exxos blog - random goings on

[exxos]

Resoldered both headers and sockets for bits 8-15.. made no odds

Random idea.. This H5 had a IMP MMU.. So replaced that out...

IMG_4174.JPG (314.93 KiB) Viewed 315 times

IMG_4175.JPG (194.25 KiB) Viewed 315 times

Only now bloody works !!!


Put back the IMP MMU...

IMG_4173.JPG (330.9 KiB) Viewed 315 times

IMG_4172.JPG (547.36 KiB) Viewed 315 times

I've checked the pins on the IMP and swapped them again.. Same issues... So IMP MMU really doesn't like my DRAM tester board for some reason

Why is it always me who ends up with these stupid problems all the time

[Badwolf]

Why is it always me who ends up with these stupid problems all the time

Can't drive (what I assume is) a massive additional capacitance? Or is the test board buffered, in which case is the IMP chip unhappy with the buffered timings?

BW

[exxos]

Can't drive (what I assume is) a massive additional capacitance? Or is the test board buffered, in which case is the IMP chip unhappy with the buffered timings?

There is no buffering on the test board..

IMP tend to be more noisy.. So I just guess that is part of the problem.. All the extra inductance and capacitance adds up to the point where the thing becomes unstable... But this doesn't happen with the Rioch MMU... But they tend to be a little more "quieter".

I suppose a theoretical fix would be to have some larger in series resistance with the RAM signals on the board.. But not really worth spending the time on it..

[exxos]

So my H5 B1 seems to have a "unsolvable" RAM fault

Capture.PNG (48.35 KiB) Viewed 258 times

IMG_4198.JPG (202.95 KiB) Viewed 258 times

IMG_4199.JPG (371.08 KiB) Viewed 258 times

Data bus seems to all switch low and give correct ROM start.

IMG_4197.JPG (345.83 KiB) Viewed 258 times

Changed the MMU, CPU, GLUE..SIMM. Resoldered the simm socket, RAM resistors, MMU, GLUE, SHIFTER..

The Only time anything changed was when I resoldered the buffer chips.. Before it would randomly change pattern moving my hand over the motherboard. Since resoldering , its stopped doing that.

DRD5

IMG_4201.JPG (385.87 KiB) Viewed 258 times

IMG_4202.JPG (261.52 KiB) Viewed 258 times

DRD7 (a couple others look the same)

IMG_4203.JPG (257.71 KiB) Viewed 258 times

So may have to change both the chips.. which is annoying.. as I don't think I have any spares

Capture.PNG (90.75 KiB) Viewed 258 times

EDIT:

Looks like the pattern just changed now the screen flashing red rather than the lines

Capture.PNG (37.64 KiB) Viewed 258 times

[exxos]

The plot thickens because I get different results on either end of that highlighted track...

Capture.PNG (59.1 KiB) Viewed 248 times

IMG_4206.JPG (292.23 KiB) Viewed 248 times

IMG_4207.JPG (249.82 KiB) Viewed 248 times

[exxos]

The track on the bottom of the motherboard continuity checks fine but it does not continuity check on the top of the board where the IC is soldered down...

I suspect a faulty via !

Capture.PNG (51.28 KiB) Viewed 247 times

Bodge wire time..

IMG_4208.JPG (454.25 KiB) Viewed 247 times

TaaaDaaaaa !

IMG_4209.JPG (226.69 KiB) Viewed 247 times

[stephen_usher]

Who makes your PCBs? They obviously don’t do proper panel testing

[exxos]

Who makes your PCBs? They obviously don’t do proper panel testing

I've used about four different places so far... But that board had been working previously...

I am going to investigate who made that board though... But on later boards I started making the vias larger,

EDIT:

JLC made those boards.. I don't use JLC for them anymore anyway. To many issues all the time.

[exxos]

Now I have the floppy drive working time to test some CPUs at 16MHz for the store

IMG_4210.JPG (365.66 KiB) Viewed 230 times

IMG_4211.JPG (310.15 KiB) Viewed 230 times

[exxos]

IMG_4213.JPG (340.09 KiB) Viewed 249 times

Now seems to have developed into more weird faults

Capture.PNG (8.2 KiB) Viewed 245 times