TurboMaster CPU incompatibility with the MK2 board

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser. Please create an account and log in to see all content by default. This is a limitation of the forum software.


Please understand that you need to create an account to be able to post, guest posting was disabled as an anti spam measure.

    Hello.

    I've been using my TurboMaster CPU accelerator successfully with your MK1 64 board with no issues at all.


    But when I try to use it with my MK2 board, the board doesn't boot up at all. Wondering if you have any thoughts on what's broken on the MK2 board that's preventing the TurboMaster from working. Would seem to me there's a loss of compatibility in the MK2 that the MK1 had.

    I own two of your MK2 boards, and they both behave the same (don't work with the TurboMaster). So it's unlikely to be an actual "failure" in my boards, but rather an incompatibility issue in the MK2.

    Is there any chance that you could suggest how to fix the MK2 so that it works as well as the MK1 board?

    Thanks,

    Thomas

    First of all, I should ask for the actual hardware that you're trying there, as "Turbo master" can mean an ealy 65c02-based accelerator, or a rather new thing that came out recently. Which one are we talking about?


    All a CPU accelerator does on the expansion port is to execute DMA on the 44-pin expansion port. This does work on the MK2 board with the Commodore REU and with the Chameleon (which also includes a Turbo), so there is no "general incompatibility", but likely a not-so-clean design choice on the accelerator side. Let's hope you're talking about "the new thing", so we can talk to the developers and see if something can be done on their side.


    Jens

    That "new" thing is an exact 1:1 clone/repro of the original one, not a new design (That's generally all Corei64 does) - so it will behave exactly the same (and will not get fixed/modified)

    Thanks for your quick reply.

    I am actually the developer of that new "THING" you're referring to.


    I should have been clearer in my original post, apologies. So here we go.

    The original TMs (V1 and V2) work on your MK2. My reproduction using a Rockwell R65C02P4 also works on your MK2. My reproduction using a new WDC 65C02 CPU in the device doesn't work with your MK2.

    HOWEVER, my reproduction using the new WDC 65C02 works on every other board I've tested (original 64's, short and long, the remade DIY solder yourself boards, Gideon's U64, the EVO and your MK1). Your MK2 however does not work with the reproduction TM using the new WDC 65C02 cpu.

    It should also be noted that both original TurboMaster versions using an adapter in the CPU socket and a new WDC65C02 CPU works on all the aforementioned boards EXCEPT your MK2. So, the problem isn't with the board or GALS, it has to do with your MK2 not being agreeable to work with the new WDC65C02 in the TurboMaster. Every other board that I've tested works in that configuration except your MK2.

    I had to go with a WDC 65C02 because the R65C02P4 variants are no longer available, and anything listed for sale these days for that chip are fakes.

    Given all this, it is clearly a compatibility issue that your board has that no other original or recreation has EXECPT your MK2. It seems apparent that something broke between your MK1 and MK2 release.


    Hope that shines a light on the MK2 compatibility issue for you.


    Thanks for your time and comments.

    Need to add something.

    I spoke out of turn on the EVO. I went back and re-tested the EVO board, and it too is exhibiting the same issue that your MK2 board has.

    What's similar between your MK2 and the EVO is an implemented/emulated PLA. Your MK1 has a socket for a PLA chip. I've tried various PLA chips in the MK1 and haven't run into issues with the TM with the WDC cpu.

    All of the other boards that work with it, including your MK1 have a socket for the PLA.

    I'm wondering if it's related to the implementation of the PLA in your MK2 and in the EVO boards that's causing this issue with the TMv3.

    I'll add one more thing to the mix. I believe that you could make your MK2 compatible given that Gideon's Ultimate 64 is compatible with the TurboMaster running the WDC cpu. So, that is an example worth noting that it's a board that implements the 64 design entirely without the need of any original chips yet still works with the WDC cpu installed in the TurboMaster. It furthermore also works while using the older R65C02P4.

    So I believe it should be possible for the MK2 to be made compatible I'd think...

    So you're copying an old design, making a change because you can't get a certain part any more, and now things don't work any more. Then you write:


    Quote from COREi8

    I am actually the developer of that new "THING" you're referring to.

    Where is your development work? I mean, did you run timing simulations? Did you make sure that the differences between the original part and the new part are accommodated for in your design? After all, it's a pretty bold assumption to make a change and hope that it will work just the same.


    If the accel is *that* old, it may even have trouble with the short-board C64 boards from Commodore (the ones with a 64-pin higher-integration PLA).


    If you want me to fix your design, we can talk about that. Let me know what your budget is - for 8-bit work, I can deviate from the standard rate of 1900,- EUR per project day - that's what I charge industrial customers.


    Jens

    Nice deflection Jens.

    I guess you didn't bother reading that it works on every original Commodore model board (short and long boards). As a product developer, I would think you'd want to achieve that level of compatibility with your MK2.


    Quite honestly, it's no concern to me that it doesn't work on your MK2 with the WDC cpu. I can make it work using an old P4 so it's all good on my end but thanks for the offer. Your MK2 is clearly not 100 percent compatible with the original C64 board and thought you'd appreciate the opportunity to improve your product. Guess I was wrong.

    Have a nice day.

    Quote from COREi8

    Nice deflection

    Says the guy who didn't answer to the question about timing models and datasheet differencies. There's a difference between "working by design" and "working by accident". I'm handing you the procedural method on a silver platter, and you just claim that my product isn't compatible, when it's still unclear which parameters you even changed. This lead me to believe (after post #4) that you don't know yourself.


    What's your assumption on the data-valid window on the address bus? What's your assumption on the window where you're allowed to drive the bus? When in the Phi2 cycle are you taking over read-data?


    Ask yourself these questions, compare with all datasheets, make adjustments, and you'll find that you have a working accelerator - with all Commodore boards (note that the C64RMK2 is indeed a Commodore board).


    Jens

    Compatibility with all the original Commodore 64 boards is my benchmark and I've achieved that, so job done.

    Have a nice day.

    Oh, so you are *really* dodging the question about technical details. Right, yeah: Copying-job done. Development job? To be done.


    Jens

    So we're continuing on with this then eh? How about you take your own advice.

    MK2 copy of a 64 done! Development to make it 100 percent, you cannot be bothered I take it?

    Come on Jens, if your MK2 was 100 percent compatible with an original 64, then this TM (that works on every other original 64 and 128) would work on yours.

    Why on earth would I waste my time making this TM work on your flawed board? It works on your MK1 (that apparently you did a better job on than your MK2).

    That's fine that your MK2 is less than 100 percent compatible with an original which I've clearly demonstrated. But to expect someone to degrade their design to make their device work with your flawed board is laughable.

    Why not just fix your board? If you want to compete with all of the other 64 reproductions out there, you're going to have to do a better job. Or, just withdraw your MK2 and bring back your MK1 (which is a much better board than your MK2). Just sayin'.

    The TM copy IS done, and it IS 100 percent compatible with all original boards. Can you say the same about your design? I've proven the answer to be NO.

    Impressive - so many words, yet not a single number in terms of preconditions like address setup time, address hold time, data setup, data hold (especially on transition into the VIC cycle).


    If you can point to where exactly the MK2 does anything wrong when you are taking a card that works in it, then swap the CPU that wasn't available back in the days in order to make it not work, then yes, I'm willing to take a closer look. Just give me the timing parameter where I'm wrong - I've been trying to steer this back on the technical track for some time now. Just give me the parameter you you believe the C64 Reloaded MK2 fails to comply with.


    If you think that your development assumptions for those numbers don't belong into this public forum, just send me an eMail. I'm even willing to sign NDA, which would keep me from making a 4MHz 65c02-based accelerator for a few years. Just get off your "blame the others" track and follow the technical path - that's the one that leads to a solution.


    Jens