

Yes, this is an ancient drive! Sue me, I collect vintage computers.

Anyway, I recently picked up a Compaq Portable with an ST-238 that
wouldn't boot, so I did the old WD ROM debug command g=c800:5, which
kicked me into the controller BIOS.

When it asked me what interleave to use (currently set to 4) I said
"what the heck" and entered 1. After a long time, the low-level format
finished, and I did the usual fdisk & format.

Here's the weird part: I ran Nortons SI, and the disk rating (compared
to a standard PC hard drive) was 0.5! That's ridiculous for a -238. I
fired up the debug routine again, thinking I would change the
interleave to 2 or maybe 3, when the routine said (again!) that the
drive was currently at a 4-1 interleave. What happened? I aborted the
routine without changing anything.

Now, it's been a loonnnng time since I messed around with rll drives,
but maybe this controller threw out the selected 1-1 interleave? It's
an 8-bit WD controller, not 16-bit.

Also, I'm pretty fuzzy on this stuff now. Does anyone remember the
optimal interleave for an 8-bit rll controller? Now that I think about
it, 3-1 comes to mind. I think you needed a faster 16-bit controller
to use 1-1 effectively. When you could, it kicked ass.

Comments? TIA!

P.S. Anyone know where I can find a working Epson QX-10 that won't
cost me an arm and a leg? I saw one for sale for $90 a couple months
ago, but I didn't move fast enough.

 

[Timothy Drouillard]

I also remember 3:1 for some reason.

 

[Guest]

Compaq Portable with an ST-238 that wouldn't boot, so I did
the old WD ROM debug command g=c800:5,
When it asked me what interleave to use (currently set to 4)
I said "what the heck" and entered 1.

I ran Nortons SI, and the disk rating (compared to a
standard PC hard drive) was 0.5! That's ridiculous
for a -238. I fired up the debug routine again, thinking I
would change the interleave to 2 or maybe 3, when the routine
said (again!) that the drive was currently at a 4-1 interleave.
What happened? I aborted the routine without changing anything.

Now, it's been a loonnnng time since I messed around with rll drives,
but maybe this controller threw out the selected 1-1 interleave? It's
an 8-bit WD controller, not 16-bit.

Does anyone remember the optimal interleave for an 8-bit rll
controller? Now that I think about it, 3-1 comes to mind.

? I think you needed a faster 16-bit controller to use 1-1

effectively. When you could, it kicked ass.

I don't remember the lowest interleave those WD controllers would
accept, but 4 was optimum for 8-bit machines, anything from 5-6 was
optimum for 16-bit machines (they either did 8-bit transfers slower or
the DMA didn't work).

An old DOS utility called HDTST128 will let you find the best
interleave and change the current interleave.

 

[Overlord]

Seems I recall the old Spinrite would test the drive to find the optimal
interleave values for data transfer speed....

Yes, this is an ancient drive! Sue me, I collect vintage computers.

Anyway, I recently picked up a Compaq Portable with an ST-238 that
wouldn't boot, so I did the old WD ROM debug command g=c800:5, which
kicked me into the controller BIOS.

When it asked me what interleave to use (currently set to 4) I said
"what the heck" and entered 1. After a long time, the low-level format
finished, and I did the usual fdisk & format.

Here's the weird part: I ran Nortons SI, and the disk rating (compared
to a standard PC hard drive) was 0.5! That's ridiculous for a -238. I
fired up the debug routine again, thinking I would change the
interleave to 2 or maybe 3, when the routine said (again!) that the
drive was currently at a 4-1 interleave. What happened? I aborted the
routine without changing anything.

Now, it's been a loonnnng time since I messed around with rll drives,
but maybe this controller threw out the selected 1-1 interleave? It's
an 8-bit WD controller, not 16-bit.

Also, I'm pretty fuzzy on this stuff now. Does anyone remember the
optimal interleave for an 8-bit rll controller? Now that I think about
it, 3-1 comes to mind. I think you needed a faster 16-bit controller
to use 1-1 effectively. When you could, it kicked ass.

Comments? TIA!

P.S. Anyone know where I can find a working Epson QX-10 that won't
cost me an arm and a leg? I saw one for sale for $90 a couple months
ago, but I didn't move fast enough.

~~~~~~
Bait for spammers:
root@localhost
postmaster@localhost
admin@localhost
abuse@localhost
postmaster@[127.0.0.1]
(e-mail address removed)
~~~~~~
Remove "spamless" to email me.