20MB HDD dead after RLL controller "blowup"

[rvdbijl]

Hi All,

I had a nice working 20MB MFM disk in one of my XTs, working off an 8-bit MFM controller (WD). I decided it would be fun to try to see if the drive would format to 30MB on an RLL controller (I remember doing that years ago, and as long as the drive is of sufficient quality, it works just fine). I bought a cheap (advertised as working) RLL controller off eBay (Everex 392). Plugged it in, and was greeted by a nice tantalum explosion on the board. Looks like it was one sitting on the 12V rail. I desoldered it, replaced it, and did not get any "hello" from the controller. Not a 1701, not any sign of the BIOS at c800, or anything else.
So I gave up and put the MFM controller back in. The disk was unreadable, so I ran a LL format with g=c800:5. That completed ok, and I could see the stepper move nicely from track to track. Reboot, FDISK - Error reading disk.. Yikes..
So I pulled out Speedstor and did a LL format through Speedstor, and on its verification it's marking every track & head as bad. I stopped it at track 30 as it wasn't finding any good ones...
I'm wondering if there is a chance in hell at repairing the drive. I'm wondering if some buffer IC on the data lines went bad due to a surge from the RLL board (as the stepper is doing its job, just no valid data is being read (or written?). Does anyone have any experience diagnosing / repairing this? I have scopes and other diagnostic tools, but am not even sure what to look for. The drive is a Microscience HH725. Also tried another MFM controller, just in case.. Same scenario.

I had been planning to replace that drive with a XT-IDE at some point, but I was looking forward to experimenting with it a bit more first. This was about the slowest drive I've ever worked with though! The seeks were atrocious.

 

[alecv]

Since HDD executes commends the Control cable is OK. Check a Data cable.
MFM ST-412 uses differential transcievers on the Data cable (similar to RS-422, at 10MHz). I.e AMD 26LS31/26LS32 .

 

[SomeGuy]

Right, check that the smaller data cable is plugged in the right way at both ends, and that it is attached to the correct header on the controller card. It seems a little unlikely that would have damaged it.

 

[rvdbijl]

Right, check that the smaller data cable is plugged in the right way at both ends, and that it is attached to the correct header on the controller card. It seems a little unlikely that would have damaged it.

I agree it seemed unlikely. But the same cable and same drive worked 15 minutes earlier on that MFM controller. In any case, somehow the cable must have developed a loose contact as I just replaced it with another cable, and it suddenly started working... VERY strange!
But thanks for the suggestions, even though I was expecting a much more complicated problem!

 

[rvdbijl]

On the same subject - (MFM drives on RLL controllers), can I assume the controller is bad if the controller gives me a 1701 error on boot with a functioning MFM drive connected? I have another RLL controller (that never smoked like the Everex!), but it just spits out a 1701 no matter what I have connected to it (jumpers are set correctly per the Stason.org site).

 

[evris80]

Not a 1701, not any sign of the BIOS at c800, or anything else.

According to the Everex manual found at minuszerodegrees (http://www.minuszerodegrees.net/manuals/Everex_Everdisk_Owners_Manual.pdf), on page 26 it says that for Microscience HH725 you have to properly select SW1 jumpers for “Option C”, as it is shown on page 24. On the same manual at page 41, it doesn’t describe a LLF procedure through DEBUG and g=C800:5, but refers to “Everdisk utility diskette” v.3 and above for this. There is a copy here https://vetusware.com/download/Everex%20Disk%20Software%203.02/?id=9595, maybe you should give it a try.

 

[modem7]

On the same subject - (MFM drives on RLL controllers), can I assume the controller is bad if the controller gives me a 1701 error on boot with a functioning MFM drive connected? I have another RLL controller (that never smoked like the Everex!), but it just spits out a 1701 no matter what I have connected to it (jumpers are set correctly per the Stason.org site).

The 1701 error is described at [here]. The possibilities there include, "A new hard disk has been added to the 'system', but the hard disk controller does not recognise the particular low level format on the platters of the hard drive."

That is why evris80 is suggesting that you perform a low level format (LLF); i.e. assuming that everything is configured properly and functional, the controller will 'lay down' a low level format that it recognises. Follow up with partitioning (e.g. FDISK) and a high level format (e.g. FORMAT).

 

[rvdbijl]

The 1701 error is described at [here]. The possibilities there include, "A new hard disk has been added to the 'system', but the hard disk controller does not recognise the particular low level format on the platters of the hard drive."

That is why evris80 is suggesting that you perform a low level format (LLF); i.e. assuming that everything is configured properly and functional, the controller will 'lay down' a low level format that it recognises. Follow up with partitioning (e.g. FDISK) and a high level format (e.g. FORMAT).

Gotta love old cables -- I tried this, but every time I got a 1701 error, no low level format tool would touch it. Not the Everex tools, nor the debug G=C800:5 (Error 80 or just simply "no disk installed". The disk works fine on the MFM controller that was originally installed in the PC (W01002S-WX2).
I have a "bucket" of old cables and since apparently the data cable was bad originally, I figured I'd try to replace the control cable as well.
Still no joy with the WD-1002-27X I have (still 1701 and error 80 / no disk). The everex, however, was now willing to LL format the drive (no 1701 on bootup), but complained about a bad track 0. Hmm.. Swapped the control cable again for a 3rd candidate and now was able to partition & format w. Speedstor. Sys'ed with DOS 6.0 and it boots! Running spinrite now to see how good/bad it is. It did clear Track 0 so far.... We'll see!
30MB on a 20MB disk, awesome!

 

[SomeGuy]

Swapped the control cable again for a 3rd candidate and now was able to partition & format w. Speedstor. Sys'ed with DOS 6.0 and it boots! Running spinrite now to see how good/bad it is. It did clear Track 0 so far.... We'll see!
30MB on a 20MB disk, awesome!

Sometimes this works, sometimes it doesn't. If it isn't working then you will probably see a pile of "bad" sectors starting about half way through.

 

[rvdbijl]

Sometimes this works, sometimes it doesn't. If it isn't working then you will probably see a pile of "bad" sectors starting about half way through.

So far so good. Spinrite checked the whole disk and found only one bad block towards the end. I still had some trouble booting from time to time, but changed out the data cable for yet another one (how do these cables go bad?!?), and that seems to have done it.
System seems stable now at 32MB RLL (612/4/26).

Thanks for everyone's help & suggestions -- I did not think this controller was salvageable. Time to build some new IDC cables I suppose! I still find it odd that some work on some controllers and some don't. I'm wondering if the sockets in the IDC connector are worn down where they sometimes don't make good contact with every pin?

 

[onre]

Captain Obvious chiming in; have you cleaned the contacts on the PCBs?

 

[SomeGuy]

I still find it odd that some work on some controllers and some don't. I'm wondering if the sockets in the IDC connector are worn down where they sometimes don't make good contact with every pin?

I have run in to bad cables often too. I think it is often where the cable runs in to the connector, or sometimes where a cable has been pinched. Possibly worn from small amounts of motion heat/cooling over the years. But usually it just seems like a WTF.

 

[rvdbijl]

Captain Obvious chiming in; have you cleaned the contacts on the PCBs?

Yup. First thing I did when it was showing signs of intermittently working and not working. I figured some contact cleaner spray wouldn’t hurt it. It didn’t help either.

 

[rvdbijl]

I have run in to bad cables often too. I think it is often where the cable runs in to the connector, or sometimes where a cable has been pinched. Possibly worn from small amounts of motion heat/cooling over the years. But usually it just seems like a WTF.

Yup. WTF alright. . It is funny how you forget things though. As I was going through this cable issue, I started to have some deja vu from when I built my own computers in the mfm/rll era. And the cable juggle seemed all too familiar. It was usually a bad crimp/press of the IDC into the connector, or a slight misalignment of the cable. What I used to do is put the connector in a vice and push it together just a bit more. That would usually work. But I guess I’ve been spoiled by SATA. It just works. Hahaha.