exxos's Atari, Amiga & retro forum

I managed to find time whilst the dinner was cooking to get out and run the tests that cryptodad suggested with the new version of ataritest and lowering the GPIO clock as he suggested just made things worse and worse so it looks like the 125MHz clock isn't enough at least for this version of the EPM240.

I wonder if there are speed grades and the one that the PCB manufacturer fitted are too slow.

The parts used on my boards were all as-per original PiStorm BOM.

Fair enough.

I'm now wondering if there's interference from the WiFi module. I think my next test will be to turn that off. Even though the Pi is close to the access point it doesn't seem to be getting a good connection anyway. I can hardwire the network.

Hello,
i am wondering if the issue could be related to different EPM240 version called Max II vs Max V. Search for EPM240 on the following git repo:
https://github.com/captain-amygdala/pistorm

I'm back from Somerset and had another play.

Firstly, I checked the EPM240 model to verify, and as Exxos said, it's the correct one.

I then tried just running the emulator at the GPIO clock frequency I was getting the best results from the ataritest program, namely 79MHz. As expected TOS tried to boot, crashed and tried to boot again.



Then I started tapping chips and wiggling things... nothing seemed to make any difference until...

Finally I pressed on the MMU and... The machine got to the desktop!

Managed to run Gembench.


Frontbench crashes and the machine is not particularly stable but it's mostly working.

I had to crank up the loopcycles in the config file just to get TOS 2.06 to boot. EmuTOS is far less stable.

I've only briefly looked at the pistorm code and don't know much about it, but I wondered at the time if core isolation would help, i.e. dedicating a single core just to the emulator. Although I imagine there's not much else running so it's probably not required.

"Frontbench crashes and the machine is not particularly stable but it's mostly working."

Congratulations on getting this far. :yay2:

Looking good so far :thumbup:

Does the diagnostic cartridge show up anything unusual as you are having weird issues?

exxos wrote: 09 May 2024 15:50 Looking good so far :thumbup:

Does the diagnostic cartridge show up anything unusual as you are having weird issues?

If I switch over to the 68000 the machine runs perfectly. Similarly if I swap the TF536 with the PiStorm.

Looking at the verilog file it seems that they're not fully emulating the bus cycle, making a few assumptions. With Gembench showing 101% ST RAM memory speed I have a suspicion that the bus timing is not right as the maximum speed should be determined by the MMU as that's controlling the bus cycle.