another simasimac

[NE555]

I recapped the SE/30 logic board. then upon start up , there is no bong, no error chime. just a Simasimac screen. I have been trying to trace things out and found a whole bunch of vias without solder in them.

Been trying to fill them with solder and a piece of wire. So far the Simasimac persists. I keep finding more and more empty vias. I am not sure if what I am doing is  actually helping in any way.

All the rails seem to get the proper voltages. LS166 has been replaced already.

what's next?

 

[Bolle]

Check the traces between ROM and CPU.

No chime usually means you have a problem there.

Vias do not always come with solder in them. As long as you did a good job cleaning everything and there is no goo left inside the vias you can leave them that way. You want to check the corresponding trace though to see if the via is actually still connecting both layers. 

 

[NE555]

thank you for that info. 

 

[NE555]

I wished someone would make a youtube video of that process

 

[Bolle]

That’s going to be a pretty boring video.

In short you get out your multimeter, set it to continuity mode and probe every single signal between ROM slot and wherever the signal is supposed to go (CPU for the address and data lines and GLU for the select and enable lines)

Repair with a patch wire in case something does not check out. Check if the ROM SIMM is seated correctly and all springs in the socket as well as all contacts on the SIMM itself are clean and are not corroded prior to doing any trace checking.

The SE/30 has a notorious problems with proper contact at its ROM socket.

 

[NE555]

It can't be more boring than some of the videos that are on youtube at the moment. I can't find any traces that are OL. But I am getting the feeling that even if a trace might not be bad It might have a resistance that makes it useless?
As I said I would like to see a video like this just to be able to compare to a baseline. Maybe someone testing the board with an oscilloscope even.
check between CPU and anything else sounds extremely generic. How many have actually fixed a Simasimac SE/30? And I don't mean one that just needed C6 and C7 replaced

 

[NE555]

all data lines from the floating processor to ROM are good. My CPU is not socketed

 

[Bolle]

You actually want to test between the actual CPU and the ROM socket because of the way the traces are run on the board.

They go from CPU passing by the ROM, then the FPU and end up at the RAM slots.

The point of failure lies between the CPU and ROM SIMM (either from the cap that’s sitting there or the battery if it leaks)

If you test between FPU and ROM it is more likely you are not detecting a possible failure.

I have been fixing a lot of boards with several issues (simasimac included) and most of them needed a lot of poking around to find the issue.

Any resistance in traces should not matter. Most multimeters will cut out at a certain resistance when set to continuity mode.

That cutout level should be low enough to not become a problem on those old boards and the switching speed and driving power of those old components.

 

[NE555]

I had some partial success. I found a dateline that has been OL and when I traced it out I found out that it ended up in a via by C12. I was able to reconnect that Data line back to the CPU. 
and now I get some distorted boot chime and a variation of Simasimac. I guess I'll have to probe the address lines next.
Thank you for following up

 

[NE555]

@Bolle Address lines 23 to 31 where do they go from the CPU? I traced 25-31 to UJ6