• Please review our updated Terms and Rules here

Tandy 4000 Boot Problem & ST-251 Problem

wesleyfurr

Veteran Member
Joined
Jul 16, 2009
Messages
612
Location
Virginia, USA
I debated where to post this, but figured this would be a better section for this machine than the Tandy section.

I have a Tandy 4000 (the big 386-16) that I picked up on ebay a number of years ago. It was listed as good, but wouldn't post. Only thing I could find wrong with it at the time was a browned capacitor. I still wanted it for my collection, so the seller gave me a partial refund and I kept it. Time went by and I learned about dead Tantalum capacitors keeping things from working. I snipped one of the leads of the bad looking capacitor and it fired up and beeped!

Finally tonight got around to replacing the bad capacitor and then hooked it up to a keyboard and monitor for further testing. First power up it did not post. Power cycled it and then it came up. At some point I power cycled it again and it took a few cycles to get it to come back to life. Looks like something may be goofy...any thoughts what might cause it to sometimes not POST sometimes?

I removed the CMOS battery a while back to make sure it didn't leak and ruin the machine as happened to my Tandy 3000 some years back. :-( So...of course it is lacking a good config. Made a boot disk and copied the setup program to it...boot from floppy and configured everything. I set it to drive type 39, which matches the 6 heads/820 cylinders/17 sectors of the ST-251 that is installed. It fires up, complains about "invalid configuration information - please run setup program", then says "hard disk failure" press F1 to retry boot. It is counting up 640k base and 1024k extended...not sure what size chips are installed, but they are 8x 30-pin SIMMS...guessing 256k each, in which case that would be the correct default. Any idea on a Tandy 4000 if it will read 1024k when there is really more installed? And what video type should I be using for a VGA card - 80 column color, or EGA? Just wondering what it is complaining about...

I'm ashamed to have to even ask this question, but it's been too long...if the ST-251 drive is dead, would that cause the "hard disk failure" message? I'm assuming non-matching disk type setting would not cause that message? How reliable are those old Seagate MFM drives these days? I was under the impression that they held up pretty well...but I may have gotten the wrong impression! The spin-up sounds seem a little off, but it's been years since I fired one up...

Any thoughts appreciated...

Thanks,

Wesley
 
There are certainly not a lot of 4000's around...I count myself lucky to have snagged the one I have at a low price. Not sure I've seen more than one other one, though that one has shown up many times at a very high price, and as far as I know has never sold. I'm guessing it was pretty pricey when new as well. If you've never had the motherboard out of it, on mine at least, there are several wires manually run to presumably fix things, as well as two resistor packs and a capacitor. Hard to imagine seeing that in a production machine...would have taken a lot of work to do that to every motherboard! I took a picture of it, I'll have to post it... Definitely an early 386. It's 16MHz and has the double-sigma symbol on it. Also has the BIOS setup you have to run from disk as well as limited drive types. It prompts for a type of 1-47, but when you hit ? for a list, I think it lists through 42 or something. When you tell it 47, it acts like that's not a valid option.

What exactly do the special EPROM's do? I would have thought SCSI would have had its own BIOS and wouldn't have mattered what the motherboard BIOS was capable of doing. The one I have is of course a Seagate ST-251 MFM drive...pretty sure it is a WD100x card...would have to pull it again to see for sure...it seems to be shorter than some of the really long full-length WD cards.

I could probably ID the SIMMS as well...just didn't bother yet. :) After I shut down my desktop last night and went to clean up a bit, I tried again and set the BIOS for EGA and the BIOS error went away, so it appears that is what it wasn't happy about. When it reboots, the hard drive grinds a moment, and then throws up the "hard disk failure" message...so I'm guessing the drive is just toast.

I'm thinking of just putting in an IDE to compactflash type setup, or just using an XT-IDE controller. Problem with 16-bit IDE with CF adapter is the BIOS limitation with regards to drive types. Does anyone know how the performance compares between an XT-IDE controller and a 16-bit IDE controller with the BIOS somehow put elsewhere?

Yes, I do know about the wonderful Tandy documentation and info. Pretty sure I've saved off a local copy of most of it...certainly an invaluable resource!

Thanks,

Wesley
 
You are not getting a 1701 error from the hard drive, so the controller is seeing the drive come ready. Was the drive setup originally with the controller? I would try a low level format and see if the drive doesn't come around.
 
Not seeing a 1701...but then again, I don't know that a Tandy would give errors like a true IBM. I'm not sure if it's seeing the drive come ready or seeing that something is wrong after it tries a while...? It makes a chattering sound, then when the sound stops is when it shows the "hard disk failure" message.

I'll have to go digging for the proper low level format utility for a Tandy...I know I had one ages ago for the 3000...not sure if that one would work or not. Unless there is a generic utility out there that would do the job? There's no BIOS on the board, so no g=c800:5 utility...if I'm remembering that command correctly...

Thanks,

Wesley
 
The one I usually go for is SpeedStor. modem7 has a page with links and a few details up here: http://www.minuszerodegrees.net/software/speedstor.htm

ST-251's make some interesting noises, and it's all pretty normal, I wouldn't worry unless SpeedStor fails to work with the drive.

Thanks for the tip, I will definitely give that a shot. Things got busy again and it got set aside for the time being.

For those interested, here's a photo of the motherboard and its underside...note the manually-installed tweaks...can you imagine adding wires and parts to every machine built and sold?

P1020862.jpgP1020861.jpg

Wesley
 
I removed the CMOS battery a while back to make sure it didn't leak and ruin the machine as happened to my Tandy 3000 some years back. :-( So...of course it is lacking a good config. [...] It fires up, complains about "invalid configuration information - please run setup program", then says "hard disk failure" press F1 to retry boot.
Did you actually install a new CMOS battery? Unless you did, you shouldn't expect to get the system to do anything useful. Config is lost as soon as a reset happens, because the CMOS is not powered by anything else than the battery. It will just complain about invalid CMOS data over and over.
 
Did you actually install a new CMOS battery? Unless you did, you shouldn't expect to get the system to do anything useful. Config is lost as soon as a reset happens, because the CMOS is not powered by anything else than the battery. It will just complain about invalid CMOS data over and over.

Resets are fine...it's perfectly happy until it has been turned off for a short time. Even a fairly brief power cycle still holds the info. Of course after a slightly longer period of being powered off, you then have to re-run setup to get things to work.

Wesley
 
Thanks for the tip, I will definitely give that a shot. Things got busy again and it got set aside for the time being.

For those interested, here's a photo of the motherboard and its underside...note the manually-installed tweaks...can you imagine adding wires and parts to every machine built and sold?

View attachment 26581View attachment 26582

Wesley
I wouldn't use any type of low-level format utility if I were you, unless it was the absolute last resort. A low-level format basically destroys everything that is stored on the hard disk, so unless you don't have the funds for it, you might as well just install a new hard drive.

A question though - Did you try installing the hard disk drive in another machine? Also, did you try connecting the hard disk controller to a hard disk drive that you know is working?
 
I wouldn't use any type of low-level format utility if I were you, unless it was the absolute last resort. A low-level format basically destroys everything that is stored on the hard disk, so unless you don't have the funds for it, you might as well just install a new hard drive.
Why would there be anything on this drive he'd want to save?

Where would he find a new MFM drive?

A question though - Did you try installing the hard disk drive in another machine? Also, did you try connecting the hard disk controller to a hard disk drive that you know is working?
It seems you don't understand the relationship between MFM controllers and drives. In most instances an MFM controller will not be able to read the drive unless the drive has been LLFed with that controller. The only reliable way to move an MFM drive from one machine to another is to move it's controller with it.
 
Stone said:
Why would there be anything on this drive he'd want to save?

Where would he find a new MFM drive?
Well, there are places that still sell old computer parts, if you know where to look. And the drive likely has old software and other data stored on it that could be useful.
Stone said:
It seems you don't understand the relationship between MFM controllers and drives. In most instances an MFM controller will not be able to read the drive unless the drive has been LLFed with that controller. The only reliable way to move an MFM drive from one machine to another is to move it's controller with it.
But wouldn't it still work if the drive was set up with another controller that was the same or similar model as the first one?
 
Since the poster was already discussing formatting, it's a popular branded machine, and the large array of software available online - I felt that was a suitable approach. Using SpeedStor can give quick answers, no reliance on CMOS (he can just select Seagate ST-251 and go), meaningful errors, lots of tests - and likely chance of solving the issue.

If data recovery was important then there are a few other things that can be checked like inspecting/swapping out cables and double checking the drive selects. On one occasion I did get caught out by a single pin in a connector that had bent up. I'd also suggest moving the controller+drive+cables in to another PC, and finding the type of card - if the card is 8 bit MFM or if the controller is RLL with it's own Boot ROM then the setup may be different. To boot an RLL drive with it's custom BIOS on my IPC AT for example, I need to set the Drive Type to "1" - using "none" or any other setting and it'll fail.

For reference this is the normal startup and spindown sequence: https://www.youtube.com/watch?v=vTu-1dLegzI (not my video)

If a computer is unable to read the first track, it can also be quite common for it to seek back and forward on the drive as well for quite a while before returning an error - generally if I hear this I check my 20 pin data cable, and if that's fine, it's almost always time for a low level format. Hopefully that's a bit more informative.
 
It would be nice to get it fired up and see what (if anything) is on the drive, but it's not a huge deal either way. Probably just as well to try a fresh LL format and see what happens. The link to the video is great...I need to listen to it again, but I don't think it sounds that nice. Too many things going on...it has been relegated to the back burner again... :-(

I've got a few other MFM drives laying around, no idea if any are good. I would probably go for an XT-IDE card otherwise...much easier to deal with...unless someone thinks an MFM drive on a 16-bit bus would be faster...? I've not looked at benchmarks or tried one for comparison...

Wesley
 
No need for the XT-IDE card in this system. Just use a standard 16-bit multi I/O controller. If you need to use a large capacity drive you can either use a software drive overlay or get the AT version of the XT-IDE BIOS and put it into an old NIC card (configured for the correct address obviously)
 
No need for the XT-IDE card in this system. Just use a standard 16-bit multi I/O controller. If you need to use a large capacity drive you can either use a software drive overlay or get the AT version of the XT-IDE BIOS and put it into an old NIC card (configured for the correct address obviously)

I guess that is certainly an option. It would have to have an add-on BIOS though...the current one doesn't even support type 47. What do you have to do to create the BIOS chip that will work on a NIC card? And how would speed compare? Might be about as easy (for me anyway) to build an XT-IDE card...
 
Back
Top