Q dropped by for approximately half a second to post this at #nesdev:
http://www.qmtpro.com/~nes/visual2a03/
It's a transistor level JavaScript simulator of 2A03, based on the 6502 simulator that was posted at visual6502.org some time ago. Pretty cool stuff.
So this is 100% accurate? Sweet, cool stuff. I need to get into this "visual" stuff....
Just messed with it. What happened to the instruction step through? I remember at one time just inputting some instructions and showing the cycle-by-cycle operation, but it seems to have changed since then, what's the easiest way to do this?
3gengames wrote:
So this is 100% accurate? Sweet, cool stuff. I need to get into this "visual" stuff....
From
http://www.qmtpro.com/~nes/chipimages/Quote:
I have achieved success: The Visual 2A03. Note that this currently contains a real 6502, so don't be surprised when decimal mode arithmetic actually works.
I've successfully tested both Sprite DMA and simple operation of the main Square channel, but it's quite possible that there are errors (shorts, missing vias) that might cause other problems - if you find any, please report them to me.
thefox wrote:
3gengames wrote:
So this is 100% accurate? Sweet, cool stuff. I need to get into this "visual" stuff....
From
http://www.qmtpro.com/~nes/chipimages/Quote:
I have achieved success: The Visual 2A03. Note that this currently contains a real 6502, so don't be surprised when decimal mode arithmetic actually works.
I've successfully tested both Sprite DMA and simple operation of the main Square channel, but it's quite possible that there are errors (shorts, missing vias) that might cause other problems - if you find any, please report them to me.
Close enough for me. Now to start tracing stuff...
Cool! Now everyone can test all sorts of little obscure behaviors! (I'm serious BTW, in case someone thinks that was sarcasm)
Now what I really, really wish is a visual 2C02. That would be awesome!
tokumaru wrote:
Cool! Now everyone can test all sorts of little obscure behaviors! (I'm serious BTW, in case someone thinks that was sarcasm)
Now what I really, really wish is a visual 2C02. That would be awesome!
That is, if there's no error in the process to convert (possibility of missing via or short he said)
So how the hell do access the assembler in this and what is the syntax mostly like?
3gengames wrote:
So how the hell do access the assembler in this and what is the syntax mostly like?
There is no assembler in it. *You* are the assembler. You need to put assembled bytes into the memory.
Pretty damn cool!
BTW, to assemble your own code here are some tips:
$A9 is the LDA #$xx opcode
$8D is the STA $xxxx opcode (remember low byte of address first)
$4C is JMP, if you don't want it to start BRK-dancing
Very cool, and glad to hear Q stopped by if only briefly. Hope he comes back to contribute to our conversations here someday.
cpow wrote:
3gengames wrote:
So how the hell do access the assembler in this and what is the syntax mostly like?
There is no assembler in it. *You* are the assembler. You need to put assembled bytes into the memory.
Well, wait, what is the one that you can type in an instruction and see the cycle-by-cycle operation? Although the guy who hand-codes his games may like this.... @_@
Quote:
Now what I really, really wish is a visual 2C02. That would be awesome!
If you can, please donate some chips to make that happen (PM me for more info).
dfrechet wrote:
Quote:
Now what I really, really wish is a visual 2C02. That would be awesome!
If you can, please donate some chips to make that happen (PM me for more info).
I wish I had some NES PPU's to send your way.
I would maybe post on NintendoAGE, I'm sure some people will send you some NES's for the cost of postage or very cheap that are beat-to-hell or have problems.
Since I live in Brazil, getting original NES hardware is not very cheap. The typical NES with a game, 2 controllers and a zapper will hardly go for less than US$60. I did pay that for the one I have in my collection, but I think the price is too high to cannibalize the PPU.
If I ever find incomplete/defective consoles with fair prices I'll be sure to get them and put their PPUs to good use.
EDIT: I went to check out of curiosity and it seems that I was very optimistic with my $60 estimate. All consoles I could find right now cost quite a bit more than that.
Alright, thanks to willing donators we should soon have enough standard composite units (RP2C02-0) to do a successful decap.
If there is enough interest and someone is willing to sacrifice his/her chip, we might also do one of the RGB units later on, but that lies in the future.
Aren't
these pictures of the decapped PPU?
tepples wrote:
Aren't
these pictures of the decapped PPU?
Yes, but they are of very low quality, the chip broke during decapsulation, and we don't have a delayered image.
According to blargg's docs, the 240Hz APU clock is generated by dividing the 21.47727MHz clock by 89490. Brad Taylor says it's generated by dividing PHI2 (on clock edge) by 14915. Can someone determine the exact behavior from this simulator/chip images?
Look at the recent contributions of User:Quietust. He's been giving notes on talk pages about his observations based on the decapped 2A03.
Thanks. That answers that question, but just leads to more. I really wish I knew how to interpret the chip images
Are there any schematics of the 2A03 made based on these decaps? I didn't see anything obviously that on Q's site.
Not full right now, but it will be completed after 2A03.
Here.
I updated circuits and uploaded PSD sources :
http://breaknes.com/download