New Toys!

[SpidersWeb]

So the 386,286 and XT turned up!

I haven't had much play time, but small adventure with the XT:

- 301 and 41 error, press F1 - then F1 doesn't work
- 41 disappears but F1 doesn't work
- both errors disappear (after about 4 boots) and she fires in to life, DOS 4.1 and a menu

Unlike my 5150 dated 1984 from Aus, this 5160 is made in new york and the back is stamped '1982' (although I'm sure it's 83 at least).
My old IBM mono monitor works, but man horrible to use - it takes about 5-10 seconds for the previous image to fade when you change screen views haha.

I had a bit of a historic laugh with it. I remember reading yesterday that Windows 2 was so disliked, that rather than installing it it was common for apps to basically run it as a shell - and this machine has Ami and Paint doing exactly that - you run them from DOS, it loads Win 2.1 but no actual Windows, just the single app - when you quit, it exits Windows too.

Still have the 286 and 386 to look at. Problem with the 286 is that it is CGA/EGA and I fail to have a suitable monitor, so I'll need to switch in a VGA card (pretty sure I have at least one, fingers crossed).
386 has a massive full length ISA memory card which could be handy.

Edit: just looked it up, 301 41 means that key 41 was stuck down, old keyboard which I guess came right after a few presses

[SpidersWeb]

Ok, so the XT works great but my 5151 monitor didn't last long and now it wont display anything - tried 3 different video cards.
When you power off the machine sometimes you see a little green dot in the middle, so pretty sure its in the signal circuit rather than power.

Naturally I had to try the other two.

The supposedly dead 386 was flipping off the PSU on power up - clearly a short. After removal I found a shorted cap on a cheap serial/parallel board. It now starts up happily BUT the hard disc is completely dead - the heads had seized in place, so I manually turned the head wheel and freed them - but when it spins up and tries to read eck the scraping noise :/ This machine had so much rust... clearly the front had been exposed to the weather over time.
It came with a Trident VGA 9600 though which is 8 bit compatible - damn handy

So moved to the 286, it's a genuine Concord 286 XT - all parts seem original, case is in good condition too!
Downsides - when they say 286 XT they don't mean an extended 286, they mean an XT with a 286 CPU - has the same SW1 as the IBM PC/PC-XT, 8 bit slots, chipped RAM but it came with some really cool stuff:

- 8bit RTC card
- 8bit high density floppy controller with BIOS (although no idea how I'm going to find the switch settings)
- genuine, not a clone, Hercules full length video card

After changing SW1 5 & 6 to ON/ON and configuring about 6 jumpers on the VGA card, I was able to watch it boot up. Unfortunately it's ST-225, which sounds healthy, was unable to boot.
So I make a 1.2Mb DOS 3.3 boot disk and put the seagate low level format tool on it.
What does it do? The drive EATS IT - heads scraped part of the disk completely clean. It's 3AM now, and I have work, so will leave that for tomorrow evening possibly.

Also interesting, the IBM PC XT has a unit that sits under the CPU which has 3 chips and a battery - I wondered why it didn't ask me for the date/time!!

With the others I'm pondering what to do. I like the Concord 286XT being a complete unit, but if I borrowed it's RAM chips to replace some broken ones in my ISA memory card, and borrowed it's case, I could have a nice 386SX/20 with Windows for Workgroups, 4 meg of RAM, and VGA that I'll actually use. The 286XT really is not expandable, I can't do much with it besides preserve it. I guess I could restore both with more time :/

[SpidersWeb]

Made a decision.

- Move chips from bank 11 on RAM card to fill in bad spots, assuming I can find the manual for the switches (its a common Everex so should be easy) to give my 386 memory
- fit IDE controller and drive to 386
- network it up with Win for Workgroups
- learn to live with the upgraded but rusty "TURBO XT" case.
- fit 3.5" floppy
- question - for a 386 BIOS that does not have auto detect or any real intelligence - is there anyway I can use my 4.3Gb Seagate? even if I can only access 100-500Mb or something would be fine! I have a 400Mb drive coming but that was for the T5200

- aquire another monochrome monitor for the 286 and another AT keyboard, hopefully hard disc can be repaired (kind of worried it wasn't parked and it's not an auto-park model and it travelled 700Km yesterday)
- thief 1.2Mb drive from 386 (hopefully that doesn't eat disks too)

- fix 5151, and leave XT alone
- find another 5151 or a CGA setup and another IBM PC keyboard so I can have the 5160 and 5150 running at the same time when it's new motherboard arrives

Also have a T5200 that needs love next week. Busy times but loving it.

Oh and I have videos and photos recorded, just need to convert them and I'll append to this thread.

Edit: found doc for the memory card http://stason.org/TULARC/pc/memory-card ... V-135.html but I accidently ordered ten spare 256 chips off ebay 5 minutes ago anywho.

[Gibsaw]

Have you got piccies?

[SpidersWeb]

Coming soooooon.
I was up until 4AM in the end, and had work this morning, so will endevour to get them + videos up this weekend (possibly tonight).

[tezza]

Good to hear of your haul.

- genuine, not a clone, Hercules full length video card

Nice!

[SpidersWeb]

Cheers, I just finished another evening (4-5 hours straight playing with the thing) - this time on the 286XT.

Leaving out all the painful experimentation:

Booted in DOS with a replacement 1.2Mb drive.
Found the hard disc! Had tonnes of goodies on it, but lots of sector not founds and it gradually got worse - to the point I couldn't even do a DIR on C anymore. I did manage to start WordStar and almost started Ami in Windows286.

Anywho used the debug g=c800:5 method and viola, found a low level utility, formatted no worries - only one bad sector and that was marked on the drive.
Rebooted, fdisked, formatted system. But when it goes to boot I get 'System Parity Error'. I'd worked out it's because I'd got the switch settings for number of disk drives wrong, so changed that and the error goes away but it won't boot off the hard disc - lights up, starts for a bit then the system crashes.

After much playing around I moved to a DOS 5 disc, formatted system no problem, copied some utils on and reboot. Didn't start, and when I get to DOS by floppy, the drive is cleared! I was reading, running programs before the reboot!

Eventually I ran MSD, which confirmed it as a IBM PC/XT with a 80286 BUT also said it had 16Mb of XMS!! I pressed M for memory and I get the same 'Parity error on system board' I had earlier. Repeated process but avoided the memory menu and went to computer information - answer was right there "BIOS: DiskCache Performance-PC's (c) 1986" (had more than that but I can't remember).

Anywho some knob replaced the BIOS chips. I checked the chips and they're the aftermarket ones - explains the lack of memory, the drive write issues, and random memory reading bug/issue.
So if anyone has a spare Concord 286XT BIOS lying around let me know lol If I get a ROM burner, might try an IBM ROM, surely an aftermarket company wouldn't have made a custom BIOS specifically for the Concord XT286?

I think the only way to build this box will be to write the hard disc using the 386 mother board - but then what will it be like to use if it has write cache and no software to control it??
May need to swap out the motherboard sadly. On the bright side, I've got a meg of 41256 chips, an 8 bit RTC card, an 8 bit high density floppy card, WD MFM controller with BIOS and a working ST225 plus a relatively tidy case with good PSU.

Back to the XT, I opened the 5151 - carefully poked around with a multi to make sure everything was 0V, spent an hour going over it, couldn't see anything visually wrong. No burst caps, no leaking caps, no corroding diode. Because I couldn't see anything and don't have the skills/tools required to diagnose, I reassembled. After it's been on for a while and has an MGA signal you can see very faint patterns - but very very faint and doesn't look like text - on power off a bright green dot appears in the centre of the screen. The last thing it did before it's failure was a power off after accidently getting in to the wrong graphics mode.

Also my copy of XtreeGold (who doesn't remember this??) arrived today - even has the original adverts, a letter from xtree explaining the importance and their support of recycling, the registration cards, 5.25 and 3.5 floppies plus the manual.
Pretty stoked with that find

[tezza]

What does it do? The drive EATS IT - heads scraped part of the disk completely clean.

Yes, I've had this happen. Bound to be extremely dirty heads due to degraded disks being used at some time.

Too bad about the IBM mono screen. These can be hard to find.

[tezza]

Anywho some knob replaced the BIOS chips. I checked the chips and they're the aftermarket ones - explains the lack of memory, the drive write issues, and random memory reading bug/issue.

Are you sure about this? Many of these clones had third party BIOSes (eg Phoenix). Maybe the Performance-PC was similar and should be there?

[SpidersWeb]

Yep very good point, and I can't be sure but my reasoning was:

- The BIOS ID advertises Peromance PCs 3 times and no mention of 286XT or Concord (advertises DiskSpeed with an explination) - didn't seem terribly professional
- The RAM test only tests the memory for DOS not the cache - like a hack of an existing BIOS rather than a proper OEM solution
- These machines have all been modded/upgraded in different ways and came from the same source
- An OEM provider would've surely ratherered a 640/384 split rather than 512/512?

Also interesting on this machine is CTRL+ALT+DEL does not work, I have to hit a cheap reset button that's been mounted on the rear of the case.

But still the possibility of it being their from factory exists, or that it was fitted aftermarket on the purchase date becuase a lot of it seems dated 1986.
I'd love to leave it for the sake of being an original 1986 machine but it's making it quite unpredictable and frustrating to work with - after some more toying around I think it also protects modification to the partition table/boot record

(In brighter news, it'll be put to one side for another day, my IDE hard disc turned up and I'm gonna try and build the 386 independantly, as long as the HDD works I doubt I'll have any complications with that)

[SpidersWeb]

boo, ST3491 = 'hard drive controller error'
Put in a 120Gb and it was fine, just config error.

Might need to get some sent from the US :/
Also it has type 46 and 47 but you can't enter the values they're all 0's, so really want one that has a type so I can avoid using Anydrive etc.

8 bit MFM controller wont work in the 386, so I'm back to playing with the 286.
Charging up my iphone and will convert some videos soon, then going to do a video of low levelling with Norton Calibrate (part of Norton Utilties 6) which does some pretty awesome tests.
Small chance I'll get no where from this, but I'm out of options anyway.

[SpidersWeb]

Some pictures.
I haven't set up a proper work area yet, and am in the process of hopefully buying a home - in which I will setup a dedicated area. So for temporary diagnosis/playing they are on carpet but it is NOT where they will be staying.

IBM 5160 with 5151 before it stopped working:


IBM 5160 RTC modification:


(the rest of the machine is stock or typical, so not much else to show on that one)

Concord XT286



On the XT286 before the hard disc completely lost it's mind, running what I think is WordStar:


Using the low level formatted built in to the BIOS on the hard drive controller:


For anyone unfamiliar with this, if you have an MFM/RLL controller with a BIOS, there is often a low level formatting utility at C800, to access it you start up debug from a DOS boot disk, then at the "-" prompt type g=c800:5 and press enter. Some controllers do not have this, and some have it at other locations, but this is the normal place for it.

[SpidersWeb]

Re-low level formatting the ST-225 this time with Norton Calibrate - which is pretty awesome.
It tests the drives performance, controller, and memory, then does interleave testing - turns out for the ST-225 4:1 is the best, and provides a 77% speed increase over the 2:1 I had before.
It's half way through now and is finding nothing wrong.

The 16 bit MFM controller comes up 'hard drive controller error' in the 386.
The 8 bit in the 386 does the same, but in the 286 it seems to work fine, just the BIOS that is messing me up.

I'm thinking if I install DOS 5 after the low level, the BIOS wont prevent me writing to it. Last time it seemed to only choose to remember what I wrote immediately after the low level format.
Update to come.... (although 512K RAM isn't much to work with have to treat it as a normal XT)

[SpidersWeb]

ST-225 working great but:

286 "System board parity error" <-- only appears when I view memory in MSD or try to boot from hard disc, suspect MFM ROM
386 the controller from the 286 does nothing, possibly because the bus is running at 8Mhz
386 with the 16bit MFM reports no error, but says 'no boot sector on hard disc', booting from floppy works fine but none of my utilities can access the hard disk and c800 doesn't work on that controller.

Frustrating. Need a few other known good parts to test - MFM with ROM and preferably IDE for the 386.

I could test the MFM card in the IBM PC XT, but I'm determined not to play with / touch that.

Also my 1.2 appears to have stopped being able to read disks, so used a 1.44 on the 386 testing.

[SpidersWeb]

Soooooo found the MFM ROM on the minuszerodegrees site, opened it in a hex editor, and the parity error is not inside it.

My current theory:

- disk cache deactivates when you boot from floppy - so no error
- because it doesn't test RAM above 512K, I think there is a memory chip that has gone bad.
- I think that chip will be in the third bank, because my suspicion is that the DiskCache is probably 128Kb and made for XT machines.
- It happens after the hard disc starts accessing, not as soon as the ROM is called.

Really need a ROM reader/burner, have to get on to that soon.
Have ten 41256 chips already coming, so will anti-static the crap out of myself, and play piggy back when they arrive.
Those were for my 386's RAM card, but I'm likely to just order a decent 386 board with a proper AMI BIOS on it.