donluca
Ninja Combat Warrior

- Joined
- Aug 19, 2015
- Posts
- 529
Hi everyone,
I'll begin this thread saying that I've already done the original stereo mod (the passive one, without the transistors) and I have to turn up the volume on my amplifier quite a bit compared to other consoles, but nothing new here so far.
So I was looking at the board right now and I saw that the signal is already amplified by the C844G quad opamp. First thing I wondered was: why the hell is it so quiet? Is the opamp running underpowered or something?
Looking at the spreadsheet it should output enough juice, on par with other consoles.
Anyway, let's say that this is how it was supposed to work and leave it at that.
Next up, is MKL improvement over the original mod (which, by the way, is just the MV-1FS summing circuit, nothing new here).
As you can see, the YM2610 output gets summed and then amplified with the left and right channel.
I can tell you right now: this is wrong.
Look at the MV-1FS schematic:
(big image)
As you can see, the left and right channel get amplified first by the 4558 dual opamps and THEN they get summed with the YM2610.
Now, with that out of the way, I'd like to talk about the choice of using a couple of NPN transistors over the original opamps used in the MV-1FS board, the 4558s, which are still available and are quite cheap.
I know a lot of you may not care for 100% accuracy and just want to get stuff done (which is great!), but I'd like to reproduce the original audio output circuit since the parts used are still available.
This means that it will not be as easy as doing the original or MKL improved mod, since we'll be "copy-pasting" the entire circuit block you can see on the right of the YM3016 (which involves the use of 3 4558 opamps and a whole bunch of other components) onto a new board, but at least we're going to have the original ("authentic", if you prefer) circuit used in the stereo MVS boards.
I'm probably doing it since I have some spare time these days.
First step will be lifting YM3016's legs from 10 to 16, so I can wire them to the new audio output board, along with YM2610's leg 27 (because I don't have a MV-1FZ schematic and I don't trust the board) to tap the other signal and send it to the new board.
Note that we could have lifted only legs 10 and 11 from the YM3016 and leave the other 5 there. Those are used for biasing and buffering the chip and we could have left the C844G opamp doing that, but since we're here doing this, we might go down the rabbit hole all the way.
I'll post pics of the work in progress!
EDIT: some useful resources for those interested:
Schematics of (I believe) a Neo Geo MVS MV-1FS and an AES (courtesy of NeoGeoDev)
https://wiki.neogeodev.org/index.php?title=Schematics
Yamaha YM3016 DAC datasheet:
http://pdf.datasheetarchive.com/datasheetsmain/Datasheets-8/DSA-151908.pdf
Texas Instrument RC4558 Dual Operational Amplifier datasheet:
http://www.ti.com/lit/ds/symlink/rc4558.pdf
NEC C884G Quad Operational Amplifier datasheet:
http://pdf1.alldatasheet.com/datasheet-pdf/view/163079/NEC/UPC844G2.html
I'll begin this thread saying that I've already done the original stereo mod (the passive one, without the transistors) and I have to turn up the volume on my amplifier quite a bit compared to other consoles, but nothing new here so far.
So I was looking at the board right now and I saw that the signal is already amplified by the C844G quad opamp. First thing I wondered was: why the hell is it so quiet? Is the opamp running underpowered or something?
Looking at the spreadsheet it should output enough juice, on par with other consoles.
Anyway, let's say that this is how it was supposed to work and leave it at that.
Next up, is MKL improvement over the original mod (which, by the way, is just the MV-1FS summing circuit, nothing new here).

As you can see, the YM2610 output gets summed and then amplified with the left and right channel.
I can tell you right now: this is wrong.
Look at the MV-1FS schematic:
(big image)
As you can see, the left and right channel get amplified first by the 4558 dual opamps and THEN they get summed with the YM2610.
Now, with that out of the way, I'd like to talk about the choice of using a couple of NPN transistors over the original opamps used in the MV-1FS board, the 4558s, which are still available and are quite cheap.
I know a lot of you may not care for 100% accuracy and just want to get stuff done (which is great!), but I'd like to reproduce the original audio output circuit since the parts used are still available.
This means that it will not be as easy as doing the original or MKL improved mod, since we'll be "copy-pasting" the entire circuit block you can see on the right of the YM3016 (which involves the use of 3 4558 opamps and a whole bunch of other components) onto a new board, but at least we're going to have the original ("authentic", if you prefer) circuit used in the stereo MVS boards.
I'm probably doing it since I have some spare time these days.
First step will be lifting YM3016's legs from 10 to 16, so I can wire them to the new audio output board, along with YM2610's leg 27 (because I don't have a MV-1FZ schematic and I don't trust the board) to tap the other signal and send it to the new board.
Note that we could have lifted only legs 10 and 11 from the YM3016 and leave the other 5 there. Those are used for biasing and buffering the chip and we could have left the C844G opamp doing that, but since we're here doing this, we might go down the rabbit hole all the way.
I'll post pics of the work in progress!
EDIT: some useful resources for those interested:
Schematics of (I believe) a Neo Geo MVS MV-1FS and an AES (courtesy of NeoGeoDev)
https://wiki.neogeodev.org/index.php?title=Schematics
Yamaha YM3016 DAC datasheet:
http://pdf.datasheetarchive.com/datasheetsmain/Datasheets-8/DSA-151908.pdf
Texas Instrument RC4558 Dual Operational Amplifier datasheet:
http://www.ti.com/lit/ds/symlink/rc4558.pdf
NEC C884G Quad Operational Amplifier datasheet:
http://pdf1.alldatasheet.com/datasheet-pdf/view/163079/NEC/UPC844G2.html
Last edited: