Miniscribe EPROM Experiment

[CelGen]

I need someone with one of the 20mb Miniscribe 8425SA drives to help me verify something.

I'm running into a number of these drives which are cosmetically okay. They pass self-test and when put into exercise mode they function totally fine. The problem comes up when you try to operate it normally. Self-test completes, homes the head and then does a weird semi-step once a second which hangs up the SCSI bus and the host machine, making it impossible to read or format. Being a stepper motor design it doesn't rely on embedded servo tracks. Simply "step to x and hope the data is there" and with the drive confirmed to be mechanically fine the problem can only be coming from the logic board itself. The only field replaceable part is the 14 pin DIP EPROM There's a possibility that a few bits have flipped after something like 30 years. I need a dump of the EPROM on a known working drive to confirm this.

 

[techknight]

Well not only do you need to compare dumps, you have to compare firmware versions as well. only those of matching firmware versions will compare properly of course. 

I can dump mine, but I need to find a parallel port PC to do it. its that willem programmer which for some reason they STILL havent moved away from parallel ports!!

My two drives work, but one of them has random seek issues. 

 

[Floofies]

I can dump mine, but I need to find a parallel port PC to do it. its that willem programmer which for some reason they STILL havent moved away from parallel ports!!

Maybe a USB-Parallel adapter would work for you?

 

[techknight]

Those dont support true parallel port emulation. They support only a special "print" mode and doesnt have a full I/O address assigned to it which is required for proper bi-directional control. 

 

[CelGen]

I keep several spare laptops around specifically because they have hardware ports. Buried away in the shed is a rig I built for the dedicated task of running my Willem and Data I/O 29B.

Even if the firmware mismatches we should still be able to verify if the chip contents is the problem. Either the drive will go full ready and accept commands off the bus or do exactly what it's doing now.

 

[techknight]

Well mine low level formats and works, so I assume thats a plus? Before that it was all over the charts doing all kinds of strange shit. So I had to force a LLF, and that fixed it all. 

 

[trag]

I can dump mine, but I need to find a parallel port PC to do it. its that willem programmer which for some reason they STILL havent moved away from parallel ports!!

I've confirmed that the parallel port on the Dell Latitude 'D' and 'E' series Legacy Port Adapter docking stations work with my Needham chip programmer. They'd probably work with your Willem as well.

So you could pick up an old but serviceable Latitude to serve your chip programmer. That conserves space.

 

[techknight]

I have a p150 compaq LTE burried around here somewhere

 

[waynestewart]

I have an Apple Miniscribe 8425SA that an Apple shop had received but never even opened the box. I opened it to make sure it worked.

It uses a 27128 EPROM. I have an EPROM Duplicator which will make copies or I have a burner that only handles 2764 and smaller EPROMs. So I can either make you a duplicate  and test it or I can read the EPROM in 2 parts

 

[CelGen]

You're actually close enough I could come down and dump it with my own Wiillem. I could always drop by in two weeks time or so while doing other things in the area if you are interested.

Well mine low level formats and works, so I assume thats a plus? Before that it was all over the charts doing all kinds of strange shit. So I had to force a LLF, and that fixed it all.

That's a good sign both that the media is healthy and your EPROM is intact. If you can dump that we could possibly do a three-way comparison.

 

[waynestewart]

Sure, that sounds good

 

[CelGen]

Waynestewart was nice enough to let me come in and spend a few minutes pulling, dumping and comparing the EPROM from his good drive to compare with mine.

The results were fairly conclusive on the verify that even though the markings on both of our chips were different, their contents matched.....so the bit rot theory has been thrown out the window.

This goes back to the other possibility that there is a fault somewhere on my drive's logic board however we were not prepared to do a board swap and test this while at his house. Anyone willing to send me a board from a unformatable/crashed drive?