Board repair help needed with a faulty Mac 512K

[TimR]

I originally posed this question a little over 12 months ago

https://tinkerdifferent.com/threads/help-needed-with-dead-mac-512k-board.499/#post-2954

and received many helpful replies.

I've recently got around to removing the 18 DRAM chips and socketing them. When I chip-tested the 41256 DRAMs 4 of them failed test so I was excited that my 512K
Mac would soon live again.

Having added sockets and fixed any bad traces, and then tested 18 chips before inserting them, I still have the same problem with the board, see attached photo.

To add some more context, here are things I know for sure about the faulty board, as I have another working 512K Mac to swap things around with.

The analog board and power supply are good, the 2nd 512K Mac board works perfectly.
The ROMs in the faulty board are good. They work fine in the good 512K Mac board.
The RAM chips in the newly-socketed faulty board are all good, I've tested and retested them twice - original fault from 12 months ago still appears.
I've checked the couple of trace repairs I made against the working board with a multimeter.

Someone previously asked about the reset buttons being stuck, I've tested the good board and faulty board buttons with a multimeter and they behave the same.

Since I posted the original request for help, I now own a scope, although it's not had much use yet, and have watched lots of Adrian's Digital Basement YouTube videos about checking signals on buses, so this is my next point of call.

I've got a capacitor meter and multimeter to check resistance values in circuit against the good board and duff board.

Is it possible the 68000 died? Did anyone else have a bad one? Can I use any 68000 chip from eBay? ex-Amiga ones seem quite common. Guessing I should socket it too?
I guess I can use the scope and diagnostic guides on YouTube to test this?

Although I removed 4x 41256 dram from the soldered in chips, that later tested faulty, this scould have been caused by the hot air gun. I still get the same box pattern at powerup. If I remove one bank of Dram chips the pattern changes to every other box in the pattern. The chips I'm using now are 8x41256-100 and 8x41256-120 in the other bank. Could this make it fail post?

Startup-up circuit has been suggested as a possible failure point previously by some. Any pointers here? I know nothing about that. Bomarc did circuit diagrams I think?

Thanks in advance for any help.

Tim.

 

[TimR]

To anyone else following this item. I've now made some more observations using a scope and multimeter.

The -5V is being correctly generated by the 7905CT voltage regulator on the 512K board. The +5V, +12V and -12V are all correct too.

The clock signal is around 16MHz at pin 1 of the LS04 and TSG and just under 8MHz on pin 15 of the 68000 processor. The 5V and GND on the 68000 are correct. If I turn the Mac off and on the Reset line on pin 18 goes low and then high. The Halt pin 17 stays low.

The address and data lines on the 68000 show signals, but don't seem consistent. Next job is to socket the processor and see if changing it does any good.

The Ram chips have been tested in an external tester and are all good. The ROMs work fine in another 512K board.

 

[Paralel]

To me, it does sound like the processor is the issue. Odd, those old 68000 processors are usually fairly tough, in my experience.

 

[TimR]

Thanks for your reply. I have some 64 pin 2.54mm pitch IC sockets coming from Mouser and have ordered a couple of old 68000 processors from eBay. Hopefully it will live again. If that doesn't work it's looking for shorted components and checking traces - at least I have a working and dead 512K motherboard to compare. I suspect that when I originally changed 1N5234B diodes on the analog board for replacements and couldn't get more than +4V output from the pot on the analog board, that the 12v rail was doing something all together different, and that caused an issue - I've seen components from other places that aren't the correct values as advertised, hoping this is more of this... I read about the same thing in the last hour about 12v feed being too high and frying things elsewhere- this 512k uses 12v to get -5v. Next time i'll check the analogue output with a multimeter before attaching the mainboard, and keep the processor, ROMs and RAM out when I do, initially. Upside of this board is it's taught me so much, probably still a lot more.

 

[TimR]

Made some more progress with this 512K Mac today. I received 2x 68000 processors from eBay and 64 pin DIP socket from Mouser. Today took out the suspect 68000 which only produced screen bars. Soldered in socket and tried a 68000 processor. Now get a raspy bong noise and what looks like a 020800 error. Tried the 2nd 68000 processor and get the same error. So both new 68000 Processors are okay, the ROMs are tested in another 512K, and the RAM checks okay in a tester. So now at least it's alive again and talking to the ROMs.

Things that are now different from when it originally worked correctly:

1- I've now got 120ns RAM in the top bank and 200ns RAM in the bottom bank. Originally it was all 200ns. (4 of my 200ns originals failed test)

2 - I've socketed the RAM, and had to fix some traces. There may be other trace issues I caused when I did this...

I've got a 2nd working 512K system I can test against. I've got a scope. Any ideas anyone?

 

[TimR]

Pleased to report that the 512K Mac is fully working again. I fitted 16 matching 41256-15 chips from eBay and it worked straight away. So they either don't like 120ns chips or/and mixed speed chips. In the end I had a dead 68000 processor and 4 dead RAM chips. I now have a 512K Mac with socketed RAM and processor.