TF1230 Beta Interest (Open Source)

[go0se]

Booting straight out of the traps to AmigaTestKit via Gotek on a newly recapped 1D.4 and a 2B still on original caps.

Both powered using the same A500 4.3A "heavy" psu with no previous know issues.

No fixes have been applied to either machine.

Both on 3.0 roms 39.106
CPU is a genuine MC68030RC50C
Memory chips are ISSI IS42S16320B - 7TL ( 128MB capable, configured at 64MB using current firmware )
10C CPLD.


More memory and ROM combinations to follow.

[go0se]

Another card built and booting. This time with Microchip 48LC16M16A2 SDRAM chips (64MB total fast ram). Mem testing OK so far.

[go0se]

A couple of questions:

Is the Enable 030 jumper currently hooked up to any logic on the CPLD? Both cards boot for me with that jumper enabled or disabled.

Will the orange coloured boot bars be featuring on this card? On the TF536 hey give me a tremendous sense of comfort and reassurance every time I see them :D

[terriblefire]

A couple of questions:

Is the Enable 030 jumper currently hooked up to any logic on the CPLD? Both cards boot for me with that jumper enabled or disabled.

Not at moment.

Will the orange coloured boot bars be featuring on this card? On the TF536 hey give me a tremendous sense of comfort and reassurance every time I see them :D

Probably not because we're on the cheaper 144 variant of the CPLD. I'll see if it can be squeezed in but I guess you would all like fast IDE instead?

[go0se]

Probably not because we're on the cheaper 144 variant of the CPLD. I'll see if it can be squeezed in but I guess you would all like fast IDE instead?

If I voted for coloured bars over fast IDE I understand that I would be in the minority :) I guess I'm going to have to learn to live with the uneasy few seconds of uncertainty before something appears on screen!

[terriblefire]

This board, if mass produced, could be £40 per board. Alas nobody would do that and i get why.

[go0se]

This board, if mass produced, could be £40 per board. Alas nobody would do that and i get why.

I can vouch for the fact that it is pretty trivial to assemble the remaining components when they come with the tedious bit done via partial assembly. £40-£45 BOM cost without a CPU sounds realistic for someone building one for themselves.

I will say in price defence as someone who has sold many's a TF card over the last few years that the costs that go unseen are:

Time spent testing each card until you satisfy yourself that it really does work as intended.
Replacing the cards that go 'missing'.
Replacing the cards that did work but then got nuked by the new owner.
Replacing the cards that just stopped working inexplicably.
Having a bucket full of PCBs that just refused to ever work having swapped out every single component more than once during the build.
Buying a lot of dud reprinted CPUs in order to get some that work at 50MHz,
Paying £30-£35 for each CPU once you eventually find someone selling legit, non reprinted CPUs, and even then, some of those don't work!
Ebay's pound of flesh.

I guess as these cards were designed by TF for the hobbyist they should be built by the hobbyist for the satisfaction of building your own card. When it becomes a commercial endeavour then people will always accuse you of taking the Gipsy Kiss.

[go0se]

Success! I've just found a really fussy motherboard, another 1D.4. Gives a mustard screen on one test card, and a black screen to no signal and occasional mustard with the other one. Motherboard boots and runs ok with no accelerator installed.

Both test cards seemed happy in the 2B. I will try some more tests with this one and see if I can narrow it down.

[go0se]

Some further observations.

Both cards still booting and mem testing without issue on the rev 2B motherboard.

On 1.X motherboards the card with ISSI IS42S16320B - 7TL memory chips goes into a black screen boot loop and eventually occasionally boots and the card with the Microchip 48LC16M16A2s gives a mustard screen.

I removed the 48LC16M16A2 chips from that board and replaced them with IS42S16320Bs and the mustard issue was replaced with the same boot loop issue as the other card.

While I had the memory chips off the board I plugged it in and it booted straight to the red diagnostic error screen as expected and on into Amigatestkit on continue. 030 detected and 2MB chip RAM.

The same issues are present on both a 1A motherboard and the 1 D.4

I know the 1A hasn't got any issues as it was serviced by Wang Computers on 15th May 1994 :lol:

@terriblefire : Any chance of a firmware build with memory left unconfigured?

DSC_1694.resized.JPG

[terriblefire]

Does your 1.1 board have the timing fixes done.. If not it absolutely will not work.