Search results

Well as to "normal" if my guess as of the cause is correct I would say "yes, but also no". Take a look at the SndWatch VBL task from the Macintosh Portable ROM: SndWatch: movea.l PowerMgrVars,A2 lea (SwVBLTask,A2),A0 move.w #600,(vblCount,A0)...

On the M5126, the source and drain pins of Q5 are directly connected, so +12V is always supplied, and Q5 would be pretty useless. My guess was they decided that being able to turn off the sound wasn't necessary, like many other questionable decisions they made when making the M5126. As to...

A simple program/interface could probably written, and then run on basically anything with serial hardware. I suppose you could even do it from a smartphone. You can actually do quite a lot from that interface. Besides running the various hardware tests, you can dump memory, execute code, and...

The test has marked the upper byte (bits 8-15) of RAM as bad, so it could be any of the chips in the lower half. While it could be a combination of bits on multiple chips (for example 3 bad bits on one chip and 5 bad bits on another), it's more likely one or more chips are bad or failing to...

Assuming the select lines are connected in order, that range would match 4 chips, but since the error is in the lower byte the two chips for that range should be U4B and U1C.

*0 loads the starting memory address and *1 the end. If you try testing RAM on two different ranges, for example 00000000-0003FFFF and 00040000-0007FFFF, and you get the same error in both, then either bit 4 is bad on multiple chips or there is an issue with bit 4 across all of the RAM.

Anything like MacTerminal or Red Ryder probably should work fine. Proper settings should be 9600 baud, 8 bits, 2 stop bits. Be warned that linked documentation is missing most of the commands and has some commands that aren't in the Portable ROM. For example to run a march test on the first...

If you're feeling adventurous and have the means to hook up a serial cable and ground pin 7 on J22, you can likely get into the test mode and possibly isolate the bad RAM there, assuming there isn't an issue with data bit 4 across all of the RAM.

I guess there's also the CudaLite with PS/2 support, and whatever Cuda version was in the early iMac.

Those solder bridges are actually a form of bodge. There's line numbers next to the pads on the edges, and there are corresponding line numbers on pads connected to the driver chips on the opposite end of the board. If for some reason there's a bad connection, this allows you to connect the same...

Sure, go ahead and mail it over if you'd like. I could probably draw up a basic schematic as well. As for the Portable logic boards, I can look at them if you want but I imagine you could probably get them running pretty easily as well. I think the only thing I have compatible with the Radius...

Well I can't tell for sure since some traces are either under ICs or on inner layers, but here's some guesses on the card and notes on RAM cards in general, if anyone cares: Since it's using 512K x 8 chips, you'll have a pair of them on the 16-bit data bus for 1MB. Because of this you only...

I can't tell from the picture, is that the 74AC138 that the bodge is on? If part of the RAM on the card is disabled it may change the card's start address, which would could be why it's not working.

That's correct, it shouldn't need anything special. Either there's a fault on the card or something else is preventing it from being added as valid memory. I'm not sure if it's just the picture, but that's a very sketchy looking bodge running over that chip. Did someone try to modify the...

That would be a modem, not sure which one as it doesn't look like the standard Apple ones.

The first card is just 1MB, the second looks like probably 6MB.

The Caboose file doesn't seem to be in that ROM set, and I think it's missing a few CUDA versions as well. 341S0853 appears to be what's on the Quadra 900's board.

It seems to work pretty good in Ghidra at least, most of the work is just matching functions up to pins to identify what the code is doing. From what I've found so far it looks like the only difference between Caboose and Egret is the addition of control for the Quadra 900 (and 950?) key switch.

The ROMs aren't very hard to decompile, you could probably reverse engineer the entire thing and recreate it on something more modern. Picking a modern chip is the fun part, as you'll likely need to adapt it to fit the footprint. I believe there's actually a third one as well, an Egret...

So I did some more testing with the BlueSCSI v2, and I believe the issue is the transceivers or buffers are asserting control lines as the SCSI controller initializes, because the SCSI controller initialization is happening before the RP2040 finishes initialization. Here's the BlueSCSI v2...