Mechanincally the machine is identical to the Model 3. The difference is in the CPU board and the keyboard. I think some later M4s have everything (including the disk controller and RS232 port) on the CPU board, the older ones like mine have the same disk controller and RS232 boards as a Model 3, the CPU board is the same size (and same connectors [1]) as a M3 one, but electtronically very different. [1] OK, there's one extra 4 pin header on a M4 CPU board. It links to the sound board stuck to the front of the screening metalwork. As regards dismantling it. Firstly, if you've never doen one before, take care. The CRT and monitor PCB are fixed ot the top case, and the CRT neck is vert close the screening metalwork in the lower case. It's very easy to reak the CRT when reassembling the machine, it's not much harder to do so when taking it apart!. The trick is to lift the case up vertically, in fact pull it forawrds slightly. Then when it's clear, turn it 90 degress anticlockwise (seen from the front) and put it down to the left side of the base assemnbly. Then unplug the cabling. Secondly, there are 2 SMPSU boards in there with mains all over the place. Of course you unplug the mains cable before taking it apart, but it's possile for stored charge to remain. Other things. If you have the type with the separate disk controller and RS232 boards, there are tapewires that link them to connectors on the CPU oard. These are very fragile, and seem to be the problem in 99% of cases when a M3 or M$ won't boot. OK, I exagerate a bit, but handle them with care. The keyaord is a simple matrix of switches (similar to the M3 keyoard in your manual (I assume) but with extra keys in the unused spaces of the M3 matrix. The switches can, and do, fail. What I did to my M4 was pull all the keycaps (make a diagram of where they go back), then desolder all the swtiches from the PCB (!). Take the whole lot apart, and then test the switches with an Ohmmerter. Don't expect a dead short (they're contuctive ruber contaics inside), but you can seprata the good from bad ones this way. The swtiches do come apart. Clean the conductive rubber part and the contacts and try aggain. This might get some of them going. The remaining duds come apart again, and I rubed a 6B (very soft) pencil on the conductive rubber part. That really reduced the resistance!. I then worked out the least-used keyoard locations (in my case the number pad) and put the dodgy/repaired switches there. I had enough good ones to fully populate the main area with a few dubious ones in the number pad. I didn't find that necessary. There are not that many small parts. The danger from CRTs is, IMHO, much overrated. The CRT in the M4 is a 12" monochrome CRT with a tension band. It's very unlikely to violently implode (due to said tension band), it's also not that high an EHT (12kV or so). I would remove the CRT/monitor PCB from the top case to clean things properly. The monotor PCB is held on by 2 scres near the bottom edge. The CRT is held in by 4 corner nuts. Either remove them separately (if you don't mind unplgging the EHT anode cap, etc), or together. The CRT will certianly need a clean!. One of those CRT nuts is _hell_ to get to, I used a very long nutdriver to get in to that corner. Oh yes, put the top case on the edge of the bench, resting on the screen face (with the keyoard surround overhanging the edge of the bench). And NEVER lift a CRT by its neck! -tony

This sounds as thought it might be slightly differnt to my keyboard. I only rememebr 2 contacts per key. On the other hand, when I ordered a replacement keyswitch for my Model 3 (about 20 years ago), I did have to specify the brand of the keyboard PCB (Radio shack apparently bought these in). The keyboards in oth my M3 and M4 are ALPS brand, FWIW. The reason I mention this is that dismantling the switches themselves will vary between brands. The good news is that (at least for the ALPS keyoard), it's a single-sided PCB, bot plated through. So it desolders quite easily, The bad news is that it's a cheap SRBP oard, and the traces lift all too easily. I've completely sesoldered several keyboards for repair. It takes me an afternoon to get them apart without damage (and a lot less time to get it back together again). I guess if you need soldering practice, though, this is not a bad project ot undertake. There arer no ICs on the keyoard PCB, so there's no risk of static damage, and the parts are not particularly thermalliy sensitive. I don;t know what drives you have (Tandon TM100s were fairly standard), but if you take them significantly apart, you'll need an alignment disk and 'scope when you come to put them back together. Again, something that doesn't worry me _at all_ now, but it might be a problem for you. If you want to take the drives apart to clean them, let me (us?) know the model number (or any markings on them). I can then give you an idea what what you can take off without losing the alignment. -tony

It is. There are a couple of locking tabs that you sqeeze inwards to free the switch. In fact that's how you get the switches out if you desolder the lot and take the PCB off. My view is that if you have about half the switches defective (as you seem to), then you probably should take the whole thing apart, test the switches, etc. You can test them in place (an ohmmeter between the pins, then press the key), there anre no other components to confuse the resistance reading. As I mentioned, it is possible to take the swtiches apart and clean the contacts and/or coat them in a bit of pencil graphite. But how long it'll last I don't know. I did it to mine, they're still working _but_ I put them in places where I'd not really use them (I rarely use numeric keypads, so I put the bodged switches there). You;'ve got far too many duds to do that, though. Not recently, nbut I have done it. The main problem with doing it on the M4 is that IIRC the scan lines are the CPU address bus, suitably buffered. Which means a lot of HF noise flying around (the scan lines are thus toggling all the time, even when the keyboard is not being read). THere has been talk on the list of using a microcontroller to interface a PC-type keyboard to a machine that expects a matrix of switches. Whether a microcotroller could keep up with the Z80 address lines and correctly output the keyboard data or whether you;d need a separate crosspoint switch chip (which is what I'd probably use anyway) I don't know. Which, IIRC, are a very close clone of the TM100. To the extent that most, if not all, parts interchange. If these are TM100-a-likes, then the postiioner mechanism is on a little subschassis which you move to do the radial alignment. I've had great success in scribng a line to indicage the position of that subchassis on the main chassis, then carefully turning hte adjusting cam screw _without loosening the positioner fixing screws_ and scribing the limits of its travel on the positioner unit. When I put the positioner back in, I line up the marks, get the cam screw to have the same limits of travel, and clamp things up. I then put an alignment disk in (:-)). Every time I've done this I've been well within tolerance on the alignment disk. On the other hand, I'd probably not risk it if I didn't have the alignment stuff to hand. Now, the index timing is pretty unimoortant, so it's OK to take off the upper clamp arm. Of course you can remove the spindle motor, belt, spindle, beaerings, etc. If you want to remove the track 0 switch, again mark the position first. I've never had any problems getting that back right. Those swtiches do fail sometimes, amazingly an identical part (it's a low operating force microswitch, about 15g IIRC), is available still. Perhaps I'm odd, but I like to keep machines as original as possible. Which includes keeping the origianl drives, PSUs, etc. -tony

A freind of mine (and he's still a friend...) once said "When you first meet Tony, you might think he's mildly eccentric. When you get to know him, you'll realise that there's nothing mild about his eccentricities" -tony

Tony Duell wrote: No, not odd at all. (Not for that at least.) I'd like to keep the machine original too, but it also has to be usable. If I can't use it, there is no joy ... I'm thinking ahead a little bit, but I'm going to be in the market for an assembler and a C cross compiler. I need to write a TCP/IP stack for this one now. ;-0 (My DOS based TCP/IP is almost done .. I'm slogging through a DNS resolver now.) Mike

I certianly agree with that. I don't collect ornaments. I collect computing machines (and related stuff). The purpose of those is to run programs, not to lokk pretty on the shelf. But I don't feel that is mutually exclusive with 'keep them as original as possible'. The 'as possible' bit covers replacing faulty parts so that the machines can run again. But I replace the smallest part possible. Obviously I don't have an IC fab line, so if a chip fails I have to replace it. Maybe not with the same type but with a compatible one, or worse with one that needs minor mofifications to the rest of the machine for it all to work. But I replace the chip. I don't replace the entire PCB. Similarly, I cna't repair a disk drive head. But I can replace the head carriage, and keep the rest of the original drive in the machine. Or, as I did recently, replace a track 0 sensor microswitch with an almost indentical part (the only difference was the manufacturer, from what I could see). Sometimes I do try to repair the original part, if there's no suitable new replacement. I've been known to rewind motors, for example. Of course what is 'possible' depends on who you are. But I try to take the view that if it's something I've not repaired before then it's a way of learning how to do something new, not a problem that I'll never manage to solve. -tony

On Fri, 26 Dec 2008, Michael B. Brutman wrote: The PT210 uses identical keyswitches. I rehabbed those by painting the undersides of the rubber domes with conductive ink -- the same stuff sold for repairing PCB traces. -- David Griffith dgriffi at cs.csubak.edu A: Because it fouls the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail?

Here's a word you don't hear often - "Aquadag". That ought to do it. GC also has some metallic ink emulsions. Cheers, Chuck

---------------------------------------- Hi It might be that one could restore it by removing a little of the surface. One could try a small piece of 220 or finer sandpaper and just remove a little of the surface that has been oxidized. Dwight _________________________________________________________________ Send e-mail anywhere. No map, no compass. http://windowslive.com/oneline/hotmail?ocid=TXT_TAGLM_WL_hotmail_acq_anywhe…

---------------------------------------- Hi It might work but I think the problem is rubber oxide not a metal oxide. After getting a new surface, Dexoit might be good to keep it from oxidizing again. Dwight _________________________________________________________________ Life on your PC is safer, easier, and more enjoyable with Windows Vista?. http://clk.atdmt.com/MRT/go/127032870/direct/01/

--- On Sat, 12/27/08, Michael B. Brutman <mbbrutman-cctalk at brutman.com> wrote: Another thing to look for would be that conductive paint that is used to fix rear window defrosters. You can get it at the auto parts store. I've used that to repair those mylar flex-circuits before, when I couldn't get the PCB trace paint. -Ian

Michael B. Brutman wrote: Mike, Has anyone mentioned using a pink eraser to clean the surfaces? I have used that to restore keys on a PET that had become unresponsive. The problem keys had a "shine" on the black conductive rubber matching the shape of PCB traces they were hitting. The pink eraser removed this shine and the keyboard worked like new afterwards. Just make sure to get rid of any eraser leavings! Cheers, Bryan

On Dec 27, 2008, at 3:28 PM, Tony Duell wrote: I only use it when I'm really mad at someone, and NEVER in front of children. *snicker* But seriously, it's one I've not heard in a while. Right along with "corona dope". -Dave -- Dave McGuire Port Charlotte, FL

Tony Duell wrote: Tony, We expect you to be around for a while because rumor has it that you can be fixed with parts commonly found in a kitchen drawer. Desoldering the single switch was fairly easy. I just wasn't looking forward to doing 70 of them. My particular key switches have four contact pins. The only inscription on the switches is 'SMK'. The trick was figuring out how to pull the switch through the metal mounting plate without having ready access to the sides of the switch. Bingo - somebody else informed me about these kits over at Erik Klein's web forum. Except for the mixing and expiring part, the kit sounds like the perfect fix. I paid a visit to Rat Shack (Radio Shack for you non-US citizens) and did not find the conductive ink pen. The store manager is pretty good about his parts, and he knew it was discontinued. I told him I was restoring a Model 4 and his face just lit up. :-) I'm going to experiment with the conductive ink pen first because it looks easy to apply. I only need to fix a few keys on this keyboard. If that fix doesn't hold up then I will go for the rubber keypad repair kit. Mike

On Saturday 15 March 2008 16:28, Michael B. Brutman wrote: It's been quite some time since I got into one of those, but what comes to mind is sharp edges. The thing will want to draw blood if you're not careful. :-) The thing there is to avoid putting any stress on the neck of the tube, or the small board that's attached to it. Doing so will *not* be much fun if things snap, been there and done that one (I was helping carry a TV into my office to work on it, and the guy who brought it over had removed the back to fit it into the trunk of his car). -- Member of the toughest, meanest, deadliest, most unrelenting -- and ablest -- form of life in this section of space, ?a critter that can be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters" - Information is more dangerous than cannon to a society ruled by lies. --James M Dakin