PDA

View Full Version : Osborne drive problem



tezza
March 30th, 2009, 02:42 AM
Although my Osborne 1 works, it has issues in that it only works for about 30 minutes a time before it collapses in a screaming heap (garbage screen, drives spin wildly etc.). Drive B also fails to read and write consistantly.

I could hear a faint intermittent buzzing every 10 to 30 seconds or so in drive B and this buzzing corresponed with the screen contracting a little. It was after a long period of buzzing (and the screen contracting alarmingly) that the machine usually crashed.

Anyway, I disassembled it tonight to take a closer look. The problem certainly lies with drive B. Not only is there a buzzing sometimes (I think from the stepper motor) but the stepper motor occasionally comes to life and moves! Just rotates about 1/4 of a turn or so all of a sudden then stops.

I put a voltmeter over the black and red wires going into the stepper motor. When there is no buzzing they read 0v. When I hear the buzzing, they showed a voltage. This could be anything from .2 to 1v. Sometimes even over 1v (sometimes 3v) and when a higher voltage occurs (usually over 1v), that's when the stepper motor moved.

I've concluded there is some faulty component on the board, sending voltages to the stepper motor which shouldn't be there. When this occurs it affects the power supply and when it's bad, it affects the power supply so much it brings the machine down.

Anyway, I'll try and get a circuit diagram for this Siemens drive and attempt to figure out what might be causing it. I've just had a quick look so far.

I just wondered if anyone had come across this before on a drive? Any idea what component might cause this?

If worst comes to worse I could just leave drive B unplugged. I've found the machine is a steady as a rock in that state, but a CP/M machine really needs two drives to be whole! I'd rather fix it if I can.

Anyway, any ideas gratefully accepted.

Tez

Chuck(G)
March 30th, 2009, 08:46 AM
My feeling is that it's not the drive, but your power supply.

The stepper on a drive will buzz if the +12 supply has too high a ripple component. My first suspicion would be the AC input filter capacitors on your PSU.

A quick check would be to try the O1 on a PC power supply. You need +/- 12 and +5 volts, which makes a PC PSU very handy for the job.

tezza
March 30th, 2009, 09:54 AM
Hi Chuck,

Thanks for your reply. However...

1. The buzzing, voltage pulses to the stepper and stepper motor movement only occurs on the B drive.
2. If the B drive is unplugged the machine is fine. It's steady and doesn't crash.

If it was a PSU problem wouldn't it affect both drives, and still be a problem even if one is unplugged?

I'll check the power supply for ripple non-the less.

Tez

Chuck(G)
March 30th, 2009, 10:06 AM
Well, the next most likely explanation is a bad stepper motor (i.e. one with a shorted winding). Before I'd write the stepper off, however, I'd check U21 and 22 (75451 drivers) on the drive electronics PCB. There's a chance that one of them could be bad.

One of the problems with early drive electronics was that current was kept flowing through the stepper windings all of the time. A change was made in the electronics for the Rev. G boards to take care of this, but if you've got an earlier (e.g. Rev. B) drive, you still have the situation, which can lead to stepper overheating.

tezza
March 30th, 2009, 10:14 AM
Thanks. I'll investigate tonight if I get some time. I think I have one of the later boards. It is a 2nd revision double density Osborne. I've got a service manual so I should be able to tell.

Before I opened it up I'd hoped it was the PSU. Much easier to fix. I will check that as well though just in case.

Do you know what readings should be coming off the pins of the 75451 drivers if they were bad?

Tez

Chuck(G)
March 30th, 2009, 10:32 AM
Do you know what readings should be coming off the pins of the 75451 drivers if they were bad?

If you try to step the drive, a bad 75451 will show 0v (or nearly so) on pin 3 or 6 no matter what the state of the input pins (1,2 and 6,7) is. The 451 is basically an open-collector AND gate, so if either (or both) input pin is low, the output will be low. If both input pins are high, the corresponding output should be nearly +12.

tezza
March 30th, 2009, 11:52 AM
Great. That's good information I can apply for testing tonight.

On thinking about it, it could actually be two issues, a ripple on the power suppy which causes a crash every 30 mins or so AND the drive problem. It did seem quite stable after I unplugged the drive, but then again I didn't run it for very long. Testing is definitely needed.

I'm reluctant to start operating disks in it, even to try to format a disk in B to get the stepper motor going. Reason is that when it does crash, it usually zaps the disk in A. With no drive B, I can't make copies.

For those that have never worked on an Osborne I can tell you the drives are the hardest things to get to. You have to remove virtually everything else in or on the case including the monitor.

Tez

Chuck(G)
March 30th, 2009, 01:28 PM
One other thing to check is if the stepper has any winding shorts to ground. Simply unplug the stepper and check all terminals with respect to ground. There should be no continuity.

tezza
March 31st, 2009, 02:41 AM
Ok, progress.

I checked the power supply. Rock stready at 12v without the hint of an AC ripple. The only thing it does do alarming is when the stepper motor fizzes a lot and the voltage can suddenly drop to 7v before rapidly climbing again to 12. No wonder the machine crashes.

Checked if the stepper motor had windings shorting to ground. No, seemed fine. As a further check I swapped over the disk boards just to satisfy myself the problem was in the boards. It was. The problem moved with them.

I checked those two 75451s . Ah ha, the one closest to the connectors was exibiting odd behaviour in pin 1.


http://classic-computers.org.nz/blog/images/2009-03-31-suspicious-readings-on-pin-1-driver-ic.jpg

Pin 1 would sit at about 5v. Then it would suddenly drop to 0.2v. Usually I'd hear a bit of crackling from the stepper motor when this happened. It would stay at 0.2v, then it would jump to 5v again. After a few seconds it would drop again. On Pin 3, it would go from 12v, down to .2 v when this happened as you would expect.

The other pins were steady, and all the pins of the other 75451 were steady. I checked pin 1 on the 75451 in the other drive and this did not show this behaviour. It just held steady at 5v.

So, the question is, is it likely to be the IC itself, or is it something else (a faulty capacitor?) pulling pin 1 low? Any thoughts? Unfortunately I don't have a circuit diagram for this drive planar. Anyone know where I might find one?

Just on the IC itself, are 75451s common? I haven't looked at any of my junk cards or boards. Am I likely to find one there?

Thanks for the help to date. It's helped me isolate it quite quickly.

Tez

tezza
March 31st, 2009, 03:03 AM
Just thought about this a bit more and figured that as it's an input (as opposed to an ouput) pin that shows the odd values, the cause would be further upsteam, not the IC itself, yes?

Tez

tezza
March 31st, 2009, 03:31 AM
Found some schematics. They were hiding in my Osborne technical manual PDF after all, just not where I expected them to be.

Tez

chuckcmagee
March 31st, 2009, 07:58 AM
Uh, yes, 0.0 to 0.8v is considered logical zero. I don't remember the bible for the 1 value, a guess would be 3.? to 5.0v. So, a input pin going to 0.2v would be extremely common!!

chuckcmagee
March 31st, 2009, 08:34 AM
Yep, pin 1 is 1/2 of a Open Collector Nand gate. I guess the Open Collector thing must reverse the logic once again. The spec sheet states that this is a AND gate, not a NAND gate, except the diagram clearly shows a NAND gate. Anyway, only when pin 1 and pin 2 are high does the NAND output go low. This turns OFF the output transistor, so if you had a +20 or +30 voltage (current limited somehow) on pin 3, the voltage would go up to the normal plus whatever voltage. This does indeed appear to flip the logic once again (ie, NAND going low makes pin 3 positive, NAND being high makes pin 3 closer to zero voltage)

There will be a test on all this at 2 p.m.

Chuck(G)
March 31st, 2009, 10:03 AM
Pin 1 of U22 is a bit odd in that it's the only output of the LS139 demux that's fed back around to the 4070 NOR gate (U13) and pulled up via a 4.7K resistor. Some probing is definitely called for. Does the output on pin 4 of U15 make sense when viewed in the context of its inputs? If not, then I'd start by replacing U15. I'd also check the section of RN5 responsible for the pullup to make sure that it hasn't opened.

tezza
March 31st, 2009, 10:21 AM
Yes, according to the truth table I have for the 75451N it's behaving as it should. The spec sheet I found shows it's a NOR gate and a low value on 1 even though pin 2 is high would drive 3 low.

Point is pin 1 shouldn't be receiving a low voltage input at all. With the drive at rest, it should stay high, like pin 1 on the corresponding IC on the good drive. That one never moves. Instead pin on the 75451 on the bad drive keeps flipping low often. Whenever it does this, I get some crackling and movement in the stepper motor.

Is it reasonable to conclude then that some upstream component is driving this pin low. Just looking at the circuit diagram this may involve pin 4 on the 74LS139 at U15, yes?

Tez

tezza
March 31st, 2009, 10:23 AM
Pin 1 of U22 is a bit odd in that it's the only output of the LS139 demux that's fed back around to the 4070 NOR gate (U13) and pulled up via a 4.7K resistor. Some probing is definitely called for. Does the output on pin 4 of U15 make sense when viewed in the context of its inputs? If not, then I'd start by replacing U15. I'd also check the section of RN5 responsible for the pullup to make sure that it hasn't opened.

Opps...simultaneous post..LOL.

We are thinking along the same lines. Just gotta get to work now Chuck, but I'll investigate the upstream circuitry tonight.

Thanks for your help.

Tez

chuckcmagee
March 31st, 2009, 10:38 AM
And I was right. The Open Collector thing works like a Inverter. So, the 75451 is 2 AND gates, not 2 NORs.

EDIT:

Well, other Chuck, I was referring to Tez's reference above to a NOR gate. I have to admit to skipping over your earlier post about AND gates, which is good. I got some mental exercise trying to figure out how those open collector guys work.

Chuck(G)
March 31st, 2009, 11:03 AM
And I was right. The Open Collector thing works like a Inverter. So, the 75451 is 2 AND gates, not 2 NORs.

My turn:


The 451 is basically an open-collector AND gate, so if either (or both) input pin is low, the output will be low.

tezza
April 1st, 2009, 02:36 AM
Ok, using the circuit diagram, IC datasheets and a multimeter, I've worked back through several logic ICs whose output was flip-flopping because the input also was.

I come to an NE555 timer IC and I suspect I might be getting close to the problem (hopefully).

Here is the relevant part of the circuit diagram...

http://classic-computers.org.nz/blog/images/2009-04-01-osborne-fix-ne555.jpg

Here are the readings on the pins. I've included the reading from the good drive for comparison. A H<--->L indicates the voltage flips from high to low and back again on occasions (accompanied by the odd buzzing and small stepper movements). All measurements were when the drive was at rest.

NE555 Readings

Bad Drive

Pin
2 - TRG H<-->L
3 - OUT H<-->L
4 - RESET H
5 - CONT 3.3 V
6 - THRES H<-->L
7 - DISCH H<-->L
8 - Vcc H


GOOD Drive

Pin
2 - TRG H
3 - OUT L
4 - RESET H
5 - CONT 3.3
6 - THRES L
7 - DISCH L
8 - Vcc H

On the bad drive's NE555 the values swing about like a barn door in a gale. The question is, is it likely to be a faulty NE555 or not? I wouldn't have thought an input (in this case the TRG) would be effected by a fault in the IC it's on unless it was the upstream circuit at fault?

I didn't have time to check the RN1 out fully. Maybe there is a short somewhere there? The voltage on pin 1 of RN1 was steady. Should the C15 capacitor be a suspect?

One unusual thing is that after some time poking around with the multimeter on the NE555 IC, it suddenly started to behave itself. There was no buzzing from the stepper for about 15 mins and no changes on the IC.? Could a low impedence from a multimeter have done this, and temporarily fixed a problem?

Or maybe it was just co-incidence? The fault is kind of intermittant.

More investigations tomorrow night but as always, any comments welcome.

Tez

Chuck(G)
April 1st, 2009, 10:20 AM
My question here would be what do you see on pin 9 of U8/pin 7 of U9; i.e. on the other side of C15?

A strong possibility at this point is a cold solder joint in the area of the NE555. Also, check the resistance of RN1, pins 1 and 2 with the power off. Better yet, bridge that resistor with one of say, 10K to see if that makes things settle down.

Packaged resistors don't go bad very often, but it's been known to happen.

tezza
April 1st, 2009, 12:12 PM
My question here would be what do you see on pin 9 of U8/pin 7 of U9; i.e. on the other side of C15?.

Hehehe. I thought you might ask that. I did measure pin 9 of U8 (which is an input pin as far as I can see) and it seemed quite stable. I didn't check it for very long though, so I might do it again as pin 3 of U9 was also unstable and it's input only comes from the pin 8 output of U8.

I didn't check pin 7 of U9. That may well be unstable as it receives an (unstable) input feed from the NE555 via U8 (in pin 13 and out of pin 11 then into pin 13 and pin 1 of U9). This is how I traced the fault back to the NE555.
This part of the circuit seems quite circular, but it maybe I haven't fully grasped the electronics of it.

Chuck, I'm assuming you have this circuit diagram, yes?



A strong possibility at this point is a cold solder joint in the area of the NE555. Also, check the resistance of RN1, pins 1 and 2 with the power off. Better yet, bridge that resistor with one of say, 10K to see if that makes things settle down.

I might give that a go. When you say "bridge the resistor" you mean over pins 1 and 2, correct?

It has all the hallmarks of a cold solder joint or a failing connection/short circuit as it is intermittent. It comes and goes, with variable frequency. Would I be correct in assuming if an IC had failed, it would be unlikely to switch intermittently between high and low on it's output pins (assuming an unchanging input). It would simply be "stuck" so to speak.

Tez

Chuck(G)
April 1st, 2009, 12:18 PM
Chuck, I'm assuming you have this circuit diagram, yes?

Yes, it's on bitsavers.


I might give that a go. When you say "bridge the resistor" you mean over pins 1 and 2, correct?

Yes, indeed. If that resistor is failing, it could create the effect. But my money's still on a cold solder joint.

nige the hippy
April 1st, 2009, 12:53 PM
If you can't find a dodgy joint (and just resolder rather than trying to see a microns-wide crack), there's always the possibility that the 555 has a failed internal bond. (probably the VCC (?)). To keep the 555 fairly permanently on you need loads of hash on the trigger input, a steady state will be blocked by the series cap.

tezza
April 1st, 2009, 01:04 PM
If you can't find a dodgy joint (and just resolder rather than trying to see a microns-wide crack), there's always the possibility that the 555 has a failed internal bond. (probably the VCC (?)). To keep the 555 fairly permanently on you need loads of hash on the trigger input, a steady state will be blocked by the series cap.

Would a failure of the cap cause the pinout readings I've reported?

I'm suspicious of all caps now. The last two repairs I've done were due to innocuous looking (but indeed faulty) capacitors.

"hash" on the trigger state? Not marijuana obviously, but do you mean high voltage pulses here Nige?

Tez

Chuck(G)
April 1st, 2009, 02:59 PM
C15 would be very low on my list. It's a .01 uF ceramic not under any particular stress--but check it anyway. My priority list would run something like this:

Cold solder joint
RN1, pins 1 and 2 possibly open or intermittent
U4 possibly bad
C15, 16 or 17 maybe.
But that's all assuming that the signal on U8 pin 9 is behaving itself.

tezza
April 2nd, 2009, 01:29 AM
Well, not EXACTLY sure what the fault was but it appears to be fixed.

Last night while probing around the NE555 area with the multimeter, the problem seemed to suddenly fix itself. Tonight when I switched on the machine I waited and waited...and no buzzing...no stepper movement? No problem.

Even so, I tipped the board over and resoldered the resistors and NE555 pins. I then put it through some pretty strenuous disk operations for about two hours. Fine. No crashing, no unusual noises, no sudden "garbage screen" shutdowns.

I figure it was a dry solder joint somewhere in the NE555 circuit. My probing around in that area with a multimeter last night, touching solder pads etc. may have just caused a temporary reconnection. Anyway, the solder has been reflowed in that area so any connections should now be solid. Hopefully that's the end of the story.

Thanks Chuck(G) and others for comments and in particular for showing me where to start looking.

Now I face the biggest hurdle yet though...


http://classic-computers.org.nz/blog/images/2009-03-31-fully-naked-osborne.jpg

...putting the Osborne back together!!

Luckily I took pictures during the disassembly :)

Tez

cosam
April 2nd, 2009, 02:01 AM
Good stuff - enjoyed following this thread.

Here's hoping it stays fixed. It often seems like these kinds of problem have a knack of returning just as soon as you've finished reassembling the machine, and that the probability of this occurring is directly proportional to the effort required to take the thing apart again ;-)

tezza
April 2nd, 2009, 02:09 AM
Good stuff - enjoyed following this thread.

Here's hoping it stays fixed. It often seems like these kinds of problem have a knack of returning just as soon as you've finished reassembling the machine, and that the directly probability of this occurring is proportional to the effort required to take the thing apart again ;-)

Yes. That exact thought had occurred to me. I'd reassemble...and the fault would be back...gleefully buzzing at me :sneaky:

However, I'm an optimist :)

Tez

Chuck(G)
April 2nd, 2009, 08:41 AM
Well, at least it works for now--and maybe for the next 20 years or so... :)

chuckcmagee
April 2nd, 2009, 11:51 AM
Good stuff - enjoyed following this thread.

Here's hoping it stays fixed. It often seems like these kinds of problem have a knack of returning just as soon as you've finished reassembling the machine, and that the probability of this occurring is directly proportional to the effort required to take the thing apart again ;-)

Gah, that is exactly what happened with my repair to the Osborne OCC2 (the bigger screen one). Fixed the solder. Got it back together. Worked about 3 seconds then back like it was. I haven't managed to get up the mental energy to tear the whole thing apart a second time yet. Per usual, bad soldering joint on a high power output transistor. Looks like a better reflow job is required.

Mike Chambers
April 2nd, 2009, 02:28 PM
My feeling is that it's not the drive, but your power supply.

The stepper on a drive will buzz if the +12 supply has too high a ripple component. My first suspicion would be the AC input filter capacitors on your PSU.

A quick check would be to try the O1 on a PC power supply. You need +/- 12 and +5 volts, which makes a PC PSU very handy for the job.

EXACTLY what i was about to post :P

mainly because it explains why the other issues after a while... garbage screen, etc. i would suspect some suspect caps in it.

Chuck(G)
April 2nd, 2009, 03:25 PM
EXACTLY what i was about to post :P

mainly because it explains why the other issues after a while... garbage screen, etc. i would suspect some suspect caps in it.

That was my thought, but Tezz noted that the problem didn't happen when the second drive was unplugged. Could be that the second drive was loading the +12 something awful. I don't know how much headroom an O1 PSU has. If it happens again, I'd hook a meter onto the +12 and look for some voltage sag when the hash happens.

tezza
April 2nd, 2009, 03:56 PM
That was my thought, but Tezz noted that the problem didn't happen when the second drive was unplugged. Could be that the second drive was loading the +12 something awful. I don't know how much headroom an O1 PSU has. If it happens again, I'd hook a meter onto the +12 and look for some voltage sag when the hash happens.

Yes, my initial thought was PSU problems too but...

1. A scope check shows no AC ripple on the powersupply (unless I'm not using the scope properly?). The only time I do see some AC "noise" is when the voltage sensitivity "knob" is right down to 0.01volts per division on the meter. I'm assuming this is just background

2. The second drive was the only one that buzzed and moved the stepper. This was the only drive that showed those odd flip-flop voltages in the ICs I checked. In the first drive those measurements were stable as. If it was a PSU problem I would expect both drives to be affected.

3. With the B drive unplugged, the system was stable. No buzzing, no crashing. A drive was fine.

As mentioned in one of the first posts, I did have the scope attached to the board to pick up the power just after it leaves the PSU, when the B drive started to hash quite badly. I then saw a huge voltage drop from 12v down to 7 v, then back up again. So it looks like when the problem occurred in the B drive, it did indeed load the +12v dragging it down. This kind of thing certainly would crash the machine.

When the buzzing is heard, the screen contracts a little and shimmers. It only did that when the drive buzzed. When the system did crash it tended to be after prolonged stepper motor "hashing" noises.

It is possible the two are unrelated and they just (by coincidence) occur together frequently. Assuming that was the case and the crashes were a PSU issue. Wouldn't you see ripple on the scope if a cap was bad?

Hopefully this fault won't reappear. I might try to run the system at length (8 hours) just to check before I re-assemble. I'm reluctant to babysit it that long though. On the other hand, if it crashes I want to be there to switch it off also, as the drives spin madly (and sometimes the beeper stays on),



Tez

tezza
April 4th, 2009, 01:53 AM
I just ran the Osborne today for SEVEN HOURS solid. No crashes, no disk noises, steady as a rock.

Considering it never made it past 30 minutes before I'd say the fault is fixed...for now. Here's hoping it stays that way.

As usual, I wrote an account of the repair up for my blog (http://classic-computers.org.nz/blog/2009-04-03-osborne-drive-fix.htm).

Thanks to everyone who helped with this especially Chuck(c). It's good to see the ugly old beast behaving itself now. (the Osborne, not Chuck :))

Tez

Mike Chambers
April 4th, 2009, 01:44 PM
I just ran the Osborne today for SEVEN HOURS solid. No crashes, no disk noises, steady as a rock.

Considering it never made it past 30 minutes before I'd say the fault is fixed...for now. Here's hoping it stays that way.

As usual, I wrote an account of the repair up for my blog (http://classic-computers.org.nz/blog/2009-04-03-osborne-drive-fix.htm).

glad all is good.... kind of. two floppy disk drives would still be nice to have in there.



Thanks to everyone who helped with this especially Chuck(c). It's good to see the ugly old beast behaving itself now. (the Osborne, not Chuck :))

Tez

lol..

Chuck(G)
April 4th, 2009, 02:33 PM
Who are you calling "old"? :mad:

Mike Chambers
April 4th, 2009, 03:01 PM
i thought he was talking about me until "old" ... although i'm 25 this month, i'm starting to get there! :eek:

hey tezza, i still think there may be a PSU issue if it can't manage to run both drives together properly. either that or your B drive has some suspect circuitry.

tezza
April 4th, 2009, 07:35 PM
i thought he was talking about me until "old" ... although i'm 25 this month, i'm starting to get there! :eek:

hey tezza, i still think there may be a PSU issue if it can't manage to run both drives together properly. either that or your B drive has some suspect circuitry.

Opps, you must have misunderstood my posts and/or my blog Mike. I CAN now run the machine with both drives together.

The machine always did behave itself with one drive. The fault was on the second drive. It seems the short there loaded the power supply too much and bought the whole machine down when it did so. That second drive is now fixed and reattached to the machines.

Unlike before, even with both drives attached the Ozzy is now perfectly stable. No buzzing, no garbage screen and crashes. I ran it for seven hours non-stop without a problem. I even tried to shock it back into being faulty with freeze spray! Tested the 12v and 5v lines yet again. No ripple.

I'm giving it a clean bill of health.

Fingers crossed it stays that way, but I guess that is the case with all vintage computers.

Tez

Druid6900
April 4th, 2009, 08:13 PM
Good work, as usual, Tez.

You might want to consider opening a Vintage Computer repair shop when you get tired of Academia :)

tezza
April 5th, 2009, 01:35 AM
Good work, as usual, Tez.

You might want to consider opening a Vintage Computer repair shop when you get tired of Academia :)

Hehehe. Thanks. Digital electronics has a certain purity that appeals. It's either right or wrong...it's either on or not...it's either high or low. If there is something wrong there IS a reason for it and it's usually possible to find it through deduction and testing.

It's logical and clear cut.

That's often not the case in my day job, which of course involves humans. They are far more complex entities than these old machines and hence are much harder work!

Fixing these old machines is a (masochistic?) form of relaxation! :)

Tez

chuckcmagee
April 5th, 2009, 01:50 AM
Buying Tandy 1100FD units has sure proved to be masochistic in my case. Well, gooooolly, the floppy drive unit ain't working right. What a surprise! Hey, I should go looking for those super rubber bands too. Too bad the nearest large store is 1 hour drive there, 30 minutes in the store, 1 hour drive back.

tezza
April 5th, 2009, 02:43 AM
Yea. Nearly all my recent fixes have been drives! They are the Achilles heel of most older systems unfortunately. :(

Tez

Terry Yager
April 5th, 2009, 02:31 PM
That's often not the case in my day job, which of course involves humans. They are far more complex entities than these old machines and hence are much harder work!

Tez

Well, we have that much in common. I've spent half my life dealing with people, the other half dealing strictly with machines, and the other half with both simultaneously. Machines, and how people interact with 'em, ya gotta try carnival work sometime...

--T