exxos's Atari, Amiga & retro forum

PIn 22 does have a 1K pull up already...

IRQ has a pull up..



Whats probably confusing things is some of the single resistors I have moved to a resistor pack somewhere instead. It was done to reduce the amount of parts/soldering/space.

I would suggest actually removing those resistors you added, possible the MFP can't work with 500R (10mA) loads..

Best way to check pins is to measure resistance with your meter from 5V to the pins and make sure you got the values there.. or some resistance. "in circuit" may effect readings, so worth checking probes both ways...

One thing to look at, is scope out the 4MHz to MMU, and E-clock to ACIA... see if there is any bad ringing on the signals...

Really need the keyboard or rs232 working so we can control diagrom.. it does have a video test which displays a image..

The video issue kinda looks like the bitplanes are getting shifted.. like some odd thing where the shifter is thinking its in monot but outputting 4 bit planes in a single mono bitplane screen.. no idea even that's even possible though...

EDIT

It does look like the bitplanes are correct initally, then the mouse gets some corruption next to it like a shifted bit-plane... but goes to quick to really tell much..


Might be worth checking UDS, LDS on the MMU with scope and make sure they are switching...

ahhhhh hold up..... Something not looking right here...

exxos wrote: 24 Sep 2018 09:29 PIn 22 does have a 1K pull up already...

Not Pin 22 - it was Pin 32 on the MFP. I measured for a pull-up on it on the STF and there was none. On the STFM there is one.

EDIT: What is probably confusing is I was reading it as the IRQ of the MFP not the IRQ from the CPU. Its actually the MFPINT line into the MFP.

Great job, Icky, for this fantastic progress!

troed wrote: 24 Sep 2018 06:54Note on RAM access: The Shifter might have glitchy RAM access without that affecting the CPU RAM access. The latter obviously is fine else we wouldn't see Desktop.

You are probably right. However, there could be another cause: if lower half of RAM was good, and upper half of the RAM was bad (only used by screen buffer), then we would get the same result.

Icky wrote: 24 Sep 2018 12:32 EDIT: What is probably confusing is I was reading it as the IRQ of the MFP not the IRQ from the CPU. Its actually the MFPINT line into the MFP.

Where is the resistor on the schematic ? I looked on ST and STFM and just see MPFINT going from MFP to GLUE, I don't see a resistor anywhere ?

Oh its R105 on the ST schematic.. that one does seem to be missing.. not sure why :roll:

exxos wrote: 24 Sep 2018 12:48 Oh its R105 on the ST schematic.. that one does seem to be missing.. not sure why :roll:

I'll take a photo of the fix later tonight for the Alpha Modifications. Will also do the GLUE Dataline Fix as well and post pics.

Looks like its very confusing with all these IRQ related lines..

Seems I wired MFP pin 26 (IRQ) to pin 2 on RP1 (STF schematics) which is correct in needing a pull up, and on RP1, BUT there is also INTR on pin 32 MFP, which should have gone to RP1 pin 2... So seems I must have confused the 2 lines, then thought I had done them , so I have confused IRQ with INTR :roll:

So yes, pin 32 of the MFP does indeed need 1K pull up..