exxos's Atari, Amiga & retro forum

It been a royal PITA but i have managed to get the Mediator working with the TF1260. Notes for any other hardware designers needing to figure out why your card goes nuts when a mediator is installed.

Firstly Mediator accesses long words at odd addresses.. My card wasnt happy with that originally but its legal on an 020 system but i guess not much else does it. So it might be a surprise use case.

Secondly Mediator has occasional very very short DSACK pulses. Nasty ...

Finally Mediator occasionally (not always) asserts DSACK1 and DSACK0 up to 15ns appart... just enough to fall either side of a clock edge.

This last one was a kicker... it meant the bus state machine thought it was acknowledging a byte then suddenly that cycle turned to a long ack.

I was able to find it relatively quickly with the logic analyzer.

I asked for help from Elbox about this but all i ever got told was PSU, PSU etc and then finally they stopped responding. In short i think the CPLDs on Mediator are a bit of a dirty hack. If your board stops responding in the middle of something i'd lay a bet its because of this.

you know. Amigahardware is always "magic" :)

Thanks for the update video. Been too hot for cranking on the computer in the evening. I did pop on my Gamecube yesterday and it was showing a temp of 48 deg c :o
Just to clarify your fix (and for os 3.9 issue) was via Firmware and not via TF PCB changes?

this is awesome news! My 1200 Tower with one of the original Mediator boards (think its LT4?) took me forever to get working stable with my Apollo 060 accelerator back in the day...I'll be putting the TF1260 in it once it's available so many thanks for tackling mediator issues & saving us users lots of headaches :D