• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Apple Lisa questions

stepleton

Well-known member
There is nothing too sophisticated that you can do through NeoWidEx or BLU until the Widget decides it's ready to talk to you.

The Widget has several subsystems, but the ones that are important today are the "controller" and the "servo". The servo is what swings the heads and the controller is what the Lisa talks to, the conductor for the whole Widget orchestra. Both of these subsystems are controlled by separate microcontrollers. The servo is pretty self-contained: in fact, the controller tells it what to do over a 57600-bps serial link! (TTL voltage levels I think, but still.) The controller is the topmost circuit board and is what the 26-pin ribbon cable from the Lisa plugs into.

When the Widget starts up, if I recall correctly, it tells the servo to reset itself, and then it waits for the servo to say that it has done this. The servo goes off and tries to swing the heads over their entire range of travel, but on your drive this is not possible because the servo hasn't freed up its sticky old mechanism yet. (Or maybe there's something inside the drive that's preventing it from articulating itself, but I suspect the former for now.) So the servo never tells the controller that it's ready.

I'm pretty sure the controller is just going to be stuck waiting forever until that servo is ready to go. There is nothing you can do with anything that talks through the controller until the servo is ready. The controller doesn't care about you at this point, it only cares about the servo.


My advice as always is be patient. We have seen signs of progress. Keep exercising the servo by resetting the drive. NeoWidEx can't perform the reset that BLU (probably) does, which involves toggling one of the signal lines in that ribbon cable, which effectively "reboots" the drive. Does BLU's reset cause the Widget to try and move the servo again? If it does, this is what I might keep repeating in hopes that the head will move a little further each time. If BLU doesn't do this, then I might try resetting the computer via the reset button on the back. You don't need to load BLU or NeoWidEx for this --- each reset ought to toggle one of these "reboots" and another servo wiggle attempt, I think.

If you don't want to keep doing this, then the ultimate Widget diagnostic tool is UsbWidEx. This is an all-in-one hardware device that (among many other excellent qualities) can talk to the servo directly. If you build yourself one of these, you can cut out the controller intermediary and reset the servo all day long, plus inquire about its feelings and do other things. I'm very glad to have one of these gizmos, but it may not be necessary for you if you can coax the servo into doing a complete swing.
 

stepleton

Well-known member
More to know about the servo:

Roughly and abstractly speaking, a "servo" is a device that moves to a specific place by measuring where it is and moving itself until it gets where you tell it to go.

The Widget's servo is an electronic, mechanical, and optical assembly that moves the drive heads to a specific track. In order to do that, it needs to know where to find the track the controller tells it to visit. Inside the drive where the arm and the platter is, there is a small piece of glass with fine lines on it that look like the lines on a ruler. The arm has an LED and a light sensor that surround this glass piece. When the arm moves, the sensor detects when the LED light is interrupted by the ruler lines. This is how it counts its progress when it moves over tracks.

The piece of glass is affixed to the drive with glue, and over the years, this glue is known to fail, and when it does, the glass piece dislodges itself. This prevents the servo from operating because it can't tell where the arm is. I've never encountered a Widget where this has happened (touching wood here), but apparently it's not too unusual. Repair is very difficult, because opening the sealed hard drive interior is required, and I don't know someone to have accomplished this.

I mention all of this because while I hope and suspect that you just need to be patient and keep exercising the drive until the arm bearing regains its full range of motion, there are other things that might be wrong with it, and this is one of them. Perhaps the servo isn't traveling because the glass piece is dislodged and it can't tell that it's moving.

But if this were the case, I don't know how the free glass piece couldn't avoid migrating to the spinning disk itself somehow and eventually cause a catastrophic failure as the drive turns itself into a rock polisher.

Let's not try to investigate this possibility right now. I still recommend trying to free up the arm. Please don't try opening the drive to have a look!
 
Last edited:

ried

Well-known member
Let's not try to investigate this possibility right now. I still recommend trying to free up the arm. Please don't try opening the drive to have a look!

I won't! Promise :) I ran it again for a few hours tonight and am not giving up hope on the Widget. We'll get that sorted at some point, I'm sure.

Thank you for sharing such detailed knowledge of these with the community, @stepleton - even newbies like me. I really enjoy learning about this stuff!

In the meantime, I'm looking at hooking up a ProFile drive to the Lisa's Parallel Interface Card. I'm having a hard time finding the correct cable for the ProFile, though. I did purchase a cheap one from Amazon, which is DB-25 (straight through) and male on both ends. It doesn't work with the Lisa's parallel port, however, because it has pin #7 (blocked on the card). It also doesn't mate up against the port on either the ProFile or the Apple /// very well, since its connector screws don't reach the machine.

Does anyone know a better parallel cable option?
 

Busterswt

Well-known member
You can make one pretty easily with two DB25 connectors and the color cable of your choosing, but a crimper comes in handy here. Unless you plan on sending that cable back, just remove Pin 7 from the cable or cut it out. I find needle nose pliers work well enough to bend the pin, or some small angled wire cutters can get in there.

IIRC the pin 7 block was to avoid having a serial cable connected to the parallel port, but I could be misremembering that.
 

mg.man

Well-known member
Does anyone know a better parallel cable option?
I've been testing / refurbing a couple of profile drives for a friend with a few Lisas we're resurrecting. The cable I have is a flat ribbon with with Pin 7 simply missing at either end. SO, it seems to me all you need is straight-through, male connectors both ends with Pin 7 missing.
 

ried

Well-known member
Well, the Widget has been run for several hours a day, probably 40+ hours in total. I'm starting to lose confidence that it's going to recover on its own, unless @stepleton says that he's seen them come back from even deeper slumber than that, of course.
 

stepleton

Well-known member
Sorry to hear about this @ried , and thanks for trying. The objective is not just to turn the spindle, though, it's to get the arm swinging.

The working theory is that gummed up lubrication is making the arm mechanism stiff and hard the move. If you can repeatedly trigger a power-on recalibration or an attempt at a surface scan (two things that the Widget does to check itself out when it boots), then the hope is that each time the arm will free itself up a little bit more and (with luck) regain its full travel.

Therefore it's not enough to just turn on the drive and walk away (since if the Widget fails its power-up exercises, it'll just give up and do very little besides sit there and turn the spindle). The idea is to keep resetting the drive, over and over, and getting it to try to swing the arm again and again. One way to do this is to power-cycle the drive a lot, but I think this is a pretty bad idea for a lot of obvious reasons. You can't low-level reset the drive within NeoWidEx (or BLU as far as I know). I think though that you might be able to get a reset to work by resetting the Lisa itself (using the reset button at the back of the machine). You can probably also do it without the Lisa by (momentarily) grounding pin 16 on the ribbon cable connector --- in fact, if that doesn't get the arm swinging, then either there's another problem or you'll need to get yourself a UsbWidEx.

What you're looking for after each reset is the Widget trying a little bit longer to move the arm around. If the activity LED on the front of the drive blinks a little longer, if you hear it swinging the arm around a little more, then this is the progress we want.


If that doesn't work, then I recommend asking over at LisaList. There's at least one person there I can think of who will be more trustworthy in their recommendations than me!
 
Last edited:

stepleton

Well-known member
But to your point in the message: if a Widget isn't making progress after a few dozen resets like this, I'd assume some other problem or a mechanism so sticky that the patented stepleton resets-n-prayers method ain't gonna cut it :)
 

ried

Well-known member
Great insight (again), thank you @stepleton. I did realize that the Widget tries doing unique things from a cold boot, so I've certainly done that 5+ times per day. Will keep at it... you never know!
 

stepleton

Well-known member
Will keep at it... you never know!

Hmm... 5+ times isn't the same as a campaign of a bunch of resets and arm movement attempts right after the other, but it's more than nothing at all. I guess if I were you, I'd try one good session of more concentrated activity, and if that's not moving things along, moving on to something else.

You've replaced all (I think?) of the electronics, so we're pretty sure the problem must be mechanical, or at least inside of the hard drive assembly itself. I'd want a second opinion from LisaList experts before trying to articulate the servo manually. It will be hard to hear much over the spindle, but do you hear any other noises that suggest any kind of mechanical interference with the arm? Any clicks or knocking sounds when it reaches the current full extent of its travel, for example? I don't have experience with a fault of that kind, so at this point I'm just passing along the kind of thing I'd be looking for.
 

mactjaap

Well-known member
Absolutely a fantastic read! I would love to see your widget revived.

I also try to revive my 4 (or 5?) widget drives over the years. NeoWidex is your friend but there can be so many problems. I will try to see if I can contribute something useful, but stepleton is a magician with widgets.
 

ried

Well-known member
Hmm... 5+ times isn't the same as a campaign of a bunch of resets and arm movement attempts right after the other, but it's more than nothing at all. I guess if I were you, I'd try one good session of more concentrated activity, and if that's not moving things along, moving on to something else.

You've replaced all (I think?) of the electronics, so we're pretty sure the problem must be mechanical, or at least inside of the hard drive assembly itself. I'd want a second opinion from LisaList experts before trying to articulate the servo manually. It will be hard to hear much over the spindle, but do you hear any other noises that suggest any kind of mechanical interference with the arm? Any clicks or knocking sounds when it reaches the current full extent of its travel, for example? I don't have experience with a fault of that kind, so at this point I'm just passing along the kind of thing I'd be looking for.

On it. I'm going to power cycle the Lisa a couple dozen times this morning to see how the Widget reacts, if at all. All electronics were replaced (controller, read/write, servo boards) except the little board with the LED attached to the platter disk enclosure. No change to the drive's behavior. I do not hear anything mechanical happening inside the drive, but that also could be because the spindle is just so noisy. I can very faintly hear the arm when it goes through its limited motion, but doesn't sound like it's hitting against or contacting anything internally.

Absolutely a fantastic read! I would love to see your widget revived.

I also try to revive my 4 (or 5?) widget drives over the years. NeoWidex is your friend but there can be so many problems. I will try to see if I can contribute something useful, but stepleton is a magician with widgets.

You and me both! Working on it. And yes, he certainly is.
 

stepleton

Well-known member
On it. I'm going to power cycle the Lisa a couple dozen times this morning to see how the Widget reacts, if at all.
I think don't power-cycle the Lisa itself --- just turn the Lisa on and reset it several times via the Reset button on the back. If I'm not mistaken, then each reset will hopefully get the Widget to try to swing the heads again.

I'm worried that a lot of powering up and powering down could do more harm than good here! For a start, the spindle motor will draw a lot of current each time you power it on, and some of the driver circuitry could be overstressed in this way. The Lisa may have parts with similar fragility.

he certainly is.
This is kind to say, but my record is not so long --- there are just not so many Widgets out there! There are also a couple given to me by a forum member that I have yet to repair. It's on my list to investigate --- my thought has been to investigate them on camera someday.
 

ried

Well-known member
I purchased one of these "Apple Lisa Widget Drive Motors" (not sure if that's the correct term or not) from eBay:

That is the component that I damaged by breaking two pins off of the ribbon cable. Is there a non-destructive way of replacing that part of the Widget? I removed the three screws from the little metallic cover disc on mine, but all that did was enable the cover to spin in place (opposite the brake). It did not allow me to remove it.
 

stepleton

Well-known member
I doubt it. While I don't know for sure, there's likely no way to remove the arm from the shaft of that motor without disassembling the drive down to its mechanical elements. If you had the equipment to replace this motor completely (including the glove box or clean room), you would very likely already have the equipment to rebuild the hardware you have now.

Keep in mind that replacing parts like this isn't just about replacing old with new, it's also about placement to very fine tolerances and likely also some sort of recalibration process.
 
Top