• Please review our updated Terms and Rules here

WDXT-GEN + Seagate ST225 lowlevel formatting take forever

S

SunDown79

Veteran Member
Joined
Sep 8, 2005
Messages
785
Location
Netherlands
So its me again with another lowlevel thingie.

This we have; one no-real-brand XT board with a Siemens 8088, stock IBM floppy controller wiht 5.25" 360K drive, hercules videocard from Tulip and the WDXT-GEN (option 300) connected to a working (tested before) ST225 harddisk.

My purpose it to test the XT-GEN card by first lowlevel formatting a harddisk, highlevel formatting after that and booting it.

I booted MSDOS 6.x and started debug, ran G=C800:5, produces all the correct questions and type and I just went for non-dynamic and interleave 3.
It starts off formatting but the further along it gets the more seeks I hear and thus it take aaaaaaages. It does step-seek-step-seek etc

What could be causing these seeks ? I was expecting it to just do step-step-step-step about 615 times and done.
Ow I did try with other interleaves but doesn't seem to matter.
 
Try specifying dynamic formatting, and specify only a few tracks. See if it finishes with or without an error, and if those tracks are readable.

Most bios formatters like that don't even verify what they write. But some formatting programs (like SpeedStor) may perform a slow track zero seek after each track.

If the tracks it formats are readable, you might just have to wait it out.
 
Good idea!

Any idea on how to go about that ?
This are the normale settings according to Seagate for a ST225 : 615 4 616 300 11 3
The first two I understand, the last two best leave alone but how about starting reduced write cylinder and write precomp cylinder, those are pretty unknown to me?
 
The RWC and Precomp deal with some drives needing special handling of the inner cylinders (shorter linear tracks mean tighter packed bits).

Precompensation is a method of compensating for the shorter track length by applying slightly altered timing when writing to compensate for "bit crowding" effects. Where to apply precompensation, for drives that need it is usually about halfway through the disk, so 300 is a good estimate.

RWC (Reduced write current) is an older method of precompensation. It appears on some drives as a signal oncontrol interface pin 2. On the ST225, it's NC, so setting it to 616 is a way to get it out of the way.
 
SunDown79 said:
This are the normal settings according to Seagate for a ST225 : 615 4 616 300 11 3
The first two I understand, the last two best leave alone but how about starting reduced write cylinder and write precomp cylinder, those are pretty unknown to me?
Click to expand...
Reference: User's guide for the WDXT-GEN

For the ST-225, the third figure, starting cylinder for reduced write current, is unimportant, because reduced write current is not applicable to the ST-225.

The fourth figure is the starting cylinder for write precompensation. I consider it unimportant in hobbyist use of MFM hard drives. A wrong setting is not going to affect the ability to format the drive. In one of Seagate's documents for the ST-225, write precompensation on cylinders 300 onwards is "recommended". In the user's guide for the WDXT-GEN, the example formatting string ("615 4 616 616 11 5") for the ST-225 has no write precompensation occurring.

The final figure represents the step rate, not the interleave. Table 4 in the user's guide translates the figure to step rate.
 
SomeGuy said:
Try specifying dynamic formatting, and specify only a few tracks. See if it finishes with or without an error, and if those tracks are readable.
Click to expand...
SunDown79 said:
Good idea!
Any idea on how to go about that ?
Click to expand...
For example, for a scope of cylinders 0 to 5, and head 0 only:

When the low-level formatting code in the WDXT-GEN's ROM asks you for the formatting string, enter 6 for the cylinder count, and 1 for the head count.

Then, to a write-then-read test the scope, you could use SpeedStor software per:
1. Navigation: [Diagnostics][WriteTest]
2. Answer 'Yes'.
3. Choose [ManualCyls]
4. Enter starting cylinder of 0
5. Enter ending cylinder of 5
6. Choose [ManualHeads]
7. Enter starting head of 0
8. Enter ending head of 0
 
What's not being said here is this: Don't be surprised if the drive ends up "pinin' for the fjords". Old MFM drives have a dismal track record of surviving 30+ years in usable shape.
 
Could be that the ST225 has died on me ofcourse (although tested it a few weeks back with success so would be bad luck)
But I will test the numbers out tonight :) thanks for all the info
 
Chuck(G) said:
Perhaps, but there are certainly a lot of 225 corpses out there... :)
Click to expand...

I have three that format up OK in my 3174 screen controller and then start to lose bits over the next few days. I think any ST225 thats been moved but not parked is liable to die. Now using Dave Gesswein's MFM Emulator and all is sweetness and light.
 
I've found low level formats will run very long and not produce an error when there's a bad cable. If in doubt about the cables, buy some new cables, it can save hours of wasted time.
 
JonnyGators said:
I've found low level formats will run very long and not produce an error when there's a bad cable. If in doubt about the cables, buy some new cables, it can save hours of wasted time.
Click to expand...

Was thinking about this as I was using a different cable then previously so changed it over, still no improvement.

I have now tested with the same cables and an F320 card and that works without any problems.
Could it be that the card is faulty ?
 
Ow and I am able to boot from the F320 formatted drive with the F300, so cables are certainly a-ok, as is also the drive.
 
Well, then it would seem that card is at least goofy. You never did say if it would format just a few tracks ok. I still suspect that a "dynamic" format might have behaved differently.
 
SomeGuy said:
Well, then it would seem that card is at least goofy. You never did say if it would format just a few tracks ok. I still suspect that a "dynamic" format might have behaved differently.
Click to expand...

Ah sorry, tested that, sounded the same, just shorter and then askes if I want to format the defective tracks and I answer N and it says "Nothing done"
So goofy indeed yes :)
 
Hmmmmm, and not to give up, I tried it again just now with default settings and it is...formatting alright...
Perhaps bad contact on the ISA bus or the pins for the cables, did check them out and they looked fine but that is all that it could have been I think right ?
 
You must log in or register to reply here.
Back
Top