SE/30 Battery victim revived

[Bolle]

I thought I should share this as a motivation for others

Got this spare board off eBay for 5€ mainly for sourcing parts.

It started like that:



I went ahead and got all the gunk off and removed everything that got in the way while cleaning:



The CPU was toast, RTC and its two diodes were more or less gone, SIMM socket had corroded to the point of no return as well.

I started buzzing out a few connections just for fun and ended up doing the whole address and data busses. Identified roughly 20 broken traces.

A rainy day and half a roll of rework wire later:







It is far from pretty but boy I couldn't believe it actually chimed right on the first try.

Got to clean up a little now and see if I can get it to boot from SCSI. The SCSI chip seems to have gotten a nice load of cap goo and is probably gone. It tries to boot from HDD the way it is now but aborts and just flashes the LED on the HDD every few seconds.

I should have a replacement chip somewhere in the spare parts box.

To be continued...

 

[techknight]

Wow... you have put WAY more effort into that than I ever would have. 

I have a few SE/30 boards like that in my scrap bin due to this. I just pillage it for parts. 

 

[Bolle]

Haha, I don't know if this is a good thing or not

Not that I need yet another SE/30 board but this was more a "could it be done?" kind of thing.

 

[techknight]

Oh I support your endever. I personally woudlnt go that far, but anyways. take one for the team. 

your SCSI issue could be a couple things. 

Did you replace the IC and crystal that is missing? the OS may not like that missing. The other thing could be a broken trace to the SCSI IC whcih is more likely. one of the address/data lines, and I have seen /IOW and /IOR break between the SCSI and the GLU. 

 

[Bolle]

The RTC and the 32.768 KHz are still missing. This should not be an issue for the rest of the system. The only thing that won't work is the clock. It will just stand still. I tried that with another SE/30 already where I removed the RTC in an attempt to implement a replacement for it in a small MCU.

I might get this project out of the drawer again now that I am in actual need of an alternative to the real RTC.

 

[techknight]

But then you need a functional RTC to reverse engineer the communication protocol and packet structure. 

 

[Alex]

To be continued...

Awesome work, how on earth did you figure out how to connect the jumper cables and which ones were needed? Through a continuity test? Did you need to refer to schematics to get it right or for some hints? Good show!

 

[Alex]

By the way, how did it chime when there are some caps missing? Such as C12?

 

[Bolle]

It only needs the audio caps (C3-C6 area) to actually produce sound. The board itself will work without the other caps (might not be as stable as with them though)

I was referring to the schematics and checking data and address lines between CPU, RAM muxes, RAM, VRAM, FPU, PDS and ROM one by one.

It still is not booting successfully yet. SCSI does not seem to work at all for the moment and it crashes with Error 11 when booting any OS from floppy.

I have a feeling this could be the RTC (which also holds the PRAM) against my first thought. I will try another good board again and pull the RTC to see how it reacts.

Might then check if pulldowns on the open lines between VIA1 and RTC will do any good. I would imagine floating lines there might throw things off the track.

The RTC experiments have been a few years ago already... the clock standing still could as well have been with RTC in place but without the crystal - I am not sure anymore if I did pull the whole RTC back then and try without it.

 

[Bolle]

Alright...

Removed the RTC/PRAM on a good SE/30 board and it is working fine without it. So there has to be wrong something else with my battery victim.

Error 11 is "Miscellaneous Hardware error". Curious what that means :/

The error is showing up early in the boot process. If I hit the NMI switch and just enter "G" it continues to boot up until the Finder loads and throws error 11 again. This time hitting NMI results in "Could not handle interrupt" which leaves me only with a restart option.

Any clues?

 

[techknight]

yea your going to need to check more traces. Especially to the resistor packs. 

Also you need to buzz out the connections between the GLU and the rest of the bus, as well as both VIAs. those 3 can cause weird hardware problems. 

 

[ktkm]

Beautiful! This really brings hope! Nice work!

 

[Bolle]

We are getting closer. After having checked all the traces from the SCSI chip before I found a broken one when going through everything once again right now.

/IOW was open between SCSI pin 18 and GLU pin 68... I must have missed that one the first time.

I still get error 11 when booting anything from floppy... I noticed however that System 6 will boot from floppy. I was able to load Lido on a floppy and use it to check the SCSI bus. It sees the harddrive connected to it and will also mount the drive.

Copied a system folder to it and ran all the tests in Lido and all passed ok. So once the drive is mounted it seems to function just fine.

When I restart the machine the harddrive won't mount automatically and it will not boot from it. If I don't put in my System 6 boot floppy it is just sitting there flashing the LED on the drive every second (just like the Jaz drive in your video @techknight)

This is even more strange than what you encountered though, as I can mount and use the drive just fine once it has been mounted manually. Really weird...

Looks like it does not set up the SCSI chip correctly on hardware resets. Connections on the SCSI chip itself are all checked and ok now. Both VIAs also check out fine.

Will check all connections on the GLU now...

 

[techknight]

Well System 6 will boot and System 7 will not if the SCC is glitchy or unreadable. 

Also /IOW and /IOR lines are commonly broken on the SCSI IC ;-)

 

[Bolle]

Stupid me... It can boot System 7 now just fine. I installed a socket for the 030 after removing the original CPU for cleaning. The socket I used did only have one of the 4 inside corner pins. The 68030 datasheet says they are all VCC, so I guessed it would be ok to leave it like that for now... seems like I was wrong.

I modified the socket to have all of the 4 inside corner VCC pins and now the error 11 thing is gone and System 7 does boot as well.

SCSI still does not work correctly though. I can mount harddrives from inside the OS but booting from SCSI still does not work - still getting blinking activity light at the :?:  screen.

GLUE and SCC traces are all ok. I did find that my board seems to be missing R35 to pull up one of the serial data lines. But I don't think that's causing any issues here.

 

[Bolle]

Case closed... put on a replacement RTC and it is working just fine now.

The other board I took the RTC from is still working without it. How strange is that? One board works without RTC and the other one does not. :?:

Now to get the PDS connector back on and see if I can get PDS cards to work again.

 

[Alex]

Brilliant work! That is one heck of a revival! Good show. I bow to you sir.

 

[just.in.time]

Great work!