//e fun

[Jackson]

Snagged me a //e without disk drives but a fully functioning monitor and a line printer for $35. This appears to be the B revision from 1983, as I checked on the motherboard and the underside of the computer. Currently, the power supply is dead; using a working supply brings me to finding out that while the lights turn on, there's no beep. Haven't hooked up the monitor to the computer yet.

Will post pictures later.

 

[Jackson]

Nevermind the pictures, let's talk about the repair journey, in which we unfortunately reach square one.

So, the white picture? Turns out it's repeating, vertical blocky stripes. I get no characters at all, as if it was seemingly stuck in full low resolution mode. I go to blow the board with my air cans. Nothing. Tried repeatedly flicking the power. Nothing. Then suddenly, flickering lines appear and I reset the computer. To my surprise, while it didn't beep, I could send characters, indicating by how every key I pressed made the screen freak out even more, and press the Ctrl+Reset key. Which also made the screen heavily freak out. 6502 must be good. So I take out all chips to test everything that could be possibly done with my TL866. Everything's good.

Now I wanted to see if it worked with a 65C02. Accidentally reversed it backwards. Vertical stripes again, but no flickering lines. Nothing else, and I repeatedly flick until I realize.... $@*$?!

Amazingly, it didn't seem to get hot, blow the magic smoke or make a giant blown fuse. I correctly rotate the chip and cross my fingers. Got the flickering lines again and now random blocks flying across my monitor. Tried to press everything while the board was on, which was a somewhat stupid decision considering I could get shocked or something could no longer work. And thus the screen froze on the next reset, with no more lines; but stripes still persisted. Ctrl+Reset no longer did anything. Wondering if the 65C02 had it's last laugh; I was amazed that it never once made a snap, crackle or pop.

Put back in the 6502B. Silly stripes again, but also the silly rapid flickering lines again. It's alive, it seems. Ctrl+Reset made funky screen effects again. I could feel the MMU getting hot when I pressed Ctrl+Solid Apple+Reset. Good. Reseat memory chips. Nothing.

Swap in RAM chips from a working //e language card. Nothing but silly stripes and silly lines. Aaargh! Touch and reseat chips, once again, being stupid while the machine was on. Now we only have silly stripes. Then the 6502 acted like the silly dead(?) 65C02. In which I can now only get stripes and no flickering lines and doing Ctrl+Reset or any other reset mode doesn't assist in letting the machine freak out. The lines to me were an indicator that the machine was at least trying to operate; now it isn't.

Now I need to test all the logic chips again. Chances are that by reverse plugging the 65C02 and flicking on the //e as many times as I could that I reduced the lifespan of some by sending the wrong signals. But as far as I know, running the test on the 244 chips near the 6502 show that they're still good, even after I goofed things up.

After all I've been through, it still doesn't solve the problem with the low resolution mode being seen instead of all text though.

 

[David_M]

I hate to break it to you, but there is nothing on the motherboard that will give you a shock. But you are correct in one thing, plugging the cpu in backward probably caused more damage.

 

[Jackson]

Holy $#@(! The 65C02 still works in my working //e, even after plugging it in backwards! It's a miracle!!

I also swapped in the HAL and the IOU chips from the defective //e before I attempted to put in the 65C02. They appear to be fine as well; will have to swap again to see if the chips I brought from my working //e got fried when I plugged the CPU backwards. Someone told me this is because the grounds on the 65C02 pinout are symmetric, so only pin D5 was affected by a Vcc current. Now to figure out what chips specifically are attached to D5; looks like the MMU and the ROMs might face issues.

I was looking forward to replacing every related chip to diagnose the problem further, so this doesn't make me panic much.

 

[Jackson]

OK, I tested the RAM I pulled from the defective //e and put it on an auxiliary card. After doing the diagnostics on a working //e to analyze the RAM on a working language card, I came to the conclusion that all the RAM that I thought was defective... was perfectly fine.

So if it wasn't the IOU, it wasn't the RAM, it wasn't the logic chips or it wasn't the HAL that stopped the board from functioning properly, then the only thing left seems to be... the ROMs, the character generator and the MMU.

 

[Jackson]

Leaving the MMU still leaves chunky blocks instead of text characters. I think the issue is more likely to do with improper contact in sockets-- will have to figure out what to do by washing the motherboard.

In the process, I also broke resistor R73. Yay!

 

[David_M]

Leaving the MMU still leaves chunky blocks instead of text characters. I think the issue is more likely to do with improper contact in sockets-- will have to figure out what to do by washing the motherboard.

In the process, I also broke resistor R73. Yay!

Have you considered that it might be one of the chips that are soldered into the board. I generally isn't a socketed chip.

 

[Dwight Elvey]

Holy $#@(! The 65C02 still works in my working //e, even after plugging it in backwards! It's a miracle!!

The one thing that MOS did that at first I thought was really strange was to put the power and ground on the same side of the IC. Having plugged a 6502 in backwards once my self and not destroyed it, I thought back to the way the power leads were attached. Maybe, this was by design intent. Maybe the fellow that laid out the chip had blown one or two parts with power and ground on diagonals. He might have realized that is was a design for disaster and changed the normal.
From a design perspective, the best place for power and ground would actually be the centers of the part. Why it was done at the corners in the first place was not the optimal.
Dwight

 

[Jackson]

No chips are soldered-- everything on this board is socketed. Also, a realization; while power was applied to D5, where D5 would actually be on the motherboard would not be affected. So no chips look to be damaged much, which all concludes to...

Bad traces. Might get the motherboard sent in to someone who has a logic analyzer when washing doesn't work, but in the meantime, R73 is a 47 ohm resistor. Will have to look for spares or get a straight up order from Mouser.