Mouse and Keyboard issue on Macintosh SE

[tony359]

Hi all
This is my first post here, I found this forum while googling for an issue I am having with an SE I am repairing.

Basically, the whole ADB stopped working in between reboots - no, I did not hot-plug my mouse! The mouse was working fine before.

The system boots up, reaches the desktop but when I move the mouse, all that happens is that the "apple" menu in the top left corner opens up. Then nothing. But I did not push the button.

The keyboard doesn't seems to do anything. I can open the debug window and all it types are "AA" for each press of the button (AA for pressing and AA for depressing) up to 2/3 of the first line, then it stops. (see pic below)

Things I checked.
- 5V is present on ADB
- GND is present on ADB
- Interference filter is ok
- I checked for continuity between ADB ports, ADB chip and VIA.
- I checked R14, R15 and the transistor driving the bus from the ADB chip side.
- ADB chip is getting a clock from the GLU chip (3.6Mhz)

Then I fired up my scope.
- ADB signal is present on the bus. Additional signals appear when the mouse is moved.
- I decoded the signal sent by the ADB chip: when just booted up, it's asking address 2 (should be keyboard by default) to "TALK". As soon as I move the mouse, the address changes to "3" which should be the mouse. That seems healthy to me. (see pic below)
- I followed the signal out of the ADB into the VIA. I see signals happening between the two chips when I move the mouse.
- I see extra IRQ signals and "Chip Select" on the VIA when I move the mouse.

The mac repair started with some memory issue so I thought that maybe the RAM addresses where the VIA stores its bits could be damaged. I downgraded the RAM from 4 to 2MB, swapping SIMMs but the issue did not change.

I swapped the mouse with another one and the same happens.

So:
- The issue seems to be on the logic board
- The ADB chip seems to be doing something.
- Signal seems to reach the VIA and get deeper in the Macintosh's guts

At this point I am a bit stuck. I am thinking that maybe the ADB is sending rubbish signals to the VIA (or maybe the VIA is) and the software/hardware receiving the signal is struggling to interpret them? I have some other SE, I could socket the ADB and VIA and test but I also like to understand things so I was wondering if any of you had any ideas of what could be going on here!

I've added some relevant schematics 'cause I know they're not so easy to find!

Thank you so much for your help and Happy New Year!

 

[stepleton]

I had a similar problem with my Classic, and it turned out to be the VIA.

Thread 1: Maaaaaaaaaaac Claaaaaaaaaaaaaaaaaaassic -- first description of the problem.
Thread 2: Very sick Classic --- could it be a bad VIA chip? -- I finally fix it, much later.

 

[tony359]

That would be great as I can source a VIA from China but not an ADB chip. Thanks for your feedback.

Can I please check with you where you saw VIA chips to purchase? I see some on ebay, pulled from systems.

Looking here: https://github.com/kr239/Macintosh-SE-Reloaded/blob/master/README.md

it says the Rockwell is equivalent to the 65C22N and the W65C22N seems to be compatible with the rockwell.

I found this: https://www.mouser.co.uk/ProductDetail/955-W65C22N6TPG-14

Does anybody know if this is going to work for me?

Cheers!

 

[stepleton]

I'm afraid I don't know where to get a new IC --- my VIA was scavenged from a friend's battery-bombed Classic.

If your 'scope investigation shows that the data going from the ADB chip to the VIA has the correct keyboard scan codes, then replacing the VIA sounds like a reasonable idea to me.

That said, I'd be sure to measure this at the VIA itself. I don't know, but maybe a broken trace between the ADB chip and the VIA could cause the problem you're seeing.

Good luck!

 

[tony359]

I've found this amazing video

and the author has the chip I found on the BOM which can be found here https://docs.google.com/document/d/1QhHpDew4saFxvt82W2TkdpnNA2W7tLZhe53Et64gpdY/edit.

So it seems the answer is yes.

BTW, I am based in the UK, if someone needs some of those VIA chips, please do let me know as I have to pay for shipment from the US anyways, maybe we could split!

Your point of probing the output of the ADB (or input of VIA of course) is actually a good idea. Difficult with a mouse but easier with a keyboard as I can set exactly the output code by pressing only once. Or I guess I could click the mouse to avoid the stream of data coming from movement.

I shall take a look indeed, thanks for the good idea. Might save me some money if I don't see the correct scan code there.

 

[tony359]

To update: a replacement ADB chip (a PIC with code from Tashtari) fixed the issue!

 

[stepleton]

Congrats!

 

[tony359]

I've spoiled the ending a bit here but here is part 1 of my repair. I trust this is not in violation of the rules of this forum.

Also please note what I said at the beginning of the video: this is my first Macintosh so please excuse my ignorance on how I approached the repair!

 

[mdeverhart]

Great video! You did a great job explaining everything, and I’m impressed how far you got given that you’re new to these old Macs. I’m looking forward to seeing Part 2!

 

[tony359]

Thanks for letting me know! I'll try and complete the edit by the end of the week! Any feedback is always welcome!

 

[tony359]

Part 2 is online! Let me know what you think.

 

[tony359]

And finally part 3

 

[Fred1212]

Late response but had same problem with my SE replacement of ADB also fixed it.

 

[tony359]

did it just suddenly fail? So weird to see the ADB side of the chip working totally fine...

 

[mdeverhart]

Glad to see you got it all working! I like the way you walk through and explain things, and methodically explain what you’re doing. Good stuff!

 

[rollmastr]

Nice, congrats on fixing up that machine. Makes me want to get another SE to fix up, especially since I have a good spare 20MB MiniScribe.

 

[tony359]

Thank you all for your nice words it was a fun journey!

‘A working miniscribe’ - I still haven’t come across one of those !

Any tips on reviving the one shown in the video? Shall I assume it’s permanently damaged or is there some procedures you’d recommend?

 

[rollmastr]

I've come acrorss two MiniScribe drives recently and one of them just worked while the other one needed some oil to the bearing of the stepper motor and a low level format just as in your video. Since your drive shows loads of bad blocks I assume you're just out of luck, but you could try moving the rotor back and forth just to clear any rough spots and retry the low level format.

 

[tony359]

with "rotor" you mean the stepper motor? Wouldn't that potentially cause damage on the surface if the head is moved and the platters are not spinning?

A friend (who also is on this forum) suggested to keep using it a bit to spread whatever substance they used to have on the platters back in the days. I was not aware there was actual contact between heads and platters back then!

 

[mg.man]

I was not aware there was actual contact between heads and platters back then!

Not when running, but when the old skool drives used to just spin down, leaving the head to "land" on the platter. When the drive has aged / the lubricant used to coat the platters has become gummy, "stiction" occurs.

View topic - Hard drive stiction