search

hoglet67 / PiTubeDirect

Bare-metal Raspberry Pi project that attaches to the Acorn TUBE interface and emulates many BBC Micro Co Processors
GNU General Public License v3.0
188 stars 23 forks source link

6502 3 MHz issue in Gecko #158

Closed mcnalu closed 1 year ago

mcnalu commented 2 years ago

The symptoms of this issue are identical to those described in #66 in that after FX 151,230,1 there is no cursor after a ctrl-break and no basic prompt after a break. All other co-pros seem to work just fine, including the default FX 151,230,0.

I used the latest stable Gecko_fix1 and am using the level shifter board I just received from Retroclinic/Mark with my BBC Model B rev3 with TurboMMC and DNFS ROMs installed. The pi is connected only with a 40pin IDC cable and I checked and reseated all connections and the tube pins on the BBC are in near pristine condition.

I found the same problem with both my Pi 3B+ and Pi 2 and repeated ctrl-breaks never got it to work.

I then tried it with Hognose-rc2 and the problem was gone! Everything I've tried seems to work perfectly in Hognose-rc2 in fact.

hoglet67 commented 2 years ago

Please can you post a couple of photos of your setup, to save me asking loads of questions.

mcnalu commented 2 years ago

Voilà

20220116_134859 20220116_134936

RetroClinic commented 2 years ago

Not sure if it's related, but I have had issues in the past, on the Elite attractor screen, where the occasional line was not removed after the next frame being drawn, leading to a build up of random lines on the screen. This has happened in Gecko sometimes as well, and under HN RC1. The fix before was to change the CPU in the beeb, as it seemed like it was a very slight timing issue with the PHI 1 Pin 3 output. The problem only occurred on the Model B, not the Master.

However, HN RC2 seems to have cleared it, and the CPUs that were causing the problem before are now working perfectly. Was this a known issue beforehand?

hoglet67 commented 2 years ago

Was this a known issue beforehand?

Not, I've not seen that before Mark.

RetroClinic commented 2 years ago

Not, I've not seen that before Mark.

Ok, it didn't happen often, and before I just assumed it was bad timing off one particular CPU, as a swap would fix it. Next time I get a chance, I'll try and replicate it with RC1 and do a video, but as I said, it's fixed in RC2, so you've done something to sort it!

hoglet67 commented 1 year ago

Whatever caused this seems to be fixed in hognose so closing.