Hello,
I've been reading this forum all night, all hot-and-bothered to figure out how to stuff a FF3j rom onto a cart, and in trying to figure out a better way to do this than cutting up a rare variant of SMB2 I found that Nintendo made the TKEPROM, back in the day.
!!!
Way cool, and useful for probably every MMC3 based game.
Obviously, you guys already knew this, but what I lack in NES development knowledge I make up for in electrical engineering. If I can get really good photos of both sides of a preferably unpopulated NES-TKEPROM-02, I can work up the design and have it fabbed - it's looking like a run of 100 would cost $4 apiece, so I'm thinking there's probably enough people here interested to make it a worthwhile venture. 20 people at $10 per would cover half the costs, I think that'd make it worth my while to float the rest.
Obviously, it would still need a transplant MMC3 (possibly also that Nintendo-marked DIP between the PRG ROMS as well, if there's no ready substitute), BUT, there are a LOT of MMC3 games out there, and it's only a QFP-44, so not terribly hard to rework.
The upshot of this is that since the board would be a duplicate of something that is known to work, there would be no real development time other than the CAD work, which would be a weekend or two (or three).
So, any interest, advice, etc?
Well sorry but how could you actually put a MMC3 on the board ? Even if you'd be able to remove the MMC3 and solder it again on the new board, it'd probably be much easier to modify the original board into a devcard.
As a hunch -- replacement TKROM boards aren't particularly useful without having cloned the MMC3 in a {CPLD/FPGA/PAL/whatever}. There's value in having a new clean board, but you've still got a limited resource of MMC3s (and desoldering them correctly).
lidnariq wrote:
There's value in having a new clean board, but you've still got a limited resource of MMC3s (and desoldering them correctly).
Few people complain about the limited supply of Super NES CICs desoldered from the limited but still large quantity of D-rarity annual sport games to put in SNES PowerPak. But I agree with you that an MMC3 or useful subset thereof would be the best solution. I wish I knew Verilog.
Not quite a good example. My SNES CICs come from suppliers in China who have hundreds/thousands sitting around. They appear to have no solder and still have the legs angled so I think they are unused.
More interesting would be to find a source for the MMC3 clones used by the Chinese pirates/companies and make a board built around those.
When I read the topic I thought someone was willing to clone the MMC3 and make new boards without having to resort to original carts at all. If the mapper itself still has to be lifted from an official board (which I imagine is not easy, because of the tiny legs), I see no reason for a new board, I can just use the one where the chip is.
A 3.3V CPLD and 3.3V regulator can be used as well....
I actually have a bunch of MMC3 clones that I bought last year from a supplier in China (DIP-40 packaging). It's embarrassing to admit since I knew I was taking a gamble on them (just more NESdev experimenting fun), but as it turns out very few of them actually work. I believe it was 2 out of 10 tested that work.. some of them were partial, one worked except for the IRQ, for example..). And these are brand new chips (well, NOS). I have a bunch more that are untested.. but with that defect rate it's too much trouble for me to want to make a board at this point.
This chip would be good for a project like this, if you're interested enough I could contact them again and see if they really want to get rid of those chips. I'm sure I don't have 100 working ones, but I'd be happy to see any of them go to use. It's not hard to test them, it's just time consuming without a ZIF socket.
A lot of working MMC3 carts get the axe to become repros. I try to buy surplus NES carts every now and then (SMB/DH mostly), but often I've been told there are no MMC3 ones because they've all been bought up for that purpose. It's cool on a smaller scale but there is a point where it's a lame thing to do. People have and will make donors out of Kirby's Adventure, basically anything available that fits.
And for an amazing coincidence my proto MMC3 boards with 3.3v CPLD/vreg arrived today
Will be a while before I can do anything with them but I am hoping use them for other things like VME7 too. Donors will still always be cheaper because even at product levels $10 doesn't cover the pcb/parts/soldering.
If it still require to destroy another cart for the mapper chip, the answer is no. If we need a solution for MMC3 board, we need a clone mapper that comes with it too.
bunnyboy wrote:
And for an amazing coincidence my proto MMC3 boards with 3.3v CPLD/vreg arrived today
Will be a while before I can do anything with them but I am hoping use them for other things like VME7 too. Donors will still always be cheaper because even at product levels $10 doesn't cover the pcb/parts/soldering.
Pics! Just sell the PCB and CPLD kits, I'd take one now w/o CIC or anything and work on the MMC3
Don't even know if everything is hooked up right yet...
http://www.sealiecomputing.com/images/repropakmmc3revAfront.jpg
http://www.sealiecomputing.com/images/repropakmmc3revAback.jpg
Edit: and I know they won't work as is, picked the wrong part number so its a 2.5v vreg
@BunnyBoy:
That' a move in the right direction. I guess that answer the OP that he doesn't need to make a board since there will be one in a near future. I support your work for this.
His could still be a much cheaper option for those who can remove the MMC3 from existing carts. Including plastics/shipping/ciclone this is going to be $20+. It's also a great learning experience to build boards
Why the oscillator? Not the same PowerPak MMC3 is it? XD
Oscillator is used in some mmc3 irq implementations, is useful for other things like nsf players, reset detection, etc. Almost certainly will be on the final boards. You can cut it off if it bothers you too much!
bunnyboy wrote:
His could still be a much cheaper option for those who can remove the MMC3 from existing carts. Including plastics/shipping/ciclone this is going to be $20+. It's also a great learning experience to build boards
I'm not saying that the OP should not do it. You're right, it must be a great learning experience.
What I'm happy about is the fact that now if someone wants to make a homebrew game, they will be able to target in the future a MMC3 like mapper and won't have to restrict themselves anymore. Now we must stop procrastinating and start homebrewin'
By the way, since the mapper seems to come already soldered on the board, making a famicom version must be out of the question I guess (not worth the investment). I need a toaster nes again.