PM8500 - av video out
- Thread starter lobust
- Start date
Yes, it shows as a second monitor like any other.
The 8500/8600 can sense when a composite or s-video display is connected.
The DingusPPC emulator does emulate this behaviour.
The bits are in a board register accessed at address F301E000. It's a 16-bit little-endian register.
Read the register in Open Firmware like this:
Composite video is 0x0080 for not connected. 0 for connected.
S-Video is 0x0040 for not connected. 0 for connected.
Unused bits are pulled high: FF3F.
We should be able to see which is stuck at connected.
I don't know if there's any Power Mac 8500/7500/7600/8600 schematics that can help you find the related components on the motherboard.
The DingusPPC emulator does emulate this behaviour.
The bits are in a board register accessed at address F301E000. It's a 16-bit little-endian register.
Read the register in Open Firmware like this:
Code:
F301E000 rw@ 4 u.rComposite video is 0x0080 for not connected. 0 for connected.
S-Video is 0x0040 for not connected. 0 for connected.
Unused bits are pulled high: FF3F.
We should be able to see which is stuck at connected.
I don't know if there's any Power Mac 8500/7500/7600/8600 schematics that can help you find the related components on the motherboard.
My apologies, I somehow missed your reply after I posted this thread.
So, I got into open firmware (after some messing about - I didn't realise these old world machines defaulted to serial io...) and did as you suggested.
The response is "FF7F"
Additional pertinent information - the a/v breakout board is no longer plugged in to the logic board. I noticed while troubleshooting that most of the ribbon cable conductors were broken at the solder joints (not impressed with the construction/soldering on this little board
), so I have completely removed it for now. The behaviour is unchanged though - it still thinks there is a second monitor connected.
FF7F means there's a connection to the composite video output. This needs to be fixed.
I don't have Power Mac 7500/8500/8600 schematics but here you can find Power Mac 9500 schematics:
https://www.macdat.net/repair/apple_schematics.html
From GrandCentral on sheet 15, there's a pin 172 DEV5CS pin. It will probably lead to a 47 Ω resistor and then to a 74F244. The 74F244 will have an input for the composite video detection and the S-Video detection.
In the Power Mac 9600 schematic this appears to go to U46. The composite video detection would be pin 17. It needs to be pulled high when there's no composite video connected.
===================
Power Mac 9600 schematics shows IOBus device #5 (selected by DEV5CS) going to U46 which has PSRNT bits for the second PCI bus slots and it has audio video pull ups which I think represent the Composite and S-Video connection detection pins that exist in the 8600 and the DingusPPC emulation.
However, for the PRSNT bits we have these connected to IOBus device #3 in DingusPPC which is shown as no connection in the Power Mac 9600 schematics. We did this because Open Firmware reads bandit2 PRSNT bits from IOBus Device #3 (F301C000). Perhaps this is a bug in Open Firmware? I suppose one could test this by moving a PCI card that has non-zero
For a Power Mac 8600, IOBus device #3 is the sixty6 video output device.
It looks like a copy/paste error where
The error was corrected in the ANS ROMs 1.1.20.1, 1.1.22, 2.26B6, 2.26NT.
But
i.e.
In the ANS ROMs,
I don't have Power Mac 7500/8500/8600 schematics but here you can find Power Mac 9500 schematics:
https://www.macdat.net/repair/apple_schematics.html
From GrandCentral on sheet 15, there's a pin 172 DEV5CS pin. It will probably lead to a 47 Ω resistor and then to a 74F244. The 74F244 will have an input for the composite video detection and the S-Video detection.
In the Power Mac 9600 schematic this appears to go to U46. The composite video detection would be pin 17. It needs to be pulled high when there's no composite video connected.
===================
Power Mac 9600 schematics shows IOBus device #5 (selected by DEV5CS) going to U46 which has PSRNT bits for the second PCI bus slots and it has audio video pull ups which I think represent the Composite and S-Video connection detection pins that exist in the 8600 and the DingusPPC emulation.
However, for the PRSNT bits we have these connected to IOBus device #3 in DingusPPC which is shown as no connection in the Power Mac 9600 schematics. We did this because Open Firmware reads bandit2 PRSNT bits from IOBus Device #3 (F301C000). Perhaps this is a bug in Open Firmware? I suppose one could test this by moving a PCI card that has non-zero
power-consumption in slot A-C to see if it has the same power-consumption in slot D-F.For a Power Mac 8600, IOBus device #3 is the sixty6 video output device.
It looks like a copy/paste error where
install-bandit2 was copied from install-kaos.The error was corrected in the ANS ROMs 1.1.20.1, 1.1.22, 2.26B6, 2.26NT.
But
install-kaos may also be wrong since IOBus device #3 is sixty6 video output and byte 0 of that is CLUT_DATA.i.e.
install-kaos may have been copied from install-bandit2.In the ANS ROMs,
install-kaos is unchanged - it still tries to read PRSNT bits from F301C000 (CLUT_DATA).
Last edited:
Similar threads
