NES Logic Levels

skrasms · Post by **skrasms** » Sun Jan 04, 2009 12:41 pm

This is a pretty simple question. What is the threshold voltage for what the 2A03 sees as a logic high coming in?

Bregalad · Post by **Bregalad** » Sun Jan 04, 2009 1:45 pm

NMOS technology (which is fully compatible with CMOS I guess)

skrasms · Post by **skrasms** » Sun Jan 04, 2009 2:59 pm

Bregalad wrote:NMOS technology (which is fully compatible with CMOS I guess)

Is 3.3V logic enough to drive it?

Jeroen · Post by **Jeroen** » Sun Jan 04, 2009 3:09 pm

Might work.......definitly not "correct though" stick to 5v for nes.

Bregalad · Post by **Bregalad** » Sun Jan 04, 2009 3:22 pm

Mmh... You'd definitely want to adds pull-up resistors and/or to test the thing extensively.

I have enough trouble making my 5V board working. The possible sources of error are so numerous that it's a headache (power filtering, good electrical connexion, problem with delays, problem with contacts, solder short-circuit, via touching another componant or solder, etc...).
You don't want to intentionally add more possible errors.

Jeroen · Post by **Jeroen** » Sun Jan 04, 2009 3:30 pm

Ya but then you'd have problems with possible revisions. With some nesses working some dont.....just stick to 5v

skrasms · Post by **skrasms** » Sun Jan 04, 2009 4:45 pm

I was hoping it would be CMOS/TTL compatible. It would save some parts/hassle

I haven't had any problems driving the 2A03 with 5V CMOS, but that's the most ideal possible case.

Looks like it'll have to be open-drain or buffered circuitry.

skrasms · Post by **skrasms** » Sat Feb 14, 2009 10:55 pm

Has anyone ever seen a 2A03 datasheet? I'd love to have one even in Japanese.

HMOS is an improved version of NMOS, isn't it? I noticed the HMOS SID chips that I have are rated for 2V being logic high with a 5V power supply. How different is NMOS?

Post by **Memblers** » Sun Feb 15, 2009 12:26 am

I wasn't aware of HMOS, but according to Wikipedia it's also known as "depletion-load NMOS". The 2 6500 datasheets I just looked at (Rockwell and Commodore) both say it's "N-channel, silicon-gate, depletion-load technology".

2A03 is a clone (excluding the patented BCD mode), I'd assume it's made on the same way..? If that's true, maybe this is accurate enough:
http://www.zimmers.net/anonftp/pub/cbm/ ... 50x-15.gif

skrasms · Post by **skrasms** » Mon Feb 16, 2009 7:25 pm

That sounds promising. I'll give it a test soon and see what magic happens.

Post by **Memblers** » Mon Feb 16, 2009 8:02 pm

Let us know how it goes. There is a lot of stuff that is 3V.

I wonder if the PPU would be the same?

MottZilla · Post by **MottZilla** » Tue Feb 17, 2009 11:38 am

Wasn't part of the issue with the PowerPAK and the FPGA needed being expensive cause it was an older 5V type? Or am I mistaken?

Post by **Memblers** » Tue Feb 17, 2009 2:54 pm

MottZilla wrote:Wasn't part of the issue with the PowerPAK and the FPGA needed being expensive cause it was an older 5V type? Or am I mistaken?

Yeah, 5V parts tend to be less available and more expensive than 3V equivalents (if equivalents exist, often they don't, and maybe never exist for new architectures).

I would think the NES can read a 3V signal is being '1', but you still won't be able to use a 5V output from the NES on a 3V system. So it's just a possible optimization, for some simpler cases.

skrasms · Post by **skrasms** » Tue Feb 17, 2009 7:51 pm

MottZilla wrote:Wasn't part of the issue with the PowerPAK and the FPGA needed being expensive cause it was an older 5V type? Or am I mistaken?

I'm looking at using some CPLDs that have 3.3V outputs and 5V tolerant inputs. As long as the 2A03 can see 3.3V as logic high then there is no problem with communication.

FPGAs are a bit of a different beast. Even 3.3V can be too much for some modern FPGAs. In that case you need a special FPGA or level shifting circuitry so that the 5V from the NES doesn't cause damage.

blargg · Post by **blargg** » Tue Feb 17, 2009 8:49 pm

Can't you just put a resistor in, so that the voltage is reduced to 3.3V via the input protection diodes and the current limited by the resistor? Slew rate will be degraded in both directions, of course.