dml wrote: Thu Aug 28, 2025 8:49 am
I didn't quite follow the last part - did you somehow bypass the DSP to get it to boot and the DSP is still suspect? Or did you do something that resolved the DSP problem? Interesting to see if the DSP is still functioning or not.
I haven't touched the DSP at all. It's failing the cart tests, but I haven't tried any actual DSP test software yet. One of the legs is completely desoldered and one doesn't seem to have a pad at all, so I suspect it's very unlikely to be working.
The machine will boot nine times out of ten from cold, but only one time out of ten from a warm reboot. I wonder if that's related to the DSP's state?
Last night I was mostly working on gettin mass storage onto it. I was getting mixed results with a BlueSCSI image I set up on Hatari, but once I took the Gotek out of my working Falcon and bootstrapped it on the new machine itself, things seem to be working. This is a nice surprise given the bus arbitration scheme is presently nothing more than a resistor connecting XBR and BG030! SDMA seems healthy, though.
I should now be able to load up some test software for various things too. At present I don't know anything about audio, Midi, printer, LAN etc. I'll probably not test them all, but some basics would be nice and a DSP test for when I get onto that will be vital.
I did flick DFB1r4 into 40MHz mode and it worked, which was a nice bonus and once again raised my eyebrows at the BG hack.
As an aside - I wonder how many of these machines are still around, in working condition? And how many of those are still in more or less factory condition with no boosters, mods and stuff attached? And how many of the dead ones died in a modding accident?
My '93 Falcon has suprisingly few mods. A clock patch went in in 2020 and I had to replace the FPU socket. The PSU and floppy have been taken out (and is stored), but otherwise the motherboard is stock. DFB1 was always planned as solderless as I was scared of mod death.
There's been plenty of abuse of the case, mind!
the third machine required extensive surgery, trace and pad rework to bring it back to life from its status as a dead board (I think it was upgrade-killed) and I'm still not 100% sure if every aspect is fully working yet. It seems to work at 16MHz but gets upset if I insert a known working DFB1 - so maybe signal integrity issues on the PCB or a problem with the expansion traces.
Anyway, looking forward to your next and I guess last vid in this mini series
Cheers! I'm editing part two at the moment, so whatever I'm doing now will be in a third part at least. Hopefully that'll be enough!
BW
