A chance comment I made to an article over at Hacker News led to me being offered a non-working Acorn System 1 for spares or repair for the cost of a donation to charity. Well, it arrived yesterday and it's in pretty good shape. The main board has some logic chips missing and the memory addressing has been hardwired rather than using the address configuration header socket, which has not been fitted. The spectra strip for the display has been repaired in the past and a second wire is broken, but that's an easy fix.



Getting the system going should be fairly straightforward, but the big question is: Do I get it going 'as is', or repair the cut tracks and fit the address select header as per the design spec. I'm tempted to do the latter so that I can use the system as a reference for my planned build of a replica because this will allow me to test the firmware in an EPROM as well as in PROM.

Any thoughts?

Some more pics...

Congrats!

It looks very good and an easy repair hopefully to original state!

If it were me, I'd attempt to restore it to the original circuitry.

Do you know which chips are missing?

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Love the hand laid out tracks. Nice thing to have. I wouldn't be too worried about it being a museum piece - getting it working seems good to me.

Yes, I have all the schematics and the tech/user manuals - it's a 74LS04 and an LS00 and I have some of those.

As well as the overall restore, I am debating whether to repair the spectrastrip to the display (quick and easy) or replace it (minor hassle, but where to get a small bit?).

As would I, if it can be done.

-Tor

That's great... Congrats'...

Now that you've got the "real thing", are you still planning to build a replica?

Sure will - having the real thing is a great bonus. I just need to find some time to make a start and sort out how to do the keyboard.

If you are okay with not having actual spectrastrip just take an old 40 pin IDE cable and make a "donor" out of it.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

The fitted spectrastrip has single core (read: stiff) wire rather than stranded, plus the pitch is different. I'll probably cut down the two broken wires about 1cm and rejoin them to the display using component leads or tinned copper wire covered with clear heatshrink.

We have that dilemma with vintage cars all the time. In the past people restored to as new (actually usually to BETTER than new) but these days things are swinging back and the trend is to preserve the vehicles past history as much as possible. It gets worse with old race cars which used to change all the time, get pranged, rebuilt, developed and so on.

How that relates to computers, I don't know As long as it's not some special prototype if all you are doing is undoing some bodgy work someone else did in the past then I say fix it back to original spec. Especially if the mods were to work around some original problem you can now fix properly.

Simon

_________________
My 6502 related blog: http://www.asciimation.co.nz/bb/category/6502-computer

SHE LIVES!!!

• Previous rework removed and the cut tracks repaired - pic below.
• Three broken wires on the display Spectrastrip repaired - one was a hairline break that I only found out when I powered up for the first time and noticed a display digit wasn't working.
• Turned-pin IC socket fitted in address select position.
• Another turned-pin socket fitted into the first, and address jumper wires soldered to match the original docs. This allows me to rewire the address select later without messing up the base socket.
• Touched up a few dry-looking joints.
• IC10 refitted = 74LS04
• IC12 refitted = 74LS00
• All existing socketed ICs lifted and reseated to clear any corroded or loose connections.

Having done all the above, I slowly wound up my bench PSU to 7V (allowing 2V drop-out for the LM340-T5 regulator) while monitoring the output from the on-board 5V regulator and the current draw from the PSU - all looked good so I hit the reset button and - YAY LIFE!!!

As could be expected, the keyboard keys need a little exercise to get them working reliably, but I entered a sample program and hit (r)un and saw the expected result. Stepping through the monitor ROM memory shows the expected hex values so that's a good day's work.

I've run out of IPA, so the next steps when some more arrives will be to deflux the boards and apply a small amount of green solder resist over the rework. I might source some more Spectrastrip at some time, but the repairs to the display strip aren't looking too bad.

Next steps:

• Try out tape saving and loading using a PC-based tape recorder emulator (don't suppose anyone's already got some Acorn System 1 progs in MIDI format???).
• Blow the monitor program into an EPROM, re-jumper the board and see if it will run - this EPROM can the be used on the copy of the system I am building until I blow the authentic bipolar PROMs.
• Test the second INS8154 IORAM device - maybe build an AY-3-8912 sound board like we had at school when I first used the System 1!!
• Actually start making my copy!!

Edit: I've now used the System 1 to test all the main chips bought online as 'new old stock' for the replica build (a few of each: 6502, INS8154, 2114) - all working fine, so if the replica doesn't power up first time it's probably my fault!

Before I sign off for now, I must put on record my thanks to Tony Edgecombe for 'selling' me his System 1 for the cost of a charity donation.