Simasimac findings `

[techknight]

This might already be general knowledge somewhere, but during my video circuit study of the SE/30, I had found something interesting.

I have performed these tests with NO ROM, No RAM, no Video ROM. Any board with the TI video RAM produced zebra stripes. Any board that had the NEC VRAM produced a weird crosshatch/herringbone style pattern. but they were all the SAME on all the NEC boards. just like the zebra stripes on the TI boards.

And this is independent of the PALs, because I can swap them between boards and it still stays the same. And the GLU is the same on both types of boards.

so maybe the TI chips have a special built in chip function procedure to define when its blank? not sure. There is no ROM present so its not running code to fill the SRAM this way. its purely generated by the video circuit, and the video circuit by itself and it not being initialized.

Not sure if this is going to help anyone out for anything, but maybe for troubleshooting reasons. So if a Simasimac has become terminal, it probably has the TI video chips and its because the system hasnt booted far enough to init the video circuit.

What brought this up, is JDW sent me his 2 terminal Simasimac boards. both of them have bad bourns network filters. I had removed the filters but the issue still remains. So i was testing this against working boards with no RAM and ROM present. This is how i discovered this.

Edit: I guess various things could cause a system not to boot all the way to initialize video. Maybe stuck address/data lines, failed address decodes, bad or missing replies from other hardware, and unexpected interrupts to name a few.

 

[techknight]

I think I found the problem with one of his boards. the Column address strobe for SIMM1 of bank A and bank B is broken connection from the resistor pack.

 

[techknight]

ITS ALIVE.... Well sorta.

It works if i fill both banks, with the circuit repair on the CASLL connection. But it will simasimac if i use either Bank A or Bank B. I have to use both banks.

So something still broke somewhere.

Edit: I booted it all the way up into my Jaz Sys 7.1 install, and it shows all 8mb i stuck in both banks, and it didnt crash.

Hey bbraun, this might be a question you could answer. How exactly does the macintosh RAM test startup routine in ROM work? does it test every cell in every SIMM?

 

[techknight]

cleaned the SIMM sockets and all is well. It now posts with just bank A ram installed like it should.

So board #1 is functional. Just have to find a new RP2 or build one.

 

[techknight]

finally got a video uploaded on the first of the two boards