How do 60 to 72 pin converters work?

Started by Salduchi, March 21, 2021, 03:07:15 pm

Previous topic - Next topic

Salduchi

CPU revision is E
PPU revision is E-0

First revision to support the triangle noise from what I understand.

P

Thank you, so the NTSC NES probably never goes earlier than those two. Very interesting, that's exactly the same version of the chips as in my 1984 HVC-CPU-07 Famicom. :)

Yeah CPU revision E is the first version with the looped/long/periodic noise feature and sometimes used to simulate a triangle musical instrument (not to be confused with the triangle wave voice). So all NTSC NES should support that, this was expected.

The PPU being E-0 was not expected though, as I heard most NTSC NES uses revision G or H. This means register $2004 (OAM data) and the palette are not readable like on revision G and later. If you run certain Codemaster/Camerica games on it you will probably see lots of graphical glitches, as these games tends to rely on reading $2004. In Micro Machines for example, the screen will be shaking in the menus. You can see this in the emulator Mesen by enabling the $2004-not-readable option. This happens on my Famicom too.

emerson

Salduchi had sent his Honeybee adapter to me for modification and was kind enough to let me run some tests and see how it works. With his permission I am posting the results of my work here.

The defeater circuit in the Honeybee adapter was tested on five revisions of the CIC chip.

CIC REV | DATE CODE
--------+---------
3193 | 8641 AC
3193A | 8718 A
6113 | 8741 A
6113A | 8914 7 BA
6113B1 | 9002 5 BA

The defeater circuit will not work if the NES motherboard has a diode tied to the CIC data line coming from the cartridge. My
motherboard is NES-CPU-07 and has this diode installed at C2. I lifted this diode out of circuit and was able to defeat only
the earlier revision CICs 3193 and 3193A.

The resistors appear to do absolutely nothing. I tested with them in circuit and removed completely and saw no functional difference.
There was no measurable change in noise on either VCC or GND lines with the resistors in or out of circuit. No resistance could
be measured between any of the resistors and the capacitor with my Pinball cartridge installed.

3193
Will start almost every time on start-up.

3193A
Will start almost every time on start-up.

6113
Could not boot game.
One long and one short period between console resets.

6113A
Will boot game but only after several attempts. Many times it will not boot.
One long and one short period between console resets.

6113B1
Will boot game but only after several attempts.
If it doesn't boot within a few seconds it seems like it never will.
Occasionally it will hang in a reset state and the power LED will be off.


Here is the schematic:
You cannot see attachments on this board.


Here are some interesting waveforms captured on power-up with different revision CICs installed.

3193A
You cannot see attachments on this board.

6113
You cannot see attachments on this board.

6113B1
You cannot see attachments on this board.

emerson

This is an image measuring either side of the capacitor. The pink signal is PPU A12 and the blue signal is the output to the console CIC. There is some noise in the signal from my soldering and desoldering irons.

You cannot see attachments on this board.

P

Heh, I didn't know the stun circuit was so unreliable. It was a good thing that you got rid of it.

Salduchi

This is great! Thank you for these results, Emerson!