2600 in 2006

I'm interested. (I'm not quite ready to take out a second mortgage to pay for a CC2 )

 

For the cable, can you get the +5v from the parallel port, or do you need an external supply?

 

And are you using JTAG to program it?

I'm interested.  (I'm not quite ready to take out a second mortgage to pay for a CC2 )

 

Aw, CC2 is awesome.

 

For the cable, can you get the +5v from the parallel port, or do you need an external supply?

 

Probably depends how beefy your printer port is. Using a battery and regulator seemed better than expecting 20+mA from a printer port.

 

And are you using JTAG to program it?

 

I'm using boundary scan on the CPLD to program the flash, but with /OE and /WE wired to the printer port.

What would be the price of a game on one of these boards? Who would assemble it?

 

Sounds cool.

What would be the price of a game on one of these boards?  Who would assemble it?

 

Those things remain to be seen. If Albert could handle surface-mount stuff that'd be wonderful, but I suspect he'd probably rather not.

What would be the price of a game on one of these boards?  Who would assemble it?

 

Those things remain to be seen. If Albert could handle surface-mount stuff that'd be wonderful, but I suspect he'd probably rather not.

Surface mount stuff is beyond the abilities of the majority of hobbyists, myself included.

 

However, I can do PLCC, which might be a good compromise. Although a PLCC version would take up more board real estate and require additional cost due to the sockets, the tradeoff may be worth it to save the cost of having it professionally assembled, and may end up costing less overall.

However, I can do PLCC, which might be a good compromise.  Although a PLCC version would take up more board real estate and require additional cost due to the sockets, the tradeoff may be worth it to save the cost of having it professionally assembled, and may end up costing less overall.

 

Well, the surface-mount chips are PLCCs, but there are also a few resistors and caps and such. It might be possible to do the board as throughhole using a socketed PLCC for the Xilinx and trying to find DIP parts for the flash and RAM, but assembly labor would be substantial doing that also.

 

One thing I'm trying to decide for my next (and hopefully final) board rev is whether to allow the device to access 96K or 128K of a 128K flash chip. I just noticed that the 128Kx8 chips are actually cheaper than 64Kx8, so there's no reason not to use one. The issue would be addressing, since my Xilinx part is pretty full.

 

If I just wire A16 of the flash to A11 of the cartridge port, that would allow 96K of the flash to be addressed. 32K would only be addressible via the $1000-$17FF bank and the other 64K only addressible via the other banks. Because the memory is ROM, information could be duplicated in the upper and lower regions if code needed to be able to access it in multiple banking areas.

 

Increasing the bottom area to 64K could be done via one of two ways:

 

Way #1 which would certainly fit would be to use the I2C output latch to control A15 on accesses to the $1000-$17FF bank. This would be rather ugly, but would work.

 

Way #2 would be to expand the four-bit lower banking latch to five bits. This would be nicer, if it fits, but it would be tight.

 

My one concern with doing these things is that unless users always use 128K binary files (sorta ugly) the device programming the chip would have no idea which data would be accessed in the upper bank and which in the lower; it would thus have to program the same data in both. Certainly doable, but it would double the time required for programming (my nice laptop, running DOS, takes about 2K per second).

 

Would anyone be interested in 128K carts?

However, I can do PLCC, which might be a good compromise.  Although a PLCC version would take up more board real estate and require additional cost due to the sockets, the tradeoff may be worth it to save the cost of having it professionally assembled, and may end up costing less overall.

 

Well, the surface-mount chips are PLCCs, but there are also a few resistors and caps and such. It might be possible to do the board as throughhole using a socketed PLCC for the Xilinx and trying to find DIP parts for the flash and RAM, but assembly labor would be substantial doing that also.

Yeah, socketed through-hole is what I meant. Sure, professional assembly of through-hole stuff would be just as costly, but at least this way it would let more of us assemble everything at home.

My one concern with doing these things is that unless users always use 128K binary files (sorta ugly) the device programming the chip would have no idea which data would be accessed in the upper bank and which in the lower; it would thus have to program the same data in both.  Certainly doable, but it would double the time required for programming (my nice laptop, running DOS, takes about 2K per second).

 

Would anyone be interested in 128K carts?

Why not?

 

Anyway, using socketed chips (PLCC or otherwise) those of us with suitable device programmers (like me) could just pop the flash chips in and program them in seconds... right? Besides, a minute is not that long.