[Coco] Upgrade MPI?

[Joe Grubbs]

I knew of the address overlap and potential software crashes that could occur, but this is the first I've heard of actual hardware failure. I wonder if this is theory, conjecture, or if it has actually happened to people?

Either way I'll likely continue to use my modified 26-3024 with the CoCo3 and reserve this one for one of my CoCo2s. 

Speaking of which, I noticed that the selector switch on this 26-3124 is detented, whereas the one on my 26-3024 has no significant detents. I just figured they were all like that. Is my 26-3024 actually broken and I've just never known?




> From: barry.nelson at amobiledevice.com
> Date: Thu, 7 May 2015 03:53:58 -0400
> To: coco at maltedmedia.com
> Subject: Re: [Coco] Upgrade MPI?
> 
> It's not a software issue, it's HARDWARE. A non upgraded MPI on a CoCo 3 attempts to output conflicting voltage levels at the same time as the GIME chip when certain registers are read. This effectively causes a short circuit as one chip attempts to output 5V and the other chip tries to pull the line to 0V. This happens for a very short time, but if the right conditions are met, the GIME chip CAN be fried. If you care about the CoCo 3 and want it to last DO NOT use an un-upgraded MPI with a CoCo 3! It can fry the GIME. The is an upgrade you can apply that does not require a programmed pal ship, just a single 74LS10 chip and a few wires. Here is a link.
> 
> http://www.coco3.com/community/2010/05/26-3124-mpi-coco-3-upgrade-1/
> 
> DO NOT use it on the the CoCo3 without upgrading it! The 74LS10 chip is a standard part and is readily available.
> 
> FYI: I have used this upgrade on a MPI I own and in addition to working correctly on the CoCo 3, I am still able to use it on my CoCo 1 as well.
> 
>