YPbPr ouput mod for SNES consoles

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
Locked
Rellikami
Posts: 3
Joined: Sat Mar 02, 2013 11:48 am

Re: YPbPr ouput mod for SNES consoles

Post by Rellikami »

Retrogamer wrote:So my parts arrived today and I rebuilt the amp.

Warning incoming pics...

Here's the old amp I used:

Image

It worked out great but my some of my colors were real weak, heres capture from my startech capture device that can capture 240p component using the old amp:
Tried using the 2N4401 with the schematic you posted for the S8050, actually got some results that were not black and white though not super impressive.
Tested the volts on Y Pr Pb Y was 3.46V from the chip and 1.56V at pin 7, Pr was about 3.52V at the chip but through the amp circuit it was 1.62V Pb was 3.48V from the chip but through the amp is about 1.58V everything looks much better voltage wise now.
here is the snes with the transistor amps Pr and Pr from chip into the amp Y from pin 7.

Image

Image

Image

Image

Now this is the main snes i use with no amp Y Pr PB are take directly from the BA6594AF chip.

Image

Image

Image

also my TV's settings for the SNES and my Wii for a comparison.

Image

Image

The SNES with the amp is doing slightly more lite colors mainly yellows compared to the one with no amp from what i can tell, of course they also have different motherboard revisions but the same encoder chips so some tweaking may be needed to get each one just right.
Now i need to get some of the amps Ziggy used when i get some disposable income and do a comparison.
i did not take before and after pictures of the one with the transistor amps that said it is a large improvement from having to have the color on MAX settings to look correct. i have some 2N3904's and some 2N2222A's i may test with this circuit as well for differences first i would like to test it on a Tube TV with Component inputs.
Pasky
Posts: 121
Joined: Wed Feb 27, 2013 9:23 pm

Re: YPbPr ouput mod for SNES consoles

Post by Pasky »

I recently switched over to this amp for my component mod.

I was also told by someone with electrical engineering knowledge that I trust, that the 220uF's next to the 10uF's are actually creating a capacitance of less than 10uF, so I removed the 220's and it's now just a 10uF to the amplifier. So after all was done I noticed the yellow's were very weak and the red's were way too strong. I added a 15 Ohm resistor before the Yb gets to the RCA jack and this makes the reds pretty much perfect imo. I also noticed the Luma from the Multi AVOUT was also a bit too strong for component. I added a 15Ohm resistor to it as well but still feel it's a bit too bright but I can live with it.

However I noticed something very strange. Since most games only give you a black screen at bootup you'll never notice this. The yellows are actually spot on at first, but then are drained some how. You can see this if you have a flashcart or demo that shows yellow on initial bootup, heres my SD2SNES bootup, notice the yellows on the snes logo:

http://www.youtube.com/watch?v=qXQYBDeAiqs

I have no idea what is causing it, but it looks like power being drained or something. I'm no hardware engineer but if someone with more knowledge who has done this mod can figure it out, that'd be awesome.
User avatar
Ziggy587
Posts: 174
Joined: Fri Oct 08, 2010 6:08 pm
Location: NY, USA

Re: YPbPr ouput mod for SNES consoles

Post by Ziggy587 »

Pasky wrote:I was also told by someone with electrical engineering knowledge that I trust, that the 220uF's next to the 10uF's are actually creating a capacitance of less than 10uF, so I removed the 220's and it's now just a 10uF to the amplifier.
Interesting. So you have it like this: S-ENC output > 10uF cap > 1M R to ground > amp input ? Have you measured the voltage on the amp's input pin? The whole reason for adding the 220uF cap (or that whole RC circuit) was to bring the voltage down before the amp. Also, I wonder if it also affects the voltage on the amp's output?
Pasky wrote:So after all was done I noticed the yellow's were very weak and the red's were way too strong. I added a 15 Ohm resistor before the Yb gets to the RCA jack and this makes the reds pretty much perfect imo.
Yes, I agree. I think red is too strong on mine as well. Which line did you add a resistor to though? Pb or Pr ?
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Re: YPbPr ouput mod for SNES consoles

Post by kyuusaku »

-The transistor amp above doesn't make any sense, it's just a weird buffer and serves no purpose here since it doesn't even transform the output impedance.

-Right, putting two capacitors in series is the same as putting two resistors in parallel, you don't want that. For the input of the NJM amp just use the 10 uF.

-You actually don't want to take Pb and Pr after the filters, the filters are for composite video; for the best component video you want all the frequency components that would otherwise distort composite video, but give a sharper component picture (the reason component is better than S-video).
Pasky
Posts: 121
Joined: Wed Feb 27, 2013 9:23 pm

Re: YPbPr ouput mod for SNES consoles

Post by Pasky »

Ziggy587 wrote:
Pasky wrote:I was also told by someone with electrical engineering knowledge that I trust, that the 220uF's next to the 10uF's are actually creating a capacitance of less than 10uF, so I removed the 220's and it's now just a 10uF to the amplifier.
Interesting. So you have it like this: S-ENC output > 10uF cap > 1M R to ground > amp input ? Have you measured the voltage on the amp's input pin? The whole reason for adding the 220uF cap (or that whole RC circuit) was to bring the voltage down before the amp. Also, I wonder if it also affects the voltage on the amp's output?
Pasky wrote:So after all was done I noticed the yellow's were very weak and the red's were way too strong. I added a 15 Ohm resistor before the Yb gets to the RCA jack and this makes the reds pretty much perfect imo.
Yes, I agree. I think red is too strong on mine as well. Which line did you add a resistor to though? Pb or Pr ?
I never checked the voltage before the amp. It seems to be fine.

I added a 15 ohm resistor between the amp and the rca jack on Pr. The red's and blue's are perfect now...it's just the yellows that aren't and I have no idea why. Because for a couple of seconds the yellows are perfect but then get drained after powering on (as can be seen in the video I made) and no clue why.
User avatar
NightWolve
Posts: 24
Joined: Tue Mar 26, 2013 8:15 pm
Location: Chicago, IL, USA
Contact:

Re: YPbPr ouput mod for SNES consoles

Post by NightWolve »

Rellikami wrote:
Retrogamer wrote:So my parts arrived today and I rebuilt the amp.
Here's the old amp I used:
http://i237.photobucket.com/albums/ff21 ... c7a3aa.png
Tried using the 2N4401 with the schematic you posted for the S8050, actually got some results that were not black and white though not super impressive. Tested the volts on Y Pr Pb Y was 3.46V from the chip and 1.56V at pin 7, Pr was about 3.52V at the chip but through the amp circuit it was 1.62V Pb was 3.48V from the chip but through the amp is about 1.58V everything looks much better voltage wise now. here is the snes with the transistor amps Pr and Pr from chip into the amp Y from pin 7.
Hey Rellikami, I'm a friend of Retrogamer's over at PCEFX and he showed me this thread some time back. I got started on the SNES mod because of what happened here and have been working on it ever since on and off depending on mood/interest. I made a current thread here, though I'll update my posts here about it if there's something of note. Anyhow, that's actually my first amp up there that you used with the S8050 (told Retrogamer about it some time back) and I was trying to tweak it when I discovered something interesting today: the Blue-Y output doesn't need amplification at all!

I'd like you to try this amp if you got the time or anybody else that has the S-ENC chip (don't know about the other one):

Luma/Y of course should come direct from the Multi-AV S-Video pin as it's perfect (redundant).

** S-ENC Blue-Y pin -> (+) 220 uF capacitor (as DC filter) (-) -> 39 Ohm resistor -> RCA Out to TV!

BOOM! 1 cap, 1 resistor. That's it! Very Simple! It turns out that the DC voltage was too strong such that when you'd plug it directly into the TV's 75 Ohm jack and that impedance resistor would ground it, it'd weaken the signal to the point where it'd look like it needed amplification... But if you just use a capacitor as a DC filter and plug that in-between, Blue-Y will wind up being too powerful actually!!

The Red-Y however, that's a different story; that needs maximum amplification and transistors are coming up short. You get the exact same results with the 2n3904, S8050 or the SNES 2SC1815 transistors (I tried all three).

The way I currently have the Red line is:

** S-ENC Red-Y pin -> diode -> SS8050/2N3904/etc transistor + 5V source -> emitter output goes to 80 Ohm pull down resistor AND to a 220 uF capacitor -> RCA Out to TV!

It's the same as my old photo that's linked, but instead of 1k resistor, use a diode, instead of 470 uF cap, use 220 uF, plus *no* 15 Ohm output resistance! You need the maximum that the transistor can amplify here, at least, that's my finding for my system. If you don't have a diode, you can leave it out. It's for protection and doesn't cause resistance of signal in this scenario if facing in proper outward direction, away from chip. Your final circuit, when ready to build on a PCB could include it if you want anti-static protection for the chip and in case of a transistor short sending +5V back to the chip, etc. Can't use it with the Cap trick for Blue-Y. Signal won't move through in that case.

Note: I am still tweaking both circuits, but lemme know about Blue-Y. If this is the case with other S-ENC chips, then the circuit can be much easier for everyone! To buy a chip if only Red-Y needs amplification may not be worth it if it pans out for most.

******************************************************

P.S. I recently bought a nice toy which has been a big help for tweaking this mod:

http://www.ebay.com/itm/120966019918?ss ... 1439.l2649
It's a F03147 HD USB 2.0 Video Audio Capture Grabber with YPbPr support! $23 shipped!

I rolled the dice in buying this because you don't know if 240p will be supported, but thankfully, it worked perfectly! So, with this device, you can take digital screenshots on your PC of Composite, S-Video and Component, side by side, page up/down very quickly on them to see how well the color level and tint match up, etc. The included software is kind of cheap, but it works with any DirectShow supporting software like DScaler and any other Movie Capture software, etc.

EDIT: Just made a really easy image for this all.
Image

Extra Notes: It's important to understand that the circuitry of a TV set that processes an analog YPbPr Component video signal varies and even if we could figure out the perfect output circuit, you're still gonna get people complaining about green levels, tint, etc... That's the unfortunate reality.

Examples: I have a Samsung LCD 19" TV that can process a 240p signal thankfully. It computes the color green perfectly. This is testable with Super Street Fighter II, the Ken stage where the water is supposed to be green and also with Super Mario World, the title screen. However, my old school Panasonic 32" CRT TV has a problem computing green from a Component signal and it's usually darker and/of the wrong tint, looks more bluish. It just doesn't compare to the brightness/tint of when you switch to Composite/S-Video. So on the same set, the circuitry for Composite/S-Video, properly displays green color as it should be. Two different sets+capture device, same circuit on console, yet green is wrong in one, the CRT.

The other issue is how it handles red, apparently. My Component-to-USB device tells me that I need a little more amplification for red, so it's almost perfect with no output resistance, just straight from the transistor with filter cap seems right. This also holds true for my Samsung LCD. But again, this old school CRT tells me that the red is a little hot and could use a 15-25 Ohm output resistor which is why my first circuit that Retrogamer posted has 15 Ohm resistors on the outputs... So anyhow, I am preferring to trust my USB device here in the view that a circuit revolving around that would produce the most universal result, while accepting that some TV sets you're gonna have differences and people may customize the circuit to fit that particular TV set.
Pasky wrote:I also noticed the Luma from the Multi AVOUT was also a bit too strong for component. I added a 15Ohm resistor to it as well but still feel it's a bit too bright but I can live with it.
Well, the Luma out of SNES units for S-Video is perfect. Both S-Video and YPbPr are referred to as Luma-Based Component video signals, so in principle the Luma from the Multi-AV out should be left alone for general mod guides. The reason you're seeing a difference is because of what I mentioned above, the circuitry of TV sets vary and even on the same set, S-Video and YPbPr processing vary. S-Video mode on my TV has options and filters that are off or don't apply when switched to YPbPr mode, etc. That's what explains all that, so just FYI in general. Point is, they are the same Lumas, just that YPbPr splits Chroma into Red-Y and Blue-Y rather than transmit them together as S-Video does. Anyhow, my 2 TVs (New LCD and old CRT) and my USB device with my SNES unit show that Luma is rocksolid when connected via the YPbPr jack.
Last edited by NightWolve on Fri Feb 07, 2014 5:11 am, edited 2 times in total.
My Ys Fan Translation Projects.
http://www.ysutopia.net/
User avatar
NightWolve
Posts: 24
Joined: Tue Mar 26, 2013 8:15 pm
Location: Chicago, IL, USA
Contact:

Re: YPbPr ouput mod for SNES consoles

Post by NightWolve »

Ziggy587 wrote:Interesting. So you have it like this: S-ENC output > 10uF cap > 1M R to ground > amp input ? Have you measured the voltage on the amp's input pin? The whole reason for adding the 220uF cap (or that whole RC circuit) was to bring the voltage down before the amp. Also, I wonder if it also affects the voltage on the amp's output?
I retested this with the Blue-Y pin, which as my circuit shows, doesn't need amplifying, so I edited my experiences here quite a bit. My findings: both the 10 uF and the 220 uF do a good job of DC filtering here I see. 220 uF more so; the voltage hovers at 0 to 1.2 millVolts, while when using the 10 uF, it hovers up a little higher, a few points, and in both cases, when you disconnect cable from video device, voltage builds up to 50 milliVolts... But both accomplish "killing the voltage off" so to speak well enough. By technical specs, you're not supposed to see video signals going passed like 750 milliVolts, so NO problem here.

However, while there's not much difference in DC voltage, there is visually signal-wise. The Blue-Y is stronger with the 220 uF. I see more blue to where the black bars at the top/bottom of a game are bluish, hence my need for that 50 Ohm output resistor. Switching to 10 uF, a little less, but signal is still strong and needs output resistor. Anyway, since your chip wanted low voltage, either cap is fine, but the question is what's better for the actual video signal... I'm still not good enough to help beyond visually, what my eyes tell me...

I wanna buy an oscilloscope so I can know what the hell I'm doing when it comes to tweaking a transistor amp actually. I have a good idea from the youtube tutorials I've been watching, but steve, my engineer friend says our chosen transistors don't need the tweaking like what you see in this NEC RGB amp and like these tutorials, http://www.youtube.com/watch?v=2UkuLxLu7AI and http://www.youtube.com/watch?v=61sBLBr-_OA . I actually built that NEC RGB amp, starting with the 10 uF, and no real difference from a simple amp, just weaker, and with steve telling me it's excessive, and simpler is fine, I guess I'm probably wasting my time... :\

EDIT: Hmmmm, just read the whole thread and caught up. Looks like Mike Moffitt had also discovered a simple 220 uF cap by itself worked and eliminated need for amplification. I wonder if we have the exact same chips... Well, the chip in my image is the exact one on my system, and I really do need amplification for the Red-Y signal, so that's weird that his works totally with caps... :\ I guess a general circuit/schematic for each particular chip won't be possible, you'll still get variance, possibly even with the same chips... That sucks...

Apologies, I wasn't caught up with this thread when I made the first post. I was watching from Retrogamer's post with my circuit image after he linked me and the last time I saw this thread, it was before Mike had started posting about his findings as well as Ziggy's later stuff. I'm very late the party here! :\
Last edited by NightWolve on Thu Mar 28, 2013 2:25 am, edited 12 times in total.
My Ys Fan Translation Projects.
http://www.ysutopia.net/
Pasky
Posts: 121
Joined: Wed Feb 27, 2013 9:23 pm

Re: YPbPr ouput mod for SNES consoles

Post by Pasky »

NightWolve wrote: Well, the Luma out of SNES units for S-Video is perfect. Both S-Video and YPbPr are referred to as Luma-Based Component video signals, so in principle the Luma from the Multi-AV out should be left alone for general mod guides. The reason you're seeing a difference is because of what I mentioned above, the circuitry of TV sets vary and even on the same set, S-Video and YPbPr processing vary. S-Video mode on my TV has options and filters that are off or don't apply when switched to YPbPr mode, etc. That's what explains all that, so just FYI in general. Point is, they are the same Lumas, just that YPbPr splits Chroma into Red-Y and Blue-Y rather than transmit them together as S-Video does. Anyhow, my 2 TVs (New LCD and old CRT) and my USB device with my SNES unit show that Luma is rocksolid when connected via the YPbPr jack.
I wasn't comparing TV's though. The luma is too strong even through capture devices. There was excessive white crush occuring on the picture, I even noticed it on my tvs as well. A 15 Ohm resistor did the trick however.
Last edited by Pasky on Wed Mar 27, 2013 9:04 pm, edited 1 time in total.
User avatar
NightWolve
Posts: 24
Joined: Tue Mar 26, 2013 8:15 pm
Location: Chicago, IL, USA
Contact:

Re: YPbPr ouput mod for SNES consoles

Post by NightWolve »

Pasky wrote:I wasn't comparing TV's though. The luma is too strong even through capture devices. There was excessive white crush occuring on the picture, I even noticed it on my tvs as well. A 15 Ohm resistor did the trick however.
What kind of capture device ?? Cool that it supports 240p via the Component input - that is rejected by the input module of most modern TV sets these days (I would guess capture devices too, but mine was good)! They're designed to demand 480i at the minimum, though they'll still accept 240p via S-Video (if that's even supported, most likely not with TV's) or Composite. It's a move by the industry to not bother to be backwards compatible with such signals unfortunately.

Anyway, I believe you, you're getting the result that you say that you are, my point was that at this time, I wouldn't bother to mention this variance in any mod guide myself. I wouldn't let it be representative of a % of units out there, etc. A working unit with decent caps still should have a rocksolid Luma signal. But if I see, say, 5+ or so more people mention that they're having to apply resistance to the Luma on their units, then I'd change my mind obviously. It'd be another official problem like to amplify or not to with the R-Y/B-Y signals, etc. The only patterns I've seen of issues with Luma before you have been with people that tapped it off the actual chip, not knowing that it was ready-to-go at the Multi-AV out and then trying to amplify it themselves.

P.S. Maybe edit your post a little and chop everything out cept for the last paragraph in what you quoted ?? ;) There were complaints of "wall-o-texts" earlier, and quoting them fully adds to the irritation, you know. ;)
My Ys Fan Translation Projects.
http://www.ysutopia.net/
Pasky
Posts: 121
Joined: Wed Feb 27, 2013 9:23 pm

Re: YPbPr ouput mod for SNES consoles

Post by Pasky »

Hmmm maybe its possible the luma ouput is different per snes revision.
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Re: YPbPr ouput mod for SNES consoles

Post by kyuusaku »

Pasky wrote:Hmmm maybe its possible the luma ouput is different per snes revision.
Or maybe people are creating broken "amps" due to minimal understanding of video signals, termination or transistor amplifier topologies.
User avatar
mikejmoffitt
Posts: 1353
Joined: Sun May 27, 2012 8:43 pm

Re: YPbPr ouput mod for SNES consoles

Post by mikejmoffitt »

From various tests both on televisions and using a scope to check levels, the Luma available from the AV out connector on my stock SFC, as well as the modded SNES I posted that started all this, should be completely compliant and proper without modification. My capture card is totally happy with it as well. Whatever buffer / amp circuit the SNES already has ought to be sufficient for it.
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Re: YPbPr ouput mod for SNES consoles

Post by kyuusaku »

You're missing the point; of course the encoder's output is "compliant" with a DC offset. Luma is already prepped for series source termination so it's irrelevant. Wiring component's Y in parallel with S-video's Y does mean however that you can't leave S-video and component connected at the same time because there will be dual-loads. A simple remedy with a single resistor...

What's really wrong is how Pb and Pr are handled, when you connect them directly a compliant signal is not what's delivered to your TV over the transmission line. And if you somehow have a lossless-despite-horribly-mismatched transmission line, that doesn't mean everyone else following your example will.

Also even more importantly the encoder's outputs are meant to drive passive filters and be looped back into a high impedance input. There's no indication that driving a 75 ohm load is OK.

---

As for the Pb/Pr amp above, here's why it doesn't make any sense.

1) The video chip doesn't need protecting in a common-collector amplifier. The resistor actually hinders the feedback.
2) 80 ohms is unnecessarily low so it leads to lots of current waste. 80 ohms also isn't a preferred value. If it's dropped to 75 ohms then it also serves as a parallel source termination.
3) 470 uF is a good value for composite video which has low frequency components for sync, Pb and Pr don't have that, so it's unnecessarily high.
4) 15 ohms R at the output serves no purpose but to marginally attenuate an already "compliant" 80 IRE peak-to-peak signal. A proper series 75 ohm termination would attenuate the level to 40 IRE which would really break things.
Last edited by kyuusaku on Thu Mar 28, 2013 3:32 pm, edited 1 time in total.
Pasky
Posts: 121
Joined: Wed Feb 27, 2013 9:23 pm

Re: YPbPr ouput mod for SNES consoles

Post by Pasky »

Ok, I replaced the 10uF going from the S-Enc to the encoder with a 220uF.

The result is the yellow stays great for much longer, but just drains away much more slowly. Instead of draining within 8 seconds like the 10uF it now takes about 30 seconds to drain. I'm not certain what's going on and why the yellows are being drained away. Anyone have any idea?

Or maybe people are creating broken "amps" due to minimal understanding of video signals, termination or transistor amplifier topologies.
I'm not claiming to be any kind of electrical engineer. I'm simply following a schematic and reporting my findings. You may very well be correct, I'm open to suggestions from anyone with knowledge on the subject.
User avatar
kyuusaku
Posts: 1665
Joined: Mon Sep 27, 2004 2:13 pm

Re: YPbPr ouput mod for SNES consoles

Post by kyuusaku »

Post your exact circuit and maybe it can be explained.
Locked