Help flashing Famicom games on kazzo/anago

[Jdoesstuff]

What about these NROM boards from Muramasa? I have a couple of those I haven't done anything with yet.
https://www.muramasaentertainment.com/p ... rom-black/

It's for EPROM, the older type of 27xxx series memory, which requires an EPROM programmer to write, and UV light to erase. Note that if you do go with a socketed chip solution, it doesn't fit back in the cart shell. And that most older boards were designed for maskROM with a different pinout than EPROM, so often rewiring is needed (for 128kB ROMs in most cases that also means fitting 32 pin chip into a 28 pin socket).

Retroblaster sounds closer to what you might want, as it at least supports writing to (specially designed) carts with mappers. But it's for NES carts only, I don't know the details exactly but I imagine it's the the same technique what CopyNES does, where it repurposes an unused pin to select between mapper/memory writes. Though there's no VRC6 cart for it, it does have support for the most common mappers.
https://retrostage.net/?product_cat=nes

So in other words, I'm pretty much wasting my time with this kazzo device as nothing I have will work with it?
Might as well return the thing while I still can...

[Memblers]

Well, the NROM one should be able work, assuming pin 31 of both the flash chips are connected to the edge connector. It's AMIC A29F040B, probably similar to AM29F040B. The script used for it might need to specifically support that chip. I was thinking people usually use 39SF040 like I do on my boards, and I use the INL-Retro with those, but I don't know what people typically use for kazzo cart writing.

[Killbox91]

I am attaching a wiring diagram and installation drawing.

[TakuikaNinja]

The kazzo & NROM board arrived today. The dumper works fine with the Linux program (needs elevated USB permissions) or within a Windows VM, though most of the scripts floating around online seem to rely on the Windows anago_wx program or its mod. I managed to dump all of the remaining carts I had, with the following exceptions:

• It does not support the Kaiser bootleg of Exciting Basket. I'll elaborate on that in the proper thread.
• Save dumping is flaky. Attempting to dump the save data from Lagrange Point (VRC7) threw an error, and there does not seem to be a save dump script for MMC5/EKROM (Just Breed).

As for the burner part...
The NROM board is the same as what Jdoesstuff has shown but the flash chips are marked as MX29F040PC (-90 for PRG, -70 for CHR, if that matters at all). I have not been able to flash the board (VRAM mirroring is always incorrect, PRG compare always fails at 0x4000), and these are the likely factors:

• The flashdevice.nut file from bakutendo contains the similar MX29F040TC/CTI chips but might not actually be compatible.
• The nrom.af/nrom_wx.af scripts may not be compatible.
• The datasheet for the MX29F040PC claims its size is 512K * 8 bit bytes, which is quite wasteful for this purpose. Is this board NROM-128, NROM-256, or something else?

[Joe]

VRAM mirroring is always incorrect

Even if you flip the H/V switch on the board to the other position?

The flashdevice.nut file from bakutendo contains the similar MX29F040TC/CTI chips but might not actually be compatible.

If it says "id_manufacurer = 0xc2, id_device = 0xa4" it's supposed to work.

The datasheet for the MX29F040PC claims its size is 512K * 8 bit bytes, which is quite wasteful for this purpose. Is this board NROM-128, NROM-256, or something else?

I'd guess NROM-256. Is CPU A14 connected to PRG A14?

[TakuikaNinja]

Even if you flip the H/V switch on the board to the other position?

If it says "id_manufacurer = 0xc2, id_device = 0xa4" it's supposed to work.

I'd guess NROM-256. Is CPU A14 connected to PRG A14?

Yes to all of those. I believe the mirroring being wrong is due to this function in anago (script_flash.c line 40):

Code: Select all

static SQInteger vram_mirrorfind(HSQUIRRELVM v)
{
	struct anago_driver *d;
	SQRESULT r =  qr_userpointer_get(v, (SQUserPointer) &d);
	if(SQ_FAILED(r)){
		return r;
	}
	if((d->vram_connection() == 0x05) != (d->vram_mirroring == MIRROR_VERTICAL)){
		puts("warning: vram mirroring is inconnect");
	}
	return 0;
}

I just checked the flash chip markings in detail. I have G021836 and J020950, which mean the date codes (according to Macronix's documentation) refer to weeks 18 and 09 of 2002 - about as old as I am. The data retention is rated for 20 years for these things, and who knows how many erase cycles have been done on them... I think it's pretty clear the seller just put in whatever bulk stock they accumulated.

Thankfully my board has the chips in sockets instead of being soldered on, so I guess I could replace them with newer chips.

[Joe]

Thankfully my board has the chips in sockets instead of being soldered on, so I guess I could replace them with newer chips.

You could try swapping them between the two sockets, maybe that'll make a difference?

In terms of new chips, it should be possible to use SST39SF010A/SST39SF020A/SST39SF040 in place of the PRG ROM (if the flashing scripts support it), but the PPU bus doesn't have enough address lines to in-circuit flash one of those chips.

[TakuikaNinja]

Swapping the chips between sockets made no difference, so both chips are definitely bad.

I guess I could also invest in an actual EPROM programmer if the CHR side can't be flashed in-circuit. It's a low priority, though - I already have the FDSKey for running NROM-style software.

[shougoki]

First off - Hello everyone. I've just created an account to add somethin in the discussion. Hope anybody's still interested after a few months

Swapping the chips between sockets made no difference, so both chips are definitely bad.

Your chips are 99% ok. I've also got Kazzo/NROM cart bundle from Ali. Mine came with AM29040B chips, though. I've been also struggling to make this cart work and what I came up with so far is as follows:
The chips are ok - tested them with a flash programmer. Soldered them onto an MMC3 repro board and they work fine there.
The small IC between memory chips is some kind of XOR gate - it's supposed to redirect 5V to OE pin of PRG ROM when WE pin is low. My cart had a bad solder joint on one of this IC's legs. Soldering it properly didn't help - the IC seems to be broken and constantly delivers 5V to OE pin of PRG ROM (Doh!)
The whole PCB seems to be badly designed, as the WE, CE and OE pins of both ROMs are constantly 5V. It's the same when the cart is inserted into Kazzo and Famicom - no wonder the cart doesn't work at all. Even when I programmed ROMs directly with a flash programmer.
I think I'll try to trace all the ROM pins to the slot connector and if data/address pins are ok, I'll just ground OE/CE pins and make some repro cart out of it.

[TakuikaNinja]

You're right - it seems the surface-mounted NOT gate IC? (which replaces the older 7400 NAND gate setup) is not doing its job. I checked the voltages supplied to the aforementioned pins while powered by the kazzo and they all report 5V. It could be old stock, just like the flash chips.

[yaiba1989]

NROM Kazzo, old cart use 74HC00, new cart use C04F SOT23-5. NROM - CNROM - UNROM (256) - UNROM (512) - AOROM - MMC1 - MMC3 - VRC2 - VRC4 - Mapper 15-45-52-69-176-178-218,... and more.