StarMax 4000 only boots via CUDA reset button?

[Mr. Ksoft]

I recently picked up a Motorola StarMax 4000/160 DT.  Seems to be in decent cosmetic shape, no obvious damage on the board, but it doesn't power up via the power button or the keyboard soft power.  Strangely it powers up immediately without pressing anything if plugged in after being unplugged for a while (a day or so?).  The power button will turn it off but then not back on.  The only reliable way to get it to fire up is to press the CUDA reset button, which turns it on every single time.  (But that has the side effect of requiring the case to be open, and it wipes the PRAM).  The system works completely fine once it's on.

Of course first thing with these Tanzania boards is to replace the battery -- I have done so (fabricated my own 4.5V battery using a 3xAA battery holder) but no difference.  Multimeter confirms that my battery setup is putting out 4.5V, so I am fairly certain it was put together properly.

My best guess is that there is some kind of issue with the soft power circuit but I am not really sure where to start with troubleshooting (possibly checking that the board is actually getting the battery voltage).  Anyone have experience with the Tanzania board that might know something I don't?

 

[jessenator]

I've never tried booting my 4400 with the CUDA button. That's interesting.

Just a crazy thought, did you try manually jumping the power switch terminals? It could be a mechanical fault with the switch perhaps, if it still turns it off.

 

[Mr. Ksoft]

I just tried that but it doesn't seem to work.  I wonder if something elsewhere in the system is keeping it "held" or something.

Also, an aside, but I can't get the damn thing to boot any CDs whatsoever.  The internal drive should be bootable, but isn't, and external SCSI isn't doing the job either.  Having an OS on this would be really helpful!

Edit:. Ok, apparently I was just really bad at picking OS discs. Retail 7.6.1 and 8.1 don't work but the Apple System Recovery 1 disc boots... So that is fine.  Still can't power it on aside from the CUDA button though.

 

[jessenator]

The 4400-specific disc images of 7.5.3 (the system restore disc) on Macintosh garden worked for me as well. These boards need a specific system enabler that not many discs had. Even the retail 7.6 disc wouldn't work for me.

I wonder, could there be a cold solder joint on the switch's underside?

 

[Mr. Ksoft]

I checked continuity across the board... So it seems the power switch flows through the CUDA reset button and something around there is controlling the soft power.  There's also good continuity from the soft power connector on the PSU to the CUDA button.  Probing the contacts around the button while plugged in seemed to trigger it to reset the CUDA and boot.  I did try manually jumping the power supply via the soft power connector, which does fire up the power but no boot or chime, probably cause I bypassed the CUDA.  I suppose I could try to reflow everything and see what happens.  A schematic would help me get further but I suspect there isn't one out there.  I couldn't find one.

 

[jessenator]

I'd be interested in a Tanzania schematic as well, but for different reasons  hope you get things sorted out.

It seems that Adreas Kann has/had some of them… I haven't tried the email on this (archived) site, but could be worth a shot: https://museo.freaknet.org/gallery/apple/stuff/mac/andreas.kann/schemat.html

Well, that solves that… I guess if anyone knows him or can get ahold of him that would be cool.

 

[netfreak]

I had a similar problem on my StarMax 3000 where after sitting for a long time it only powered on once then seemed to be dead. Had to use the motherboard reset button a few times but since then it seems to power up normally each time I try. I'm also running it without a battery.

 

[Mr. Ksoft]

I did some investigating and reflowed some suspect pins but no good.  Still only boots from the button.  What's bothering me though is the fact that I can trigger the CUDA reset by simply probing the pins on the button-- just touching the pins on the sides cause it to reset and boot (not a normal boot, the clock and everything still get reset).  Not completing the circuit with the multimeter... Heck, the multimeter wasn't even on, so not sending a voltage through it either!  Might this be a hint as to what is going on?

 

[Mr. Ksoft]

Yes, I'm reviving this old thread of mine, as there's more info to play with now.
I still have this StarMax and it still has the same problem.

However, the news is that I saw in another thread that the impossible has happened and the Tanzania schematics are now out in the wild! No idea when that happened, but they're available now! And, I have a scope now. Two valuable tools in the fight against this haunted machine. So, I've been tracing out the power circuit and the CUDA Lite and trying to see what I can find. (For anyone looking at these, the relevant sheets are mostly 16, which has the CUDA and power switch, and 24 which as the power supply connections) I'm still mostly a software guy, though, so I can describe what I've observed but not put all the pieces together yet.

• CUDA Lite is properly receiving power, +5V trickle from the power supply is present
• Observations on power-related pins on the CUDA:
  • Pin 17 is unlabeled on the schematics, but the CUDA emulation from MAME says this should be "trickle sense" for soft power systems. It is always high at 5V, so I guess that is the trickle power. I think this is correct.
  • Pin 18 (FILE_SERVER) is always low until the power is turned on, then it goes to 5V for a second or so and drops back to zero. Based on the notes in the MAME code, it says "pull up for soft power". It does indeed get pulled up, but I'm not sure if it should only be pulled up when the system turns on or if it should always be pulled up.
  • Pin 24 (KEYBOARD_PWR_ON_) which is connected to the ADB port and should support ADB power-on, is always low, and doesn't do anything, even when pressing the power button on an ADB keyboard.
  • Pin 25 (CHASSIS_SW) is held high until I press the power button, and then it goes low. This appears to be expected behavior, so the power button is OK.
  • Pin 26 (PFW) comes from the three pin power supply "soft power" connector as "PFW1_", this should be the equivalent of the ATX "PWR_OK" signal. It is low until the power supply is fired up (such as by the CUDA reset button) which I also believe is expected behavior.

I am guessing that for some reason or another, the CUDA is not sending a signal to the power supply to power on... except when the CUDA button is pressed, for whatever reason. I want to make sure of this, however, I can't track down where that is happening. What I am not seeing on the schematic is how that power on signal is getting sent to the power supply. I guess through the soft power connector, but that just appears to have the 5V trickle supply, the "PFW1_" (assumed "PWR_OK") signal, and ground. Unless "PFW1_" is somehow doing double duty as the equivalent of both ATX's "PWR_OK" and "PS_ON" signals and I don't understand it because I just don't have the knowledge yet... which is likely. What am I missing?

 

[treellama]

Oh interesting, last time I tried to use the Starmax it wouldn’t power on. I just chalked it up to “Starmax things”. Will have to try CUDA reset.