"New Drives" for MIO

[+Larry]

I thought that instead of having this info "buried" in the extremely long New MIO thread, it might be easier to find drive info in a new topic.

 

If I have found all the info correctly, the following drives are thought to work with the firmware 1.4xx:

 

ST-157N-0 -- (Vintage 50-pin SCSI -- requires 512-byte/sector, low-level format. Note: this is an RLL 2,7 drive, according to the Seagate docs)

ST118273LC -- (18GB, 7200 RPM, 80-pin -- requires 80 -> 50 pin adapter)

 

This is all that I have found listed -- any others?

 

If you're successfully using another drive with 1.4xx, please post it here.

 

-Larry

[+bf2k+]

ST-157N-0 -- (Vintage 50-pin SCSI -- requires 512-byte/sector, low-level format. Note: this is an RLL 2,7 drive, according to the Seagate docs)

 

How did you LL format it in 512 byte? Are you getting 32 MB partitions w SDX 4.42?

 

I have a St157N-1 that I could not get to work with either firmware.

[sl0re]

With the adaptec adapters to MFM and RLL mentioned in the manual, every drive I tried worked... and I tried some real odd balls as they were all I could afford to keep for myself at the time (vs. selling to other people).

[+Larry]

ST-157N-0 -- (Vintage 50-pin SCSI -- requires 512-byte/sector, low-level format. Note: this is an RLL 2,7 drive, according to the Seagate docs)

 

How did you LL format it in 512 byte? Are you getting 32 MB partitions w SDX 4.42?

 

I have a St157N-1 that I could not get to work with either firmware.

 

Yes, I believe that is a problem if you have previously tried to format it to 256 bytes/sector. There are a couple of ways around this. For me the easiest thing was to LL format it with the Black Box formatter at 512 bytes/sector. If you have some vintage MS-Dos software, you likely could do the same thing.

 

Sparta is very nice, but I typically don't use it, so I have only tried MyDos.

 

-Larry

[+Larry]

With the adaptec adapters to MFM and RLL mentioned in the manual, every drive I tried worked... and I tried some real odd balls as they were all I could afford to keep for myself at the time (vs. selling to other people).

 

Hi Slore-

 

Yes, the Adaptec solution works great if you have or can find the older MFM/RLL hardware.

 

But I'm referring to firmware 1.4xx with 512 bytes/sector *and* newer embedded SCSI drives. (OK, the ST157N is borderline, but I still had one from my ST days, and it works great with 1.4xx.)

 

BTW, are you indicating that 1.4xx works with an Adaptec/MFM drive with 512 bytes/sector? Although I'm not sure why you would want to do that -- 1.4xx is somewhat slower than 1.1.

 

-Larry

[sl0re]

With the adaptec adapters to MFM and RLL mentioned in the manual, every drive I tried worked... and I tried some real odd balls as they were all I could afford to keep for myself at the time (vs. selling to other people).

 

Hi Slore-

 

Yes, the Adaptec solution works great if you have or can find the older MFM/RLL hardware.

 

But I'm referring to firmware 1.4xx with 512 bytes/sector *and* newer embedded SCSI drives. (OK, the ST157N is borderline, but I still had one from my ST days, and it works great with 1.4xx.)

 

BTW, are you indicating that 1.4xx works with an Adaptec/MFM drive with 512 bytes/sector? Although I'm not sure why you would want to do that -- 1.4xx is somewhat slower than 1.1.

 

-Larry

 

I missed the bios reference.

[+bf2k+]

But I'm referring to firmware 1.4xx with 512 bytes/sector *and* newer embedded SCSI drives. (OK, the ST157N is borderline, but I still had one from my ST days, and it works great with 1.4xx.)

 

I guess with MyDOS you are only using 256 of the 512 bytes/sec (16meg partitions)?

 

I would have thought that the 1.4b2 fw would not have been necessary with the ST 157 N drives since old 1.1 worked with older embedded SCSI drives (like ST225N).

 

All this gets so confusing

[MEtalGuy66]

But I'm referring to firmware 1.4xx with 512 bytes/sector *and* newer embedded SCSI drives. (OK, the ST157N is borderline, but I still had one from my ST days, and it works great with 1.4xx.)

 

I guess with MyDOS you are only using 256 of the 512 bytes/sec (16meg partitions)?

 

I would have thought that the 1.4b2 fw would not have been necessary with the ST 157 N drives since old 1.1 worked with older embedded SCSI drives (like ST225N).

 

All this gets so confusing

My ST-157N, and ST-138N both work fine with original 1.1 firmware. They only format to half capacity, but they are rock solid stable..

[+warerat]

But I'm referring to firmware 1.4xx with 512 bytes/sector *and* newer embedded SCSI drives. (OK, the ST157N is borderline, but I still had one from my ST days, and it works great with 1.4xx.)

 

I guess with MyDOS you are only using 256 of the 512 bytes/sec (16meg partitions)?

 

I would have thought that the 1.4b2 fw would not have been necessary with the ST 157 N drives since old 1.1 worked with older embedded SCSI drives (like ST225N).

 

All this gets so confusing

 

The problem is that newer devices have a default 512-byte block size and it is usually fixed. Those old MIO low-level formatters set 256-bytes as the block size which new drives will ignore. That means the minimum transfer size during the data phase has to be 512 bytes. From the low-level firmware point of view, we always transfer 512-bytes during any I/O, even when you request 128 and 256 bytes. It's messy enough to have to pad the transaction with dummy acknowledge pulses for short transfers but it is simpler than trying to calculate fitting multiple smaller sectors into one block, hence the 1:1 sector mapping.

 

Total partition capacity is a loose term with 1:1 mapping-- it depends on the density you lay the filesystem out with. For example, if you define 720 sectors in a partition, the raw usable size in native DD 512 (if you use SDX 4.4X) is 360K. But if you're going to use a double-density (256-byte) DOS, then the capacity halves itself to 180K using those same physical sectors and you'll only use the top half. And so on to 90K for single (128-byte), you'll only use the top quarter. At the end of the day, you're always transferring 512-bytes (at least with the 1.4 firmware). Whether you use all the capacity in the sector is dependent on the DOS you pick.

[+warerat]

For those that are having problems with new drives-- check the power to the termination, especially if you're using passive termination. BIG HINT: THE MIO DOES NOT SUPPLY TERMINATOR POWER!

 

I made a change to my MIO to do this, but if the above doesn't make any sense to you and you don't completely understand what that means, you may cook something.

 

I've tried three different Seagate 73.4GB drives, and a few SGI branded IBM 4.5GB drives and they all work.

Edited April 7, 2009 by warerat

[+bf2k+]

My Terminator plugs into the end of the 50-pin cable after the drive and has the word 'ACTIVE' printed on it. This confuses me a little since to me (with my background) 'active' means the terminator supplies its own power but I don't know how this could be true unless the thing has a battery in it - no way to tell - it is a sealed unit.

[kurtm]

My Terminator plugs into the end of the 50-pin cable after the drive and has the word 'ACTIVE' printed on it. This confuses me a little since to me (with my background) 'active' means the terminator supplies its own power but I don't know how this could be true unless the thing has a battery in it - no way to tell - it is a sealed unit.

 

Active termination means it has a voltage regulator to better match cable impedance. Passive is just resistors.

[+warerat]

My Terminator plugs into the end of the 50-pin cable after the drive and has the word 'ACTIVE' printed on it. This confuses me a little since to me (with my background) 'active' means the terminator supplies its own power but I don't know how this could be true unless the thing has a battery in it - no way to tell - it is a sealed unit.

 

Active termination means it has a voltage regulator to better match cable impedance. Passive is just resistors.

 

As we never push this bus past 50KBytes/sec passive is completely acceptable.

[kurtm]

Active termination means it has a voltage regulator to better match cable impedance. Passive is just resistors.

 

As we never push this bus past 50KBytes/sec passive is completely acceptable.

 

 

Wasn't trying to imply otherwise. Just trying to explain the difference. And actually, Active might be trickier for us since the MIO doesn't supply Termpower.

[+warerat]

Active termination means it has a voltage regulator to better match cable impedance. Passive is just resistors.

 

As we never push this bus past 50KBytes/sec passive is completely acceptable.

 

 

Wasn't trying to imply otherwise. Just trying to explain the difference. And actually, Active might be trickier for us since the MIO doesn't supply Termpower.

 

I'm with you...

 

For those of you having termination issues, read this: http://www.scsita.org/aboutscsi/SCSI_Termi...n_Tutorial.html

 

Especially part 6 if you're using SCA to narrow with the MIO (which is what I'm doing).

[+bf2k+]

OK... now that I have put together a new 800XL machine, I finally got my Willem working and burned a 1.4b3 MIO eprom. Plugged it into my ICD 1m MIO and it worked the first time (that's scary - nothing I do regarding this thread or the other one has worked the first time!)

 

I don't know exactly what to attribute this to, but the MIO has not locked up (red activity light on solid) a single time in 10 tests copying 450-500 files per test in varying amounts of nested folders. It could be one or more of the following:

1. I am using a different 800XL from my previous testing. This one has a Rambo 256k, Bryan's Video mod, and AtariMax 32n1 OS. I am using the XL/XE OS with reverse basic (K). (I always use this version for testing).

2. In my original SCSI setup the "active terminator" was over 6" away (a no no) from the last drive on the chain - with a different cable, I have moved it to 1" away.

3. The last tests I did were using 1.4b2 firmware - these used 1.4b3.

Those are the only differences I can think of from my previous bout with this.

 

These tests were all done on a Seagate ST31055N 1gb drive.

 

I used a mixture of 256b/512b sectors in different tests. SDX 4.42 and SD 3.2g were used. I got the best results using 256byte sectors with SD 3.2g.

 

In all of the tests with done SD 3.2g, the copying finished successfully every time. In most of these successes, ICD Cleanup 1.3 report 0 errors. In 2 of the successes, cleanup had to fix the free sector count or bitmap diffs, but no trashed filenames or x-linked files.

 

All of the tests using SDX 4.42 512b either trashed files, filenames, dir entries, x-linked files or aborted with DOS error 150. I really want to try to figure out why I have no luck with SDX 4.42. Maybe I need to do the PBI stabilizer mods? Or maybe I need to put 4.42 in my real SDX cart instead of the Maxflash cart I am using now. (I don't know why that would make any difference). I do know (as Warerat has pointed out), the 512b writing is faster than the 256b so maybe this additional speed is exposing some other flaw in my hardware (like PBI stability).

 

If anyone is interested, I will post my test result spreadsheet although at this time, I may not be finished with this round of tests.

 

Comments, Suggestions?

[+Larry]

Hi bfk2+-

 

I'm certainly interested. FWIW, your errors sound reminiscent of termination issues that I can remember from long ago.

 

I finally got around to running drac030's RWTEST.COM using my MIO 1.4b3 + ST157N, and it returned (apx) 6300 B/sec READING and 6500 B/sec WRITING. I previously could tell it was a little slow, but just how slow was a surprise.

 

I've ordered another 50-pin drive - a more modern Seagate Hawk-2, ST31230N. So I'll see how that goes. If this drive works and is still slow, then I'm going to attempt to duplicate warerat's setup -- drive, termination power, etc.

 

Did you ever get your ST157N reformatted with 512-byte sectors?

 

-Larry

[+warerat]

You still have some problem, as CLEANUP is detecting changes.

 

How are you providing power to the active terminator?

 

Using your external terminator, make sure J2 (on drive) has jumpers only on 3 and 7.

[+bf2k+]

You still have some problem, as CLEANUP is detecting changes.

 

How are you providing power to the active terminator?

 

Using your external terminator, make sure J2 (on drive) has jumpers only on 3 and 7.

 

My J2 jumpers are on:

1 - Term Pwr to Bus

4 - Parity Disable

[+Larry]

I'm getting ready to try out an ST31230N drive with the MIO. It is the only drive on the bus.

 

Among other options the drive jumpers allow the following:

...

Termination enable/disable

Term. power supplied from drive/not

Term. power to SCSI bus/not

Term. power from SCSI bus/not

 

I need Termination enabled (end of SCSI chain).

Can I use "Term. power from drive?"

But I would disable Term. Power to SCSI bus?

 

So, the termination power would come from the drive, but not be applied to the SCSI bus (back to the MIO)?

 

I saw the post showing the MIO-supplied term. power, and I can do that, but if I can provide it from the drive it would be easier.

 

Thanks,

Larry

Edited April 20, 2009 by Larry

[+warerat]

I'm getting ready to try out an ST31230N drive with the MIO. It is the only drive on the bus.

 

Among other options the drive jumpers allow the following:

...

Termination enable/disable

Term. power supplied from drive/not

Term. power to SCSI bus/not

Term. power from SCSI bus/not

 

I need Termination enabled (end of SCSI chain).

Can I use "Term. power from drive?"

But I would disable Term. Power to SCSI bus?

 

So, the termination power would come from the drive, but not be applied to the SCSI bus (back to the MIO)?

 

I saw the post showing the MIO-supplied term. power, and I can do that, but if I can provide it from the drive it would be easier.

 

Thanks,

Larry

 

Yup.

[+bf2k+]

As of this evening, after tearing my hair out for several weeks , I have 2 SCSI drives working Rock Solid with the 1.4b3 firmware using a stock 130XE, ICD 1Meg MIO (unmodified except for FW). My tests were done with 2 DOS versions (SpartaDOS 3.2g and SDX 4.42) using a mixture of densities (256b and 512b per sector) and these drives:

• Seagate ST32430N - 2.1gb
  
• Seagate ST31055N - 1gb
  

Both of these drives can supply term power to the SCSI bus which is necessary on the unmodified MIO as Warerat has pointed out.

 

All of my problems were related to:

• Bad (intermittant) SCSI ribbon cable - replaced it with a new one
  
• Problems with 2 modified 800XL's - replaced them with a stock 130XE
  
• minimal SCSI termination knowledge - Warerat helped to fix that also
  

I feel the problems are solved based on 40 or so tests copying well over a total of 2 gb of A8 files with Cleanup reporting 0 errors. Before resolving the issues, my success was around 25% with virtually no reliability after the copies were done.

 

And both drives are pretty fast. KMK's RWTEST on the ST31055N reports:

 

512bytes/sector

Write: 38125.4088

Read: 41335.9696

Avg: 39730.6892

 

256 bytes/sector

Write: 21226.5789

Read: 23099.5124

Avg: 22163.0456

[+Larry]

As of this evening, after tearing my hair out for several weeks , I have 2 SCSI drives working Rock Solid with the 1.4b3 firmware using a stock 130XE, ICD 1Meg MIO (unmodified except for FW). My tests were done with 2 DOS versions (SpartaDOS 3.2g and SDX 4.42) using a mixture of densities (256b and 512b per sector) and these drives:

• Seagate ST32430N - 2.1gb
  
• Seagate ST31055N - 1gb
  

Both of these drives can supply term power to the SCSI bus which is necessary on the unmodified MIO as Warerat has pointed out.

 

(snip...)

 

 

Congrats on getting these two drives working!

 

Unfortunately, my very nice (silent and cool) ST32130N does not work, but I'd like to check all my jumper positions versus yours.

 

Could you please post the list of your jumper positions with the Seagate descriptions (TE, ME, MD, etc.)

 

I am glad to see you got a couple of Narrow drives working, since that avoids adding adapters.

 

Thanks,

Larry

[+bf2k+]

...

 

Unfortunately, my very nice (silent and cool) ST32130N does not work, but I'd like to check all my jumper positions versus yours.

 

Could you please post the list of your jumper positions with the Seagate descriptions (TE, ME, MD, etc.)

 

Sure... remember this is for UNMODIFIED MIO (no term power from MIO).

 

ST31055N

 

 

Something to remember - some Seagate drives have Parity 'enable' jumpers and some have Parity 'disable' ones. My ST32430N has PE.

Edited April 23, 2009 by bf2k+

[+Larry]

...

 

Unfortunately, my very nice (silent and cool) ST32130N does not work, but I'd like to check all my jumper positions versus yours.

 

Could you please post the list of your jumper positions with the Seagate descriptions (TE, ME, MD, etc.)

 

Sure... remember this is for UNMODIFIED MIO (no term power from MIO).

 

ST31055N

 

 

Something to remember - some Seagate drives have Parity 'enable' jumpers and some have Parity 'disable' ones. My ST32430N has PE.

 

Thanks.

 

Not to be redundant, but you have all the jumpers as shown, or are some of them not attached (e.g. ME, DS)?

 

Yes, my drive is marked for Parity Enable "PE" And, sorry, I keep transposing my drive numbers -- it is a ST31230N.

 

-Larry

Edited April 23, 2009 by Larry