SNES/Famicon PCB Revisions and RGB Video

[Link83]

Btw: at the weekend I looked into the ghosting issue of the S-CPUN. The user rama (not registered here) hints me into the direction that the issue is related by overshoots in the RGB signals. This can be seen here: yellow line is the signal coming out of the S-CPUN and the blue one after the THS7314 with load. Shown is a jump from 0 to max.

Spoiler

The overshoots can be reduced by adding a resistor between the AVCC pin of the S-CPUN (pin 155) and Vcc. A resistor of 10 ohms seems to be enough, higher values just further reduce the amplitude levels. However, the overshoots cannot be completely avoided (at least not in my SNES setup) with the simple resistor fix.
As mentioned the resistor also reduces the brightness. In my SNES (SNSP-CPU-1CHIP-01) 20 ohms reduced the Vpp at R, G and B to 0.7V as wanted. Also all Vpp were equal with this resistor, which was not the case before. I tested with a potentiometer.

Here is how my SNES looks at the moment. (I haven't a 20 ohm resistor, so I took a 30 ohm one)

Spoiler

Just as a note
A warning: lifting pin 155 at the S-CPUN can be a tricky and one can easily broke something. The cartridge slot is right in the way.

I just came across this post and wondered if lifting S-CPUN pin 155 and connecting it to a 20 ohm resistor is currently the most 'correct' way to fix the Vpp levels on a 1CHIP SNES? (and as a by-product also reducing the ghosting)

Also, is there any risk of damage to the S-CPUN by adjusting the AVCC level in this way?

I also read borti4938's post here:-

I wanted to redesign it anyway. Hence I can add a solderpad for disabling the filters anyway.
At the moment I'm thinking about leaving away the brightness adaptation from and keep it outside e.g. by soldering additional resistors on top of R6, R7 and R8 or replacing them or by soldering a resistor in front of AVCC.
What do you think?

I would be very interested in an optional 1CHIP RGB Bypass board revision without the input resistors.

It would also be great if all the RGB Bypass boards used a THS7374 (instead of a THS7314) with a jumper pad to disable the LPF.

Hope you will consider it borti

[sofakng]

Hey all... I'm new to retro gaming and RGB output, but I've gone ahead and purchased a 1CHIP-03 SNES. (I have no interest in the Mini or Jr. because I always had the regular SNES as a kid)

Is it correct that to obtain the best RGB output that I should install a THS7374 amp? Does that also restore CSYNC?

Is this the correct Borti board for adding the amp to the 1CHIP-03? If so, how do I disable the filter which it also sounds like I need to do for the best quality?

[borti4938]

At the moment I'm at this point:

Spoiler

This board is intended to be used on every SNES - either 1Chip or the good old 3Chip. A quick review through the properties:

• THS7374
  • 4 Channel buffer for red, green, blue and /CSYNC
  • LPF enabled by default but can be disabled by closing J1
• Input to the THS7374 (R, G and B):
  • Opportunity to add brightness correction with R11, R21 and R31 (ONLY need for 1Chip SNES and ONLY if AVCC pin is NOT lifted and reconnected to Vcc with a resistor (as shown above))
  • DC part of the signals first removed C11, C21 and C31 (loosely say: signals are decoupled from the SNES now)
  • Then another DC offset added with R12, R22 and R32
• Output from THS7374 (R, G and B):
  • 75 Ohms (39 Ohms in PAL systems) in series (R13, R23 and R33)
  • optionally 47pF from MultiAV points to GND (missing on SNES mini; C14, C24 and C34)
• Input to the THS7374 (/CSYNC):
  • Signal dropped down to Vpp of 0.7V (values / pulses from 0.7V down to 0V and back) using a simple voltage divider (R41 and R42)
  • 2.3V of Vpp also possible by changing values of R41 and R42 (highest possible Vpp input to the THS7374)
• Output the THS7374 (/CSYNC):
  • big cap to remove the DC component (needed on some modern TVs; my Samsung e.g.) which can be optionally left out (there is a jumper below)
  • 75 ohm in series to the MultiAV
  • optionally 47pF from MultiAV points to GND (missing on SNES mini and 1Chip-03); C44)
  • BE CAREFUL:
    • /CSYNC goes to pin 3 of the MultiAV.
    • In PAL systems there is 12V
    • On all NTSC systems except 1Chip-03 and SNES Mini / Jr. there is the stock /CSYNC
    • if pin 3 is not free, free pin 3 or remove C43, R43 and C44 from the modding board

Here is a picture how it looks installed in a Mini (don't look on the blue PCB - this is a multiregion mod which has nothing to do with the stuff discussed here):

Spoiler

I used again the option to lift AVCC and reconnect over a resistor to VCC

(unfortunately I only had a SMD 0603 resistor )

I guess I will publish the design files quite soon.

I also had in mind the places for R11, R21 and R31 can be used for a small cap (e.g. 220pF) to remove the overshoots (see this post here) on the RGB outputs a bit more. But I haven't tried it yet.

I just came across this post and wondered if lifting S-CPUN pin 155 and connecting it to a 20 ohm resistor is currently the most 'correct' way to fix the Vpp levels on a 1CHIP SNES? (and as a by-product also reducing the ghosting)

I my point of view this is as good as the brightness correction using the resistors. But as it mentioned, it is the only method so far to remove the ghosting. I have played Super Mario World at the weekend on a 1Chip console without the 20ohm resistor. Especially in level 4 I was extremely annoyed by the ghosting such that I switched of the console, got my console with the pre-resistor installed and went on with gaming. I definitely have to do the mod in my 'main' SNES, too

Also, is there any risk of damage to the S-CPUN by adjusting the AVCC level in this way?

No! The only risk to damage the S-CPUN is to damage the S-CPUN (or the mainboard) during the modding work. Due to the cartslot there is not much space to work on pin 155. It's quite risky to damage neighboured pins, pads on the mainboard or to brake of the pin.

I would be very interested in an optional 1CHIP RGB Bypass board revision without the input resistors.

It would also be great if all the RGB Bypass boards used a THS7374 (instead of a THS7314) with a jumper pad to disable the LPF.

Both points considered

Is this the correct Borti board for adding the amp to the 1CHIP-03? If so, how do I disable the filter which it also sounds like I need to do for the best quality?

Wait for the 'new' version.

[tjstogy]

At the moment I'm at this point:

Borti, you're literally a genius. Question- how does this differ from the older version aside from the 7374 chip? Also, maybe you could make a deluxe version with IGR....

[Einzelherz]

What's all these add-on color chips you kids are doing?

[borti4938]

Do you mean the chips on the blue PCB? The blue PCB is just a multiregion modification (SuperCIC, uIGR, regionpatch and clock circuit). I have added a comment to the picture that the PCB has nothing to do with the RGB stuff discussed here.

At the moment I'm at this point:

Borti, you're literally a genius. Question- how does this differ from the older version aside from the 7374 chip? Also, maybe you could make a deluxe version with IGR....

You switching the LPF on and off using the uIGR. Possible, but with the current chip there is no pin free for that. Anyway - I guess 99% of all people set the LPF on or off, respectively, according to their 'feeling' how it looks better on the current setup.

Unfortunately I don't know how the DAC inside the S-CPUN is designed. If there are filters inside, the LPF of the THS7374 can be switched off. If not, the LPF has to be left switched on to have it 'correct'.

[Einzelherz]

I meant the one you described that does something to the output.

[Harrumph]

This looks amazing, like always Borti. Maybe it was mentioned earlier, I must have missed it, but what would be the purpose of using this on a multi-chip SNES (you write it can be used for all SNES revisions)?

[sofakng]

Will anybody be selling these chips pre-assembled? It looks like they contain several SMDs which are difficult for beginners (like myself!).

I believe one of the websites (retro fixes?) was planning to sell pre-assembled chips but I'm wondering if they would be using the older board design, etc.

[bobrocks95]

I'm super late to this thread, but I did my best to skim through it all- is this about an RGB amp that bypasses the built-in ones on the multi-chip and the 1-chip consoles? Does it make the multi-chip look about as good as a 1-chip, and what's improved on the 1-chip (which I thought everyone thought looked great)? These are probably all dumb questions, but I'm really late to the party here...

[Link83]

Wow the new V4.0 looks brilliant borti4938! It seems to do everything I wanted from a new RGB Bypass board

I just have one suggestion though - I realise this may make the PCB slightly larger, but maybe add a jumper to enable the C-Sync output? Otherwise I can see some people not understanding the installation instructions or forgetting to remove one of the components (C43, R43 and C44) which will either connect the stock C-Sync to the new C-Sync, or even worse to 12V! Since only two motherboard revisions need the C-Sync replaced (SNES Mini and 1CHIP-03) It would make sense to have a jumper.

...I also would love to see a revision of this board for the early model N64's with a VDC-NUS chip

[Seraphic]

Hi borti,

LPF enabled by default but can be disabled by closing J1

You switching the LPF on and off using the uIGR

So the LPF can be disabled permanently by closing "J1" or turned on/off by using user-in-game-reset?
And uIGF is a additional mod?

Wait for the 'new' version.

So there will be a new version of this RGB THS7374 board soon then?

[ApolloBoy]

Will anybody be selling these chips pre-assembled? It looks like they contain several SMDs which are difficult for beginners (like myself!).

I believe one of the websites (retro fixes?) was planning to sell pre-assembled chips but I'm wondering if they would be using the older board design, etc.

Seconding this. I'd love to finally have sharp RGB on my SHVC-CPU-01 Super Fami.

[mvsfan]

Managed to buy a Super Famicom with serial S2502XXXX for a measly US$19 on ebay. It would most likely be a 1CHIP-02?


No. That console is not a 1-chip. they start at UN301. you most likely have an APU.

Oh, nevermind. YOUR super famicom probably IS a 1-chip.

[Sid]

Managed to buy a Super Famicom with serial S2502XXXX for a measly US$19 on ebay. It would most likely be a 1CHIP-02?


No. That console is not a 1-chip. they start at UN301. you most likely have an APU.

Oh, nevermind. YOUR super famicom probably IS a 1-chip.

Hehe, you had me confused until I realised that you had yourself confused.

I don't have the tool to open it up and check the exact version, but it's certainly a 1CHIP based on the image I get vs a definite non-1CHIP. There's certainly no carefully inspecting the little details to squeeze out a perceivable difference - it's absolutely night and day.

I am left wondering if a brightness mod should be on the cards though. The brightness borders on being unacceptably garish.

[mvsfan]

yeah at first i was going on a US snes serial no. then i saw super FAMICOM.
heh.

yeah if your getting a real good image, its likely a 1-chip.

the tool to open it is not expensive. Look up Gamebit screwdriver on ebay.

I prefer getting a 1-piece screwdriver rather than the bits because on some consoles the bits dont fit right.

[RGB32E]

Looks like Borti just released his V4 Bypass PCBs on OSHPark.

SNES RGB Bypass Board V.4a // RGB+CS Version
SNES RGB Bypass Board V.4b // RGB Only Version

More info on GitHub

[ApolloBoy]

I prefer getting a 1-piece screwdriver rather than the bits because on some consoles the bits dont fit right.

A lot of those bits are also cheaply made and strip very easily.

[bobrocks95]

Someone have mercy I still don't know what this is

Will 3-Chips magically look like 1-Chips? What will 1-Chips look like?

[Harrumph]

Will 3-Chips magically look like 1-Chips?

According to this thread, no. The image will be "better" but not with sharper pixels.
Maybe Borti can put up some comparison pics?

One thing it can do is give csync to PAL consoles. But otherwise, yeah I'm a bit unsure if the install effort would be worth it...

[sofakng]

Is anybody going to pre-assemble and sell these new boards from Borti?

I'm decent at soldering but I've never attempted any SMD components and I'd rather purchase something already assembled.

[bobrocks95]

Will 3-Chips magically look like 1-Chips?

According to this thread, no. The image will be "better" but not with sharper pixels.
Maybe Borti can put up some comparison pics?

One thing it can do is give csync to PAL consoles. But otherwise, yeah I'm a bit unsure if the install effort would be worth it...

Ah, thanks. It will be interesting to see some before and after pictures. I guess a PCB for the sharp pixels mod is what I should be waiting on if I want 1-Chip-like results.

[borti4938]

I've tested the sharp pixel mod in a PAL and a NTSC SNES if you mean that on here.
Read on Circuit-Board this post (German) and the following. Or to keep it short, I gave up with this mod due to the following reasons:
- noisy / shaky image on all of my TV sets
- high impact on the composite video output - the output became useless
- hardly understandable

This was the point where I started a second try with the THS7314 / THS7374. The S-PPU2 is not designed to drive low loads. This is the reason why the 'Bypass' board always sits behind the first or second transistor after the S-PPU2s RGB outputs.

Will 3-Chips magically look like 1-Chips?

According to this thread, no. The image will be "better" but not with sharper pixels.
Maybe Borti can put up some comparison pics?

One thing it can do is give csync to PAL consoles. But otherwise, yeah I'm a bit unsure if the install effort would be worth it...

Correctly summarized! To make it magic, one might need a sample and hold circuit. (An approach by micro at Circuit-Board.de (also in German))

Unfortunately, my before-and-after pictures with the board were useless. So I throw them away. I also only have one 3Chip-SNES at home at the moment. But once I have a second, I will make a second try for before and after. However, everybody who has the equipment to make professional comparison pictures feel free to make some!

So the LPF can be disabled permanently by closing "J1" or turned on/off by using user-in-game-reset?

If you don't have a region patch installed, you can use pin 7 for switching on and off the LPF. In general, most people do not change it.
If you want to have an extra pin, I have to change the PIC (what I don't want at the moment).

Looks like Borti just released his V4 Bypass PCBs on OSHPark.

SNES RGB Bypass Board V.4a // RGB+CS Version
SNES RGB Bypass Board V.4b // RGB Only Version

More info on GitHub

Yeah, finally
At the moment, my free time is very spare. So I'm happy that I found the time to write the READMEs. I hope that I haven't forgot anything and most of it is understandable. If not, just forward questions to me or the forums where I'm around. I will then try to improve the instructions.

[Link83]

Nice work borti4938

In the readme I noticed this:-

- for the 'regular' 1Chip model: http://retrorgb.com/snes1chip.html (using 750 Ohm resistors)
- for the SNES Mini/Jr.: http://retrorgb.com/snesminipremade7314.html (using 1.2k Ohm resistors)

I have read the assemblergames thread and i'm still confused why the 1CHIP-01/02/03 require different resistor values to the SNES Mini/Jr. They all use the same S-CPUN and S-RGB chips, so why do they require different resistor values?

Also might you have any suggestions on the correct capacitor value/location to reduce the vertical line issue? I have read multiple different solutions in different threads, but believe that adding a capacitor between +5V and Ground on the 7805 may be the best solution?

[mickcris]

was just building up a 4a for a mini and noticed i did those resistors wrong. the prep txt files have them backwards

2. Special components (for 1Chip-03 and SNES Mini/Jr., read notes):
a) for 1Chip consoles (not Mini/Jr.)
- R11, R21, R31: [1]
3x 1.2kOhm resistor 0603 package (0.125W, 1% tolerance should be fine)
- C44: 47pF/50V ceramic capacitor 0603 package [2,4]

b) for SNES Mini/Jr.
- R11, R21, R31: [1]
3x 750Ohm resistor 0603 package (0.125W, 1% tolerance should be fine)
- C14, C24, C34, C44:
4x 47pF/50V ceramic capacitor 0603 package [4]


thanks for making these boards borti! they work great!

[borti4938]

My fault - SORRY! I will correct it very soon THANKS YOU for reporting

- for the 'regular' 1Chip model: http://retrorgb.com/snes1chip.html (using 750 Ohm resistors)
- for the SNES Mini/Jr.: http://retrorgb.com/snesminipremade7314.html (using 1.2k Ohm resistors)

I have read the assemblergames thread and i'm still confused why the 1CHIP-01/02/03 require different resistor values to the SNES Mini/Jr. They all use the same S-CPUN and S-RGB chips, so why do they require different resistor values?

The SNES Mini/Jr. has a higher load applied to the S-CPUN RGB outputs than the big model 1Chip consoles (150ohm vs. 160ohm)

Also might you have any suggestions on the correct capacitor value/location to reduce the vertical line issue? I have read multiple different solutions in different threads, but believe that adding a capacitor between +5V and Ground on the 7805 may be the best solution?

I never had the issue. But I read something about 470uF (min. 16V). Also, soldering a 1000uF (min. 25V) to C58 (big model) / C52 (Mini) in the US consoles might be helpful!

[chomel]

Ok, first of all, FANTASTIC WORK. I've read that Borti is working on something new on this thread:http://shmups.system11.org/viewtopic.php?f=6&t=58863
For future boards, would it be possible on a PAL SNES to connect CSYNC to the Amp and to direct it to pin-9 of the multi AV out to replace composite signal?

If not, would it be possible to direct CSYNC signal directly to the multi AV out? I know PPU-2 outputs CSYNC on pin 100 which is directed to the video encoder. But on PAL SNES, CSYNC is replaced by 12V.

[Turom]

I don't know who maintains this spreadsheet but it helped me a lot for my search, so here's my contribution that may be added to the list:

• S24941268: SNS-CPU-APU-01
• S25116727: SNS-CPU-1CHIP-02

Il will soon make some capture with my XRGB Mini to compare the two once I recieve the PSUs.

[chomel]

I see the new bypass on Borti github have been online for a month. I'd like to mod a 3chip SNSP but I have a question.

For 3-chip, it's written SNES_RGBAmp_wCSYNC: possible on 3Chip boards (but not tested / described here). But in the How to install section, there are only informations to connect R, G and B to the SNES motherboard. But nothing about Csync. Is it because, you can't use the 4.1a bypass for 3-chip or did I miss something?

[syboxez]

I see the new bypass on Borti github have been online for a month. I'd like to mod a 3chip SNSP but I have a question.

For 3-chip, it's written SNES_RGBAmp_wCSYNC: possible on 3Chip boards (but not tested / described here). But in the How to install section, there are only informations to connect R, G and B to the SNES motherboard. But nothing about Csync. Is it because, you can't use the 4.1a bypass for 3-chip or did I miss something?

All NTSC SNESs except for the 1CHIP-03 and the Mini have CSync already outputting on pin 3 of the multi out. PAL consoles instead have 12v on pin 3 as is needed by some PAL TV's SCART inputs.