Search the Community
Showing results for tags 'grom'.
Found 4 results
Hi, I am currently with Ciro in his place, we did dump his original cartridge TI Logo II - Deutsch. We did use the GramKracker device and also verified our finding by reading the grom memory when the module was in and all GramKracker functionality was off. The cartridge uses - 8K Rom - Grom 6000 (bank 3) - Grom 8000 (bank 4) - Grom A000 (bank 5) - Grom E000 (bank 7) The oddity is the Bank 6 is unused but nevertheless its memory space is reserved, as we see the last grom starts at E000 It is matching the "information" from the PC99 dump from Mike Wright, but it appears the rpk dump on whtech is using the content of bank 7 at bank6 if you examine the .bin file. Both dumps allow you to start the cartridge, but I can imagine that once the content of the last grom is required only one will work. So the question is, can the Grom Bank be modified without breaking functionality? Br Klaus
I didn't want to hijack the 512K cart thread, so I thought I would post this here: I actually have a cart, custom built for me by Jens-Eike (it's with him in Germany) that is a 64K bank-switched cart with 1K of ram permanently mapped in to the end of the cart ROM space (so there is 7K or ROM per bank, and the same 1K of RAM in every bank). It also has two hardware stacks on it but there's issues with the stacks - we can do a push (post increment) but pops are proving very tricky (pre-decrement) - something to do with a lack of appropriate signals in the cart port to trigger the pre-decrement, so that might not make it into the final design. We'll have to see. It's a shame because the stacks are lovely: Simply write to an address (I think it's >7FFE) and it pushes that value to the hardware stack. Do a read from the same address and it does a pop! The RAM for the stack is taken from the 1K of RAM. Of course, you don't need to use them if you don't want them, but I want them for a TF V2.x which would have a hardware stack (and thus be faster). It's currently implemented in discrete logic, but I'd be happy to hear to from anyone that is willing to tackle the problem with programmable logic. As I say, there are issues that need to be solved. I spoke with Gary Smith about it here in the UK. He has looked at the problem and agrees that it's a tricky one. He has a tentative solution using flip-flops to induce a delay (if I have understood it correctly) between detecting a read from the stack port (triggering an address decrement) and performing the read. I'd like to get the whole thing into one chip: That is: 64k ROM, 1K RAM and all logic. If anyone is interested is doing a design please let me know. Jens-Eike and Gary has expressed an interest, but it's hard to keep the momentum going - understandable enough... We all have lives outside of the TI
I played the cartridge version of Parsec and the GROM emulator version (off my CF7) last night and the real cartridge 'seemed' slightly more responsive. Was that true or am I just seeing things? IS there any difference between the two.