Does the NTSC 1CHIP-03 need the 750 Ohm brightness mod?

[sofakng]

I've just received an NTSC SNES 1CHIP-03 and I haven't tested the brightness (using the 240p test suite), but according to RetroRGB all 1CHIPs are too bright.

Do I still need to perform the 750 Ohm mod on my SNES?

I'm planning on purchasing Borti's latest RGB Amp board which I think also fixes this, but the latest revision isn't ready so I'm interested in doing this mod first.

[Asbrandt]

Generally, the 750ohm fix to the current sinks off the video DAC is considered the most "correct" fix that remains relatively easy to install and also works for the video encoder, while not changing the output termination of the amp away from the proper 75ohm.

--

If you wish to be more precise; 731ohm is the target value for the full-size 1CHIP systems, 732ohm being the closest available value, if you're planning a Mouser/Digikey order.
(For anyone with a Mini system who may read this post, the target value for those is 1.051K, 1.05K being the closest available value.)

It might seem silly to use closer values that tend to only be available in tight tolerances when the original tolerances were pretty loose, but I do it anyway just to avoid compounding tolerances, since I'm only doing personal one-offs where the increased price of those parts isn't really a big deal.

[sofakng]

Thanks so much for the detailed reply!

I wish I had asked sooner since I already ordered 750 Ohm resistors per the RetroRGB website. (After looking again on DigiKey, it looks like these would be a better match? (Yageo MFR-25FBF52-732R)

Like yourself, I'm only performing mods for myself so I don't mind paying extra for the better components.

[Asbrandt]

It's not a huge deal if you got 750's, if you got a good handful of them just try to sort out a few that are under 750ohm rather than above and it'll be fine.

[Einzelherz]

I don't notice a brightness difference with my SHVC, 1Chip-01, or 1Chip-03.

Edit: never mind my input if this is in reference to the add on RGB bits.

[Asbrandt]

I don't notice a brightness difference with my SHVC, 1Chip-01, or 1Chip-03.

Edit: never mind my input if this is in reference to the add on RGB bits.

Thanks to clamping (I think?) it usually manifests as the colors being washed out instead of overall brightness.

[tjstogy]

The RGB bypass on my 1chip 03 looks simply amazing. I have the 750 ohm hooked up, to a pvm 20l5 and it's almost unfair how good it looks. It looks like an emulator. Problem is, the pvm is so good that it makes my other systems look crappier. The Overscan on the NES and Genesis are always apparent and the n64 looks like shit in its own way. I'm on the lookout for a consumer tv next.

[borti4938]

Generally, the 750ohm fix to the current sinks off the video DAC is considered the most "correct" fix that remains relatively easy to install and also works for the video encoder, while not changing the output termination of the amp away from the proper 75ohm.

But this is not the correct way in my point of view! The best way to fix the brightness is to lift pin 155 of the S-CPUN and reconnect it to Vcc via a 20 Ohm resistor.



(I only had a 30ohms SMD0603 resistor for this but will replace it with a 20ohm resistor soon.)

This also drastically reduces the ghosting issue present with the 1Chip consoles.
(the ghosting isn't noticeable on the screen anymore but I can still measure it with an oscilloscope.)

I've just received an NTSC SNES 1CHIP-03 and I haven't tested the brightness (using the 240p test suite), but according to RetroRGB all 1CHIPs are too bright.

Do I still need to perform the 750 Ohm mod on my SNES?

I'm planning on purchasing Borti's latest RGB Amp board which I think also fixes this, but the latest revision isn't ready so I'm interested in doing this mod first.

If you want to use V3 there a brightness corrector on the board which only effects the RGB.
The next V4, which will come very very soon, has the option to either apply the brightness correction with the three resistors or to apply an external brightness correction (hopefully with the method presented here).

[Pasky]

Removes or reduces? I recall you mentioning this fix before and said it didn't completely eliminate it.

[borti4938]

you are right - it reduces the issue in such a way that it isn't noticeable any more. On the oscilloscope I still can see this but not on the screen anymore
I have edited my post.

[sofakng]

Borti,

Will you be selling assembled versions of your v4 board? I don't have any experienced with SMD soldering and I'm not sure I can assembly it.

Also, what do you suggest I do before I use your v4 board? (ie. should I perform your pin 155 mod, or use 3 resistors?)

[RGB0b]

But this is not the correct way in my point of view! The best way to fix the brightness is to lift pin 155 of the S-CPUN and reconnect it to Vcc via a 20 Ohm resistor.

I've never tried that...and would NEVER recommend a beginner try that either. It's SO EASY to break the pins on the S-CPUN, as they're very brittle. It might be the better solution, but it's definitely "experts-only".

I'd love to test a SNES that had that done though, just to see how it compares. I've had a few people email me saying that adding the 750 ohm resistors helped with the ghosting/ringing. Actually, a few people said that using the THS7374 mod (filter off, with 750 Ohm resistors) completely eliminated the ghosting. They all live pretty far from me, otherwise I'd have liked to try the stock S-RGB encoder and see if the resistors "fixed" the ghosting on that as well.

I wish everyone working on SNES research could all meet somewhere, bring their SNES' and scope all the different revisions and fixes. LOL, maybe we can pick a convention in 2017 and make it the retro tech weekend...

[Asbrandt]

But this is not the correct way in my point of view! The best way to fix the brightness is to lift pin 155 of the S-CPUN and reconnect it to Vcc via a 20 Ohm resistor.

This one is new to me. I'm bad at google and didn't manage to find a pinout of the S-CPUN / RF5A122 anywhere, what's this doing exactly?

[sofakng]

Yeah, the S-CPUN pins look awfully small and I don't think I'm skilled enough (or brave enough) to risk damaging my 1CHIP-03 model.

It looks like that I'll use the resistors for now and then modify as needed when Borti releases v4 of his board.

[borti4938]

But this is not the correct way in my point of view! The best way to fix the brightness is to lift pin 155 of the S-CPUN and reconnect it to Vcc via a 20 Ohm resistor.

I've never tried that...and would NEVER recommend a beginner try that either. It's SO EASY to break the pins on the S-CPUN, as they're very brittle. It might be the better solution, but it's definitely "experts-only".

Definitely only for "experts". The major problem here is that there is extremely tight space between the S-CPUN and the cartridge slot. Even a wired resistor hardly fits in between.

Here is what I did a few minutes ago

Spoiler

I've had a few people email me saying that adding the 750 ohm resistors helped with the ghosting/ringing. Actually, a few people said that using the THS7374 mod (filter off, with 750 Ohm resistors) completely eliminated the ghosting. They all live pretty far from me, otherwise I'd have liked to try the stock S-RGB encoder and see if the resistors "fixed" the ghosting on that as well.

Well, that is absolutely against from what I experienced!
It's not a long time ago where I modded a SNES Mini. In my first try I used the 'three-resistor-method'. I played SMW and in the fourth world (bridge level) I was annoyed by the ghosting that I stopped playing.
By looking at the oscilloscope I've seen that the 'three-resistor-method' has no impact on the ghosting (one can see it by observing over-/undershoots from low to high values and vice versa). So I replaced the 'three-resistor-method' by the single resistor at S-CPUN pin 155.
BTW: the filter of the THS7374 has nothing to do with ghosting That's another effect what we can see here. Actually the LPF even reduces ghosting a bit at edges with large output value differences and put the low pass filtering on top.

I'd love to test a SNES that had that done though, just to see how it compares.

I would send you my SNES but shipping from Germany to US (and back) is not that cheap.

This one is new to me. I'm bad at google and didn't manage to find a pinout of the S-CPUN / RF5A122 anywhere, what's this doing exactly?

Pin 155 is the analoge power supply for the analog RGB and sync outputs. The resistor limits the current which consecutively reduces over- and undershoots when high brightness differences has to be driven.
I a range of 1 to 15ohm the over-/undershoots are reduced (as well as the amplitude a bit). But as the output is overdriven (compared to the standard Upp) the reduction of the amplitude is not an issue.
Moreover, with 20ohm you got Upp = 0.7V on R,G, and B, what you need for the S-RGB (and/or the THS7314/THS7374) input.
The steepness of the signals (sharpness of the outputted picture) is not influenced in that area.

[tjstogy]

Borti can you be on bobs podcast please?

[Voultar]

Generally, the 750ohm fix to the current sinks off the video DAC is considered the most "correct" fix that remains relatively easy to install and also works for the video encoder, while not changing the output termination of the amp away from the proper 75ohm.

But this is not the correct way in my point of view! The best way to fix the brightness is to lift pin 155 of the S-CPUN and reconnect it to Vcc via a 20 Ohm resistor.



(I only had a 30ohms SMD0603 resistor for this but will replace it with a 20ohm resistor soon.)

This also drastically reduces the ghosting issue present with the 1Chip consoles.
(the ghosting isn't noticeable on the screen anymore but I can still measure it with an oscilloscope.)

I've just received an NTSC SNES 1CHIP-03 and I haven't tested the brightness (using the 240p test suite), but according to RetroRGB all 1CHIPs are too bright.

Do I still need to perform the 750 Ohm mod on my SNES?

I'm planning on purchasing Borti's latest RGB Amp board which I think also fixes this, but the latest revision isn't ready so I'm interested in doing this mod first.

If you want to use V3 there a brightness corrector on the board which only effects the RGB.
The next V4, which will come very very soon, has the option to either apply the brightness correction with the three resistors or to apply an external brightness correction (hopefully with the method presented here).

Attenuating the signal (while still driving a 75 ohm load) is plenty appropriate. Typically, it's always best to attenuate on the output via the series resistor so that when you source-terminate it becomes 1v or 700mv (whatever signal level you're working with) on the device end.

[borti4938]

Don't forget: we are not at the "output side of the console" here but overall you are right
Also, the aim of my method is not to simply attenuate the signal from around Upp=770mV to Upp=700mV. It's also decreasing current flow and reducing the over-/undershoots. Ok, one could also install resistors in series for each output pin (R, G and B) but this would mean to either cut traces or to lift three (neighboured) pins! I would guess this method could be replaceable to my one. Not surprised that one would need a 20ohm resistor for each line...
Nevertheless, simply reducing the load (as be done with the 'three-resistor-method' discussed here and presented on retrorgb.com) increases the current flow.

[zeruel85]

Don't forget: we are not at the "output side of the console" here but overall you are right
Also, the aim of my method is not to simply attenuate the signal from around Upp=770mV to Upp=700mV. It's also decreasing current flow and reducing the over-/undershoots. Ok, one could also install resistors in series for each output pin (R, G and B) but this would mean to either cut traces or to lift three (neighboured) pins! I would guess this method could be replaceable to my one. Not surprised that one would need a 20ohm resistor for each line...
Nevertheless, simply reducing the load (as be done with the 'three-resistor-method' discussed here and presented on retrorgb.com) increases the current flow.

Hey borti4938, I have a SNES Mini with one of yours PCB, the v3.3b with THS7374 onboard. Do you suggest to do also the '20Ohm mod' to pin 155 or in this case is irrilevant? I don't understand in which case you recommend this mod.

Thanks!

[Voultar]

The RGB bypass is totally unnecessary for 1CHIP systems. There isn't much visual difference in the driver outputs between the 7314 and S-RGB enc. outputs. The S-RGB encoder may have some very subtle LPF going on, but in my opinion this is negligible. The THS7314 is a lot heavier on that than anything. The only real utility/benefit that I find for it is with the SNES Mini, as the installation is a bit more forgiving than utilizing the on-board encoder:

Though, I did design a board around the THS7374 specifically for both variants:



A parasite board for the SNES Mini that simplifies the "3 wire" installation, with properly terminated C-Sync:



Nonetheless, it's always good to hear other ideas for these things!

But regardless of what is done, the most important thing is to maintain signal/load integrity, and that it always "appears" to be a 75 ohm output on the TV/Monitor side of things. That's where the bulk of your shimmering and ringing derives from; Improper amplitude and/or source termination.