ACA500 Green Screen Errors - EXC27

Quote from stevenwarren82

Checkmate Case and updated Zorro Adapter

Please remove and try with the recommended configuration. The ACA500plus is only guaranteed to work if connected directly to an A500 board. Nothing may be plugged between the accelerator and the A500 - especially not an unbuffered Zorro expansion.

Quote from stevenwarren82

I am aware of the timing issues with the Checkmate case and Zorro adapter, but these seemed to be easily overcome in the past by ensuring a good connection and powercycling the Amiga.

I haven't seen any case where praying and re-trying has improved actual design flaws.

Jens

Quote from stevenwarren82

works perfectly.

until it doesn't, because the whole system isn't made for that kind of bus load. And it certainly isn't made to be powered by an ATX power supply - see our PSU FAQ.

Jens

And yet some Amiga 500's with ACA500plus(e-clock modded) + ACA1234@50MHz + A2048 + Indivision ECS V4 + X-Surf-500 + GRAKA 500 Zorro II RTG all plugged into a second release Checkmate 1500 plus backplane board and powered by the worlds cleanest MeanWell modular power supply .. just keep ticking along ... since May 2021 here ...

That's as good or better service life than your average unmolested A500 typically sports
I mean these are 40 year old beasts... designed to survive their 90-day to 1-year limited warranty, sold by a company that was notorious for cost cutting, substitutions and spotty quality.

Sure .. my build has had some issues like:
- MicroMys with select model PS2 mouse compatibility
- Lyra with timing being incompatible with select games
- ACA1234 had some serious issues straight from the factory
I wouldn't hold those issues up and say that they alone made the associated products unsuitable for use with Amiga, any more than I would say the second release Checkmate 1500 plus backplane card makes the ACA500plus unsuitable for use.

Sure .. there could be RLC marginality issues, and there almost certainly was a ground plane issue that caused signal reflections with the first release Checkmate 1500 plus backplane. However, the second release backplane cleared up the no-op condition, and I've not seen it demonstrated or proven, one way or the other, that there is a problem, just some idle speculation and cast aspersions.

That your Amiga 500 and my Amiga 500 and other peoples Amiga 500's are working fine in this configuration suggests the concern is likely overhyped / overblown.

It's clear we don't get factory support for our use case ... but given the great fun ... I think most of use are happy with the trade.

I am incredibly glad to read that a quick board swap got you back up and running

Quote from ljmarent

the second release backplane

..is something I'm reading about for the first time now. I know that I've sent extensive e-mails to the maker(s) back then, but never got a reply. I can't say if they implemented everything I suggested, and I certainly don't expect a thank-you, given that I'm still not willing to give support for the setup.

Jens

Jens

I'm sure we chatted about the second release Checkmate backplane card once before, I don't expect you to recall, as it's not a supported product.
What I'll do is post a couple pictures here, so that it's not just me 'saying' that there is more than one version and we don't waste time searching for a long ago chat.

This is the first release:

And this is the second release:

You can see how the first release ran a couple traces right through the ground plane.
All the signals travel perpendicular to those two traces on the reverse side of this two layer copper card.
There was never any detailed explanation of the changes or a testing report supplied by Checkmate, at least that I ever saw.
Nor was there testing report from iComp that showed there was any issue with the second release Checkmate backplane.

We all just accept that it's outside of the installation guide of the ACA500plus.

The first release generally resulted in a no-op condition with the ACA500plus.

The second release simply 'worked' better than the first, from my own experience and from reading and watching the videos of others.

I wasn't privy to your communications with Checkmate,

but I'll guess the second release surely does not incorporate all that you might have suggested.
By which I mean:
- it has still got an ATX power connector and no minimum load or power conditioning.
- it has still got a Zorro II 'like' slot without buffering.

Checkmate \ Rob Cranley, did make visually and functionally apparent improvements after your thankless efforts.

Moving forward...

Now that there is a 'Bluster' CPLD replacement for 'Buster' a four layer version of this backplane with two buffered Zorro II slots and better regulated power supply would be terrific. I've got your 300 fancy German hamsters ready to trade.
We'll all just sit here and play games on our Checkmate 1500 plus's until someone rings the bell and tells us our upgrade is ready.

Quote from ljmarent

Nor was there testing report from iComp that showed there was any issue

No need to - this board is completely without bus buffers, so it's out of spec. Why would I waste any time on that?

Also, ATX PSU may be OK if you provide a point-of-load regulator. I don't see that on these pictures either. So again: Why bother if fundamental design practise is ignored, just for the sake of providing "something", not telling the customer that this is never going to work reliably with "anything Zorro"?

Jens