I finally got the accelerator working on his minimal configuration installed on a A600 rev 1. It took me several tries and I went down down several troubleshooting red herrings before acquiring a cheap oscilloscope that helped me toubleshoot what was going on.
I used LemaruX plcc to dip64 adapter to bridge the accelerator to the main board https://github.com/LemaruX/Amiga-A600-68k-PLCC2DIP-v2
Here's some photos of the setup and sysinfo
Sysinfo reports 18mhz but my cheap oscilloscope says the clock is at 14.20mhz
Please take a look and let me know if you think that these values are in line with the accelerator expectrd performance. 1.10 performance increase seems suspiciously low, but I confirmed that value removing the accelerator and the values reverted to 1.0 and 7Mhz, so the accelerator is surely doing something. Also, Frontier ran much smoother according to my eyes.
It was really difficult to tune the clock to a frequency stable enough to make the A600 initialize or complete the workbench boot. The diagROM ran no matter how unstable the clock, while the kickstart rom needed some fine tuning of the clock with the oscilloscope before even attempting to properly boot and read from the IDE device.
As you can see I installed a variable resistor in place of R3. Don't know if that was the cause oor the solution of my woes.
In my attempts, I tried using the CDAC signal directly soldering a wire from the motherboard to the proper accelerator input (and changing the jumper accordingly). I did not see any substantial difference in the stability of the input or the output.
A couple of notes:
Using version v2 of the adapter prevents the A600 case to close down. Version V1 of the adapter positions the accelerator in a better spot, closer to the keyboard connector. The problem using that version is that V1 lacks the right screw-mount to clamp down the "vampire" socket tight enough against the onboard 68000 chip. I need to modify the adapter project and make it equally as long as the V1 version, but I also need to add back the 4 screw-mount needed to clamp it down to the HDD motherboard cutout.
I don't plan to have the ram installed right away as I want to test it further adding 4mb fastram through a PCMCIA ram card (still to be built). Alternatively I could fork your project and build the PLCC adapter directly to the accelerator board, but that may be a little too difficult considering my limited capabilities on circuit and PCB design.
Hello Mathesar. Thank you for this great project.
I finally got the accelerator working on his minimal configuration installed on a A600 rev 1. It took me several tries and I went down down several troubleshooting red herrings before acquiring a cheap oscilloscope that helped me toubleshoot what was going on.
I used LemaruX plcc to dip64 adapter to bridge the accelerator to the main board https://github.com/LemaruX/Amiga-A600-68k-PLCC2DIP-v2
Here's some photos of the setup and sysinfo
Sysinfo reports 18mhz but my cheap oscilloscope says the clock is at 14.20mhz Please take a look and let me know if you think that these values are in line with the accelerator expectrd performance. 1.10 performance increase seems suspiciously low, but I confirmed that value removing the accelerator and the values reverted to 1.0 and 7Mhz, so the accelerator is surely doing something. Also, Frontier ran much smoother according to my eyes.
It was really difficult to tune the clock to a frequency stable enough to make the A600 initialize or complete the workbench boot. The diagROM ran no matter how unstable the clock, while the kickstart rom needed some fine tuning of the clock with the oscilloscope before even attempting to properly boot and read from the IDE device. As you can see I installed a variable resistor in place of R3. Don't know if that was the cause oor the solution of my woes.
In my attempts, I tried using the CDAC signal directly soldering a wire from the motherboard to the proper accelerator input (and changing the jumper accordingly). I did not see any substantial difference in the stability of the input or the output.
A couple of notes:
Using version v2 of the adapter prevents the A600 case to close down. Version V1 of the adapter positions the accelerator in a better spot, closer to the keyboard connector. The problem using that version is that V1 lacks the right screw-mount to clamp down the "vampire" socket tight enough against the onboard 68000 chip. I need to modify the adapter project and make it equally as long as the V1 version, but I also need to add back the 4 screw-mount needed to clamp it down to the HDD motherboard cutout.
I don't plan to have the ram installed right away as I want to test it further adding 4mb fastram through a PCMCIA ram card (still to be built). Alternatively I could fork your project and build the PLCC adapter directly to the accelerator board, but that may be a little too difficult considering my limited capabilities on circuit and PCB design.
Keep up the good work!