restoring a second OSI

[dave]

I am really enjoying watching the progress of your restoration of this machine.

Dave

[nama]

I'm still a little concerned about the drive, and of course still have not tested it with a OSI formatted disk. I have put in an unformatted disk and it seems to spin well, and came out no worse for ware. I did leave it spinning again (actually alway spins so I can't really avoid it) and after about 15 mins it stopped and made a loud humming noise. I quickly turned it off. On power ups again it was spinning, but it had loud scratchy sandpaper on metal like noise again...apart from the noise it seems to be spinning ok. I pulled it apart and couldn't find anything obvious.

Exactly how noisy are these units?

EDIT: So I pulled the unit apart again. Had a close inspection. Tiped the unit on it's side, and this time I let it run for about 2 hours keeping a constant eye on it incase it freezes up again. Only once did it seem like it was going to seize, but it didn't. Unfortunately that hasn't seemed to help the issue very much. I have downloaded the manual and I will take a look to see if there is anything I can do to eliminate the noise.

[nama]

I have pulled the drive apart and pulled out the main spindle and bearings. Actually once I loosened the nut the unit spun pretty freely, but I have decided I will replace the bearings anyway as one bearing looks to be in a bad condition. It spins ok, but it may not have any grease left inside and is pretty noisy and loose. Thats where I think the issue with the noise lies. The bearings are SSR-4ZZ with a flange. I think that are fairly common. I'm not sure why it seems like the nut was over tightened and the bearings and spindle not spinning freely, but I'll look into that when I reassemble the unit.

[nama]

Picked up two bearings today and replaced the old ones. The difference is night and day. The mechanism is so quite now, and spins smooth with a nice whirr noise. Great!
Still have no idea if the drive actually works or not.

[dave]

That's really neat. Would you care to post a little tutorial on changing the bearings, and can you show which nut was overtightened? I imagine many of these drives may have end-of-life bearings.

[nama]

I can do one better than that. I can point you to Terry's (AKA Tezza. Yet another Kiwi) site, where you can find the MPI 51 manual. Maybe this would be great to link to from your site.

http://www.classic-computers.org.nz/blo ... -drive.htm

Terry is an avid collector of old computers, especially the Australian and New Zealand TRS-80 clone called the Dick Smith System 80. The MPI 51 was standard issue in the System 80 drives.

Page 71 of the manual has a cross section diagram of the area I was working on, and Page 69 and 71 has a step by step disassembly and assembly guide. The screw that was possibly over tight was part #59 in the diagram.

[dave]

[Moved from "Files. . .Looking for OSI 542C Keyboard Schematic "]

Merry Christmas!

I only have the 542 and 542B schematics. I checked the SAMS manual, and it appears to only have the "B" revision. However, the only difference between the "B" and "C" models, according to Mark's OSI page is the addition of a safety time delay on the reset button. You can certainly get the rest of the keyboard running without the reset circuit, which is likely entirely separate. In fact, most OSI owners I knew disconnected the break key from the reset and wired it into the keyboard matrix, and put a separate reset button on the back of the machine.

If your 542A keyboard works well, then the cable is most likely OK. Also, if the computer still works with the bad keyboard plugged in, and there's no smoke, then you should be in good shape for debugging.

The actual keyboard function is based on the 2 8T26's, and the two 7475 latches between the 8T26's and the keys. The 542B and C have to multiplex the sound functions with the keyboard function, accomplished with the LS138 and a couple of 7404 inverters. So, you may want to check the R/W line (pin 3 on the 74138, pin 10 on the connector--you may want to double check to see where pin 10 goes on the C version.) Then check to see if you're getting pulses on pins 4 and 13 of the 7475's (U5 and U6 on the 'B" version). Then, check the pins 9,10,15,16 on each of the 7475's. If these are good, then you probably have good 8t26's and 7475's, as well as the '138. If you see inactivity on one, then you can suspect the 8T26 (more likely) or 7475 associated with that stuck pin. To check if it's the 7475, just look at the corresponding input--if the input is pulsing but not the output, then the '75 is bad. Otherwise, suspect the 8T26.

One thing about this keyboard is that If someone plugged the keyboard in backwards, then several chips, or every chip, could be damaged.

Good luck!

Dave

[nama]

I started working on fixing the 452C Keyboard. As all but the LM393 are socketed, and many of the IC's are the same I thought the simplest way to fix it was to but a few IC's (6 in total. I even managed to track down some 8T26's at Akihabara) and just replace each IC one at a time and test. If the replaced IC didn't fix the problem, then I put the origin IC back and move to the next one...well, that was the plan. Unfortunately I went through and replaced every IC (13 in total) but the problem didn't resolve itself...

Now I can think of a few reasons for this:
- The problem doesn't lie in the IC's tested, and is in another part (resistors, diodes, capacitors or the LM393)
- the problem is a bad socket
- the problem is in multiple IC's (or components) and replacing one at a time would not find both issues.

Anyway, I thought I'd post what I have least about the keyboard.
With shift lock on the keyboard seems unresponsive. Pressing H, D or M has no effect.
Starting the machine with shift lock of and then applying shift lock has the same effect as above.
With shift lock off, pressing ANY key jumps to the Monitor, but pressing keys after this has no effect.

Having a very quick look with the logic probe, many areas of the circuit seem to stop pulsing when shift lock is activated...no idea why or even if this is helpful info for trouble shooting. I'm going to look at the 542B schematic to see if this gives any insight into the shift lock weirdness.

If anyone has any ideas of what I should do next, please let me know. Again I would greatly appreciate any help.

[billdrom]

The 542B has extra functionality over the 542A keyboard.
Such as the tone, noise and A15 board interface circuits.
Maybe you can just remove the unneeded chips and just concentrate on the keyboard interface itself.
Look at page 7 of the 540A schematics for the only required chips.

http://osi.marks-lab.com/boards/schematics/OSI540A.pdf

Only two 7475 latches and two 8T26 bus drivers and one 7402 are required for the keyboard along with 8 diodes and pullup resistors.
It should be easier to troubleshoot with only these parts installed.

Good luck!

Bill

[nama]

Hi Billdrom,
Thanks for pointing me in the to that schematic. Sounds like a good idea to eliminate the parts that are not involved with the keyboard.
One thing I notice straight off is that my keyboard doesn't have 74LS02's. The list of IC's are as follows:

74LS04
DM7417
74LS75
74LS193
74LS138
74LS390
8T26

Looking at the 542B schematic I believe the 74LS138 and 74LS04 are doing the job of the 74LS02. So as per your suggestion, I think I'll remove all other components so I can concentrate on the Keyboard section only. In fact I think I will replace all 6 components at one time, and work backwards from there .
First thing to do is take some high-res photos of the keyboard so that I don't forget what goes where.

Cheers