TF536 on A300/A600

[andipexton]

Hi I appreciate this is quite an old thread, I have started looking at this and I have rolled my own adaptor for the TF536 and I have it in the A600 but it is only working with KS 2.04 (i.e. A500+ ROM) 3.2 it hangs on a grey screen. DIAGROM runs through ok.

With 2.04 I can run Elite II no problems and it's smooth

Obviously I need a later ROM to use IDE but I'm guessing the issue is when it is trying to use the TF536 IDE which conflicts with the internal IDE and PCMCIA

I'm running the firmware I found on the forum for A500/A600 - later/latest firmware just hangs after the rainbow bars regardless of ROM (Not tried with DiagROM)

My question is would firmware with just CPU and Fastram for the A600 be possible, allowing the internal IDE and PCMCIA to function, I know IDE will be a little slower but I would prefer compatibility over the possible speed increase

If there is a more stable configuration/combination of KSROM and firmware available already please point me at that

[terriblefire]

For me this is too far off the beaten path.

A600 isn't supported nor is KS 2.0. 536 is already supported on 500/CDTV/2000/1000/A570 on 3.1, 3.9 and 3.2.. that's already 15 combinations to test.

i tried the 536 in the 600 and i never got it to work.

recommend buying an A600 specific card.

[terriblefire]

I can expand briefly on why the TF536 will not work reliably on the A600...

Basically the TF536 generates its own E clock and it relies entirely on this. In the A600 the E-clock on the onboard 68000 cannot be disabled or shut up and thus as soon as you get past the colour bars and get to something that uses a CIA you will get a hang unless you are incredibly lucky and get the two clocks in phase.

There is nothing i can do about this without a specific A600 designed fw and I am not going to do that. The card is not designed for the A600.

[andipexton]

I can expand briefly on why the TF536 will not work reliably on the A600...

Basically the TF536 generates its own E clock and it relies entirely on this. In the A600 the E-clock on the onboard 68000 cannot be disabled or shut up and thus as soon as you get past the colour bars and get to something that uses a CIA you will get a hang unless you are incredibly lucky and get the two clocks in phase.

There is nothing i can do about this without a specific A600 designed fw and I am not going to do that. The card is not designed for the A600.

Ah!

Thank you, you have been more helpful than you think... if Terriblefire won't come to the A600, maybe ... I'll post back if I have any updates

[terriblefire]

I can expand briefly on why the TF536 will not work reliably on the A600...

Basically the TF536 generates its own E clock and it relies entirely on this. In the A600 the E-clock on the onboard 68000 cannot be disabled or shut up and thus as soon as you get past the colour bars and get to something that uses a CIA you will get a hang unless you are incredibly lucky and get the two clocks in phase.

There is nothing i can do about this without a specific A600 designed fw and I am not going to do that. The card is not designed for the A600.

Ah!

Thank you, you have been more helpful than you think... if Terriblefire won't come to the A600, maybe ... I'll post back if I have any updates

Just to add the TF536 is one of the cards with the no derivative licenses. Before you go modifying fws etc.

[andipexton]

Ah!

Thank you, you have been more helpful than you think... if Terriblefire won't come to the A600, maybe ... I'll post back if I have any updates

Just to add the TF536 is one of the cards with the no derivative licenses. Before you go modifying fws etc.

No intention of doing anything to infringe or replicate your work
I did an experiment modifying the A600 to Take E from the TF over Gayle interesting but not any more promising as yet
I did same experiments with 68k as well as a control

Speaking of the earlier cards i.e. the TF530 do you know if anyone modified one of those for the A600

[terriblefire]

Just to add the TF536 is one of the cards with the no derivative licenses. Before you go modifying fws etc.

No intention of doing anything to infringe or replicate your work
I did an experiment modifying the A600 to Take E from the TF over Gayle interesting but not any more promising as yet
I did same experiments with 68k as well as a control

Speaking of the earlier cards i.e. the TF530 do you know if anyone modified one of those for the A600

TF53x all have the same core code for 6800 emulation so they'll all behave exactly the same.

[andipexton]

TF53x all have the same core code for 6800 emulation so they'll all behave exactly the same.

... I guess no chance you're going to open source the firmware?

Thanks for entertaining my questions

I'll probably publish my failed attempts, and partial success for entertainment sake and move on reluctantly...

[andipexton]

TF53x all have the same core code for 6800 emulation so they'll all behave exactly the same.

Sorry... One last question... How predictable is the phase of E to the clock on the TF, i.e. does it have a known initial state on coming out or reset or is it like the OG 68k and could be anything? Just figured if you know it saves me a load of experimentation

Thanks in anticipation

[terriblefire]

TF53x all have the same core code for 6800 emulation so they'll all behave exactly the same.

Sorry... One last question... How predictable is the phase of E to the clock on the TF, i.e. does it have a known initial state on coming out or reset or is it like the OG 68k and could be anything? Just figured if you know it saves me a load of experimentation

Thanks in anticipation

It's unaffected by reset... it needs to be. There is no way to know the relative phases of the TF and or the A600 E clock. the solution is to redesign the TF FW to use the E clock as an input rather than an output.

if this was a trivial thing i'd have done it by now