Lisa widget drive with a banging head

[mactjaap]

Inspired by the long read about the Lisa and problems with the widget drive I started to test my own. I have six of them and some spare part. Only one of them is working and only if the mechanics are having a good day and cooperate. Sometimes I have to help the head a little bit.

Two of the disks behave the same way. They spin up, you hear the loud click of a releasing brake and the they start to “bang” with it head. I have tried to gently stop this with my finger, but that has no use. It will bang from start to end of the disk.

My questions, does anyone knows what this means and is it possible to repair?

This is a short video on YouTube:

 

[stepleton]

That's a new one to me. I'm not sure why it's doing that. One theory is that the coarse positioning function of the servo isn't working, and it's swinging the arm back and forth against the stops without getting feedback that it's moving at all. Are all the cables and other connectors plugged in and secure?

While the drive is off: if you turn it over gradually, can you hear anything moving around inside? It would be a sad problem if the coarse positioning graticule inside the drive came unglued and were no longer in its correct location. I don't know if the behaviour you observe would be caused by that.

 

[Paralel]

Does it actually hit anything with this "Bang" or is it just a result of the quick movement?

 

[mactjaap]

Hmmmm.

Does it actually hit anything with this "Bang" or is it just a result of the quick movement?

I don’t know. What should I look for without opening the drive?

 

[stepleton]

What should I look for without opening the drive?

As in this video from @ried I think you may be able to find the black piece that the flat flex cable enters and see whether it is moving through the full extent of its travel.

Here are some other things to look for:

Are all the cables and other connectors plugged in and secure?

While the drive is off: if you turn it over gradually, can you hear anything moving around inside?

 

[mactjaap]

Thanks for the update. Will do this research.

 

[mactjaap]

Checked the cables and they are OK. Checked for something moving inside. Nothing.

I have a disk which is handled very rough in the post. There’s something inside … I think the glass part which is loosened. But not on this disk.
I need a moment to connect the drive again and will film the movement.

 

[stepleton]

Nuts, well, so much for the simple stuff.

Have you got an oscilloscope? I wonder if it would be possible to scope some of the lines on the servo board and narrow down the problem.

Having a UsbWidEx would also be extremely helpful here for isolating the issue, but there aren't a lot of those things around. The idea would be to connect it to the servo and see what happens when you issue commands to it directly.

 

[mactjaap]

Thanks agian! Yes I have an oscilloscope… but never worked with it… but maybe we could give it a try!
Could be helpful and interesting for other widget owners too.

I have connected the FloppyEmu so we can do a NeoWidEx exploration.

 

[stepleton]

I suspect NeoWidEx will not tell us very much at this stage --- the servo is probably not reporting that it's in good health, which means that the controller won't be ready to talk to the Lisa or to NeoWidEx. But it might be worth a try.

Here is the haystack where we will need to look for our first needle: the servo schematics.

https://lisaem.sunder.net/LisaSchem/WidgetServo1.gif

https://lisaem.sunder.net/LisaSchem/WidgetServo2.gif

Let me know if anything obvious leaps out at you

What we need first of all is an understanding about how the servo works on the hardware level --- something I don't have at the moment. A careful study of the schematics would probably produce this, given enough time --- another thing I don't have a lot of right now.

There is plenty of stuff here that can go wrong. All kinds of analog circuitry that helps the servo know where it is and how much it's moving. For a traditional troubleshooting strategy, we'd take steps like these:

1. Understand better how it all works, maybe dividing the circuit up into separate sections that have different functions.
2. Come up with theories for how malfunctions in these sections could affect the servo.
3. Identify theories that match the bang-bang behaviour you're seeing.
4. Come up with ways to detect through measurement which of the theories is actually responsible for the behaviour.
5. Applying these methods, identify the component whose malfunction is actually causing the problem.
6. Replace or repair that component and hope it fixes things.

For an example of what I mean in option 4: we could suppose that the problem is coming from (a) part of the mechanism that measures the head location or (b) part of the mechanism that controls the motor that moves the head. To distinguish between the two, we'd want to show that measurements of the signals made by the head location system are correct/faulty and show that measurements of the control signals/voice coil actuation are faulty/correct. (In reality, we probably want to divide the servo system into more granular functional components.)

All of this is to say that we have a tricky troubleshooting problem on our hands --- and we also haven't completely determined that the problem rests in the servo board --- it could be somewhere else!

It would probably be a good idea to get started learning how to use your oscilloscope. Do you have any convenient way of practicing this skill --- maybe probing around a battery-powered radio or an Arduino-driven circuit or something like that?

How are you with understanding schematics and electronic circuits? (I would rate my abilities as "amateur" at best!)

 

[mactjaap]

Thanks for your information. It will be a long road to revive this widget!

I will start to switch parts. Busy weekend so I hope on Saturday.

 

[mactjaap]

Switched upper board and servo board. Same result. Head is banging. I’m thinking of opening the hard drive to see if the glass part is still attached.

 

[stepleton]

Here's a less risky approach to consider:

1. Take a look at the servo board schematic; maybe find a friend who likes looking at schematics if this is new for you
2. Identify the signal line that carries the coarse servo positioning information from the arm to the servo board
3. Practice using your oscilloscope to detect and display signals from electronics
4. Check to see whether there's a signal on the signal line you found in (2). You'd probably find a square wave or other repeating pattern that comes and goes with each back and forth sweep of the arm across the finely notched lines of the graticule (the glass piece).

If you see this pattern, you'll have a pretty good idea that the graticule is still in place, and you won't have had to open the drive. You will have eliminated one source of the problem and can move on to the next.

There are not many Widgets in the world. Would it make sense to play it a bit safer and give one the very best chance of being restored? You can always open it up later, and the opportunities to learn in the meantime could be substantial. (That's one way to look at a hard-to-diagnose fault! )