I'm ordering four more of the Sintech PCI cables and tackling this hack alongside my SuperIIsi™ Sub-Basement PDS hack, so it's time to be a bit less cryptic.
The puzzle was meant for Bunsen, who needs to get a Sound Card into the NuBus adapter of his 6100.
By moving the Accelerator, extending the PDS to the underside of the 6100 MoBo, all the PDS pins are available to the NuBus Adapter and the G3 card.
The reason for the 6100's NuBus adapter's finicky performance when it comes to G3 Cards is more than likely the same as the problem I identified with the SE/30 vs. IIsi PDS pinouts. I tested the IIsi NuBus Adapter stacked on top of of Straight up Hacked bare SuperMac Video Card RA PDS adapter out of curiosity when I'd already proved that the adapted adapter worked by running the Radius color Pivot II/IIsi card atop the poor critter.
When that didn't work, further research turned up a "NuBus" signal on the IIsi PDS that's missing on the SE/30 PDS and is unimplemented on the bare SuperMac Riser Card. So the IIsi NuBus adapter won't work in the SE/30's PDS without synthesizing the NuBus signal, as was obviously implemented the Expanse NuBus Chassis interface.
It seems highly probable that Sonnet didn't bother to implement this same/similar signal on the PDS Passthru of their Accelerator, as it was intended for a the AV/HPV Cards "normally" installed in the X100 PDS. It turns out that the NuBus Signal is likely present on the MAXpowr G3 PDS Passthru, but not the Sonnet Crescendo/NuBus G3.
The only curiosity is that the Crescendo/NuBus worked at all in this scenario, which leads me to believe the problem may be a more problematic combination of implementations. Address space stomping and the NuBus Card's PseudoSlot ID might be rearing their ugly heads in some strange combination.
So I'm checking it out, more later, or should I start a new thread about my proposed implementation of this interesting hack? :?:
The puzzle was meant for Bunsen, who needs to get a Sound Card into the NuBus adapter of his 6100.
By moving the Accelerator, extending the PDS to the underside of the 6100 MoBo, all the PDS pins are available to the NuBus Adapter and the G3 card.
The reason for the 6100's NuBus adapter's finicky performance when it comes to G3 Cards is more than likely the same as the problem I identified with the SE/30 vs. IIsi PDS pinouts. I tested the IIsi NuBus Adapter stacked on top of of Straight up Hacked bare SuperMac Video Card RA PDS adapter out of curiosity when I'd already proved that the adapted adapter worked by running the Radius color Pivot II/IIsi card atop the poor critter.
When that didn't work, further research turned up a "NuBus" signal on the IIsi PDS that's missing on the SE/30 PDS and is unimplemented on the bare SuperMac Riser Card. So the IIsi NuBus adapter won't work in the SE/30's PDS without synthesizing the NuBus signal, as was obviously implemented the Expanse NuBus Chassis interface.
It seems highly probable that Sonnet didn't bother to implement this same/similar signal on the PDS Passthru of their Accelerator, as it was intended for a the AV/HPV Cards "normally" installed in the X100 PDS. It turns out that the NuBus Signal is likely present on the MAXpowr G3 PDS Passthru, but not the Sonnet Crescendo/NuBus G3.
The only curiosity is that the Crescendo/NuBus worked at all in this scenario, which leads me to believe the problem may be a more problematic combination of implementations. Address space stomping and the NuBus Card's PseudoSlot ID might be rearing their ugly heads in some strange combination.
So I'm checking it out, more later, or should I start a new thread about my proposed implementation of this interesting hack? :?: