• Please review our updated Terms and Rules here

Memory Size Error and Non system disk on 5170

I'm wondering if maybe the battery is dying and loosing the settings for the memory and drive. Possible?
1. If the battery voltage goes below an acceptable value (for more than a few minutes), CMOS/SETUP contents are lost, and the 5170 will generate a 161 error on power up.
2. If CMOS/SETUP contents unexpectantly change, the 5170 will generate a 162 error on power up.
 
Ok, so you guys were right about the system files. We booted on a floppy and there was the C:. We copied the system files over and were able to boot from the HDD directly. There's a new problem, though. As we were using the computer over an hour we kept hitting unreadable sectors until the drive seems to have failed all together. It doesn't seem like the drive is accessible at all now. When I boot the machine I get a 1780 Error, Disk O failure. The HDD light flashes once and then nothing, not even seek noise. On entering MSDOS installer it says the HDD is not writable. Our best guess is the drive heated up and misaligned ever so slightly. I don't know what to try next.

We did figure out the memory problem. It was just set up as having more than it actually did. Counting chips gave us the exact number the computer was looking up before the error. Punching that into the setup cured it.
 
I don't know what to try next.

You do what has been suggested. :)
Chuck(G): If you're curious as to what's on the hard disk, you might try booting from floppy and see if you see anything. Otherwise, low-level format the HD and install a new system.

A low level format is standard maintenance to be done at regular intervals and there's not much point in fooling around unless you want to see what's left on the drive or experience what a drive in need of a LLF acts like.
 
Depending on the hard disk, you might need to do a low level format of the HD. That would create a new set of tracks which would be aligned relative to where the drive heads now reside.

A low level format will not help if the drive suffered a several head crash in the past nor will it help if the controller card is going bad.
 
Here I am, head in hand, waiting for the AT Advanced Diag Disk to finish the system check. I think it's been 30 minutes already.

10:40 AM. An hour and a half. I know I'm supposed to wait awhile for this to finish, but is it stuck at this point?
 
Last edited:
You've very probably got issues.

Strip the system back to minimum (floppy and display--if you have a hard disk, disconnect it) and run the diagnostics. If they pass, start adding cards back.

How far does the system check get?
 
I press "0" to enter the system checkout and that's as far as I get. While it's sitting there the HDD thumps a few times and the floppy is spinning but not seeking.
 
I wonder if your floppy is reading correctly. Have you verified it on the system used to create it? Maybe we should start there. Did you configure the 5.25" on the system that you used to create the floppy as a 1.2M 5.25" drive in the BIOS setup?
 
Last edited:
I don't believe he verified it. We used the same disk to run the setup option and put in the correct memory size. Is there any other way to LLF without the diag disk? I know debug is out, but is BASIC an option?
 
You have the issue of getting enough stuff into the machine to load and run diagnostics and utilities It should be possible with BASIC, particularly if you've got a serial port on your PC--and another to match it on a "server" system connected with a "null modem" cable. But I'm not aware of anyone who's tried doing it recently.

A 3.5" drive might be your best bet at this point. Adapters for the power plug, as far as I know can still be had--and you can use a more modern floppy cable to hook it up, so you don't need to hassle with the edge connectors. Then you can create the boot disks on a machine that handles 3.5" media.
 
When my 286-XT arrived it booted fine too, but it was really just sucking the very last of the magentism off the disk, 3 boots later and it couldn't see anything - did a low level format and viola as good as new.

For getting data on to a machine, null-modem cable for the win. In fact I used it yesterday to load up my desktop 386 with goodies (needed WFW311 to get on the network ;) ). I got my cable delivered next-day for around $15NZD (10US?), so they're still out there, and easily made if you're so inclined.

I use FastLynx 2 (FX2). It has an upload mode which can copy itself to a host thanks to the DOS command 'ctty'. So you don't even need to copy it on to a disk first. It's an easy to use uttility with built in help etc. It usually works even if your DOS folder is missing, because it uses the same defaults for uploading as a standard DOS boot (2400,n,8,1)
Usually "ctty com1:" or "ctty com2:" is all you'd need on the client, then select Upload from FX2's menu on the host. (and I beleive ctty is part of command.com).

Once it finishes I find I have to reboot, then load it up, and you can use either machine to transfer/delete files.

FX2 and a lot of the DOS utils for drive formatting etc easily fit on floppies.

Edit: the above assumes you have at least one other machine with a serial port. If not, you may need to get yourself a hack machine to bridge the gap. Parallel is an option but can't remember if ctty works with it, FX2 does.
Edit2: if you fit that AMI BIOS you mentioned, I'm fairly certain you can Low Level Format with it
Edit3: I'm also assuming you have a spare disk with space on it - FX or FX2 are quite small but not tiny, in the 150Kb+ range from memory.
 
Last edited:
Well it needs a DOS boot disk to use the ctty command (which he has, unless the floppy broke? his earlier posts said he did and he browsed the HD)
Also needs to be enough disk space on the floppy to hold the FX2 files after the copy. Either the boot disk or another disk.

Can do a youtube clip of the procedure if needed.

For a machine with no disks at all, I guess you could pet it and admire it's good looks?
 
Last edited:
Ok. I took every card out of the machine except for the ega card and the disk controller. After that I was finally able to get through the system check menu to the format menu. At that point I followed the instructions for a conditional format at which point I got a 1706 Write Error and was told to do an Unconditional Format. I followed through with that saying "no" to any defects because I didn't know of any, obviously. Again, I got a couple of 1706 Write Errors.

When I tried using Speenstor it threw a Parity Check 2??? at me and froze the system to the point of needing a hard boot. I ran the app from the disk fine in DosBox. So, is the drive dead or what?

Update 1: I just ran a surface analysis. It gave me a 1706 again but then started at cylindar 731 and started working down. I'll update again when it's done.

Update 2: The scan stalled out at cylinder 265. I shut the system down for the night.
 
Last edited:
I think I've pinned the "Parity Check" error and the system analysis hang to the 370 emulation boards. I put the expanded memory back in and all was still well. There maybe something odd in the emulator that trips up system utilities.

I ran the surface scan again from the adv diag disk and it went through the same motions, but this time produced another 1760 write error at the end. I tried running it again but it wouldn't even start. I then ran SpeedStor again and this time it loaded (without the 370 boards). Speedstor couldn't do anything with the disk. Everything I ran it came back with a write error. I even tried parking the heads. That got a seek error. I did a test of the controller board and that came back fine.

So, what's next? Is it a doorstop? I'm a little disheartened at this point. My original intention was really to swap the bios and controller out and put in an IDE drive and a 1.44, but as I worked on this with you guys - and seeing that even the drive cables were still taped together - I began to want to preserve this machine in it's original configuration. I especially feel bad about losing the data on the drive. It seems like it was a school computer. There was assembly language files from different authors, some of which I had the names as they were embedded in the code. Some were accounting softwares, one was a battle ship game. I think they may have been writing it for upload to a school mainframe? Maybe a university? Who knows. Anyway, it's all gone. But I'm glad I at least saw it.

If there's anything else I can do short of $1000 worth of physical repairs to the drive I'm willing. I know the data is gone, but the machine could live on in it's original state at the very least.
 
Back
Top