A IIGS owner had asked for help on FB and I offered to help thinking it may be something simple they just weren't able to do themselves. Turns out my thinking was flawed!! This board has a problem and I'm running out of ideas and could use some help. The system came with a REMicro ROM01 chip and this is a ROM01 board.
The intial problem was the computer boots (with nohting connected) to the check startup device screen. This is where things start to go south... the alterntate chararacter set characters do not appear correctly. For example this is how the screen looks:
So that seems clear the alternate characters map isn't getting used correctly.
Running the system test restuls in an error of:
System Bad: 03011500
Appears to be a problem with bit0 of the soft switch at $C015. But this is where things start to get confusing...at least for me.
If I break the startup so I could get to a prompt there's a bunch of crap being "typed" in.
If I attache a 5.25 drive to the drive port and boot computer inspector boots, gets to the main menu then starts responding to key presses. But there's no ADB nor //e keyboard connected. Because there's no connected keyboard but keyboard data I suspect the ADB micro or more likley ADB GLU could be pumping data onto the data bus. I though with the regular IIs a strobe signal gates keyboard data from getting onto the data bus, but... I really never studied this part of either a II or GS circuit.
Checking the ADB Micro ADB data line (at the end of the cable all the traffic is a repeating message that looks like this:
This is my best guess: looks like it's fairly clean with the computer looking for a keyboard, and after timing out waiting on a response tries again. I didn't see anything else. I also didn't check the pin directly on the ADB micro, that may be something I should check. I checked at the port which I think it on the other side of a transistor and resistor. BTW tons of TPs listed on the schematics, but finding them on the board.... not an easy task!
Other suspects are the 3 245s, but a quick check the data one appears to be functioning well, A and B signals match.
The FPI or MEGA II could also be factoring in, but I don't know how to check those. They (and the ADB chips) are just black boxes... I would love to see what's happening inside (code or logic). The SRAM checked out good, I added sockets ot make addition testing on RAM easier. But I ran through all 10 41464s on a TopMax memory check and swapped with known good chips from a PC graphics card.
I've been digging through Jim Sather's books on the II and IIe and the GS technical reference (not really a great resource in my opinion). I'm definately in new waters but have some good GSes I can check signal on to see if there's differences.
I'm also using a known good 12-851 supply, but I've also used a good AA-11j041B, I've replaced all the caps on the motherboard. I've swapped out the 816 and ROM wiht known good ones. and the board still fails.
Any thoughs on how to proceed? I don't like the soft switch error, becuase that's address decoding is internal to the Mega II, right?
Can someone also explain how to properly use the '<!--break-->' keyword? I've never been able to get newlines to work with the AF interface and this time I tried something new after it was suggested and it's let me down...
I guess I should also add, you can access the CDA menu on boot up but you can not do anything there. Chararacters are not correct and the keyboard does not allow the user to interact with the menus.
The facebook IIgs group might be a better place to ask
Anything $Cnnn is a hardware control line.
Thanks Wayne, I've had an open discusson on the II group (there's a lot of crossover with the GS) but not getting a lot of help. I figured here may be a better opions, but I'm also pretty sure I'm dealing with things outside the knowledge of 99.9% of the users. It's also interesting the folks I woked with on II products have been quite too. That said, i'm hoping to grab one or two helpful ideas. Without some better design documentation it's really like dealing with black magic.
LaserMaster, yes, thanks. I know the Address Decoder handles those and the soft switchs are fairly well documented (including in James Sather's boosk). With the GS that logic is not clear. I believe much of that handling happens in the Mega II or the FPI, but I don't know it's defintely different and more complex. LOL
Two other thing that aren't totally clear are how keyboard data works on the GS. It appears keyboard data is palced on the data bus in phase0, like is done on the II. But that data appears to be getting out fairly frequently.
The other thing is the memory flow on the GS, there's SDBUS (Slot data), MDBUS(Main Data), and DBUS (816 data). So maybe keyboard data ins't getting on the bus at incorrect times because I can load things from disk. The disk uses the internal drive port, so I believe that's bypassing SDBUS, and coming from the IMU wihch is MDBUS. IMU and ADB GLU use MDBUS which likley has some arbitor that may block keyboard data while IMU data is active? That could explain how progams can load. I may try a disk II and see if the SDBUS data stream can bypass any interaction with the ADB GLU data. I see the 74245 are responsible for the Addressing direction control, but I haven't found many if the important data management magic pieces. I need ADB MCU and ADB GLU code and or logic schematics.
I have confirmed how the 245's work on my good GS. The next step will be checking the 245s are working correctly. That's really all I got right now.