Super retro trio tear down - Taking apart the SR3.

[MaxWar]

The RetroGen Adapter is VERY quiet. There's very little VDP noise in the audio output, both on the stock mixer and the Mega Mixer, so little, in fact, you can hardly hear it. I'm not too sure what causes the exaggerated VDP noise on the Super Retro Trio and even the RetroN3, but I'm starting to think it may be attributed to some problem with the 5V rail to the cartridge slot. The RetroN3 and RetroGen Adapter attempt to limit the current to the cartridge slot by putting a 10ohm series resistor, which not only causes major VDP noise, but also causes a heavy voltage drop with both Virtua Racing (after faking !LO_MEM by bridging that pin with address line A23) and the Mega EverDrive. The Super Retro Trio, on the other hand, simply wires 5V directly to the cartridge slot, which might also contribute to heavy VDP noise. My RetroGen Adapter has a 1N4005 diode between the 5V rail and the cartridge slot, which drops the voltage to 4.1V for cartridges. It reduces VDP noise and is at a high enough voltage to prevent Virtua Racing or the Mega EverDrive from causing problems.

Now, I've run out of 1N4005s, so I can't see if that fixes anything on the Super Retro Trio's Genesis side. If you have some on hand, I can take a look at the motherboard for you and point out where to solder the diode.

Ace, I reply here to not drive the mega amp thread off topic.

I tried following that lead and removed the 5v pins from my SR3 cart slot. Yeah, I did just that because it was easier and more easily reversed than cutting traces.


From there I used a Grind Stormer bootleg genesis cart and attached a Wire to its 5V. This allowed me to do some tests by putting stuff in series with the 5v goind to cart.


I tried with Diodes in series ,I had a 1N4004, but the console refused to work with that. I would get 3.6V on the cart.

I then tried with Inductor coils. Still did not work, but when I put a big enough one it worked!
And lo and behold, it actually reduced the VDP noise level. It is still there but it sure removed some edge off it. I would say it went from seriously annoying to tolerable.

I wanted to do more testing with different diodes but at that point my SR3 completely stopped working. I eventually discovered that the connections on the Gray ribbon cables were broken.

I guess that was an inevitable consequence of manhandling the console pcb for too long, damn cheap design can`t take a beating. Until I replace all those wires with proper replacement and a Drakon load of hot glue to keep them still I cannot do more experiments.

But so far so good. Progress has been made.

[AlmostOriginal]

Until I replace all those wires with proper replacement and a Drakon load of hot glue to keep them still I cannot do more experiments.

[sean]

I removed r83 and now I don't get any video and the solder pad under the 3 is burned off.
So, were does that point go.
thanks,
Sean

[MaxWar]

I removed r83 and now I don't get any video and the solder pad under the 3 is burned off.
So, were does that point go.
thanks,
Sean

Uh?
You removed R83 on the genesis part? Why did you do that?
Also, what do you mean by Under the 3. The ''Northern pad'' of R83?

[sean]

tumblr_n5hva7oJix1rc7qlzo1_1280.jpg
I removed R83 on the NES Side.
It's Suppose to make the video better on the NES and now no video.

tumblr_n5hvvzhHKM1rc7qlzo1_1280.jpg
I also was going to do the compatibility fix for MMC4/5

[MaxWar]

tumblr_n5hva7oJix1rc7qlzo1_1280.jpg
I removed R83 on the NES Side.
It's Suppose to make the video better on the NES and now no video.

The left side of R83 that you removed connects directly to the composite output. The right side connects to two vias that go under the board.
Where did you get the instruction to remove this resistor. Link please?

Also, a trick to remove surface mount resistors without damage is use Flux and wick. Just stamp the wick flat on the component and melt the two sides at the same time, then it sticks to the wick. It is super easy.

Edit: I looked some more into the board and looks like r83 connects only to video output on the left. So no wonder you have no video.
You probably needed to bridge it, not remove it. In which case removing it is utterly unecessary, you just need to put engough solder on top of it to bridge it.

Now that you have broken the right pad, you should scrape the coating around the pad. I think it connects to two traces. You should connect those spots to the other pad with thin wire. Alternatively you can inspect the board and find where the traces connect and route longer wires.

[sean]

The left side of R83 that you removed connects directly to the composite output. The right side connects to two vias that go under the board.
Where did you get the instruction to remove this resistor. Link please?

Also, a trick to remove surface mount resistors without damage is use Flux and wick. Just stamp the wick flat on the component and melt the two sides at the same time, then it sticks to the wick. It is super easy.

Edit: I looked some more into the board and looks like r83 connects only to video output on the left. So now wonder you have no video.
You probably needed to bridge it, not remove it. In which case removing it is utterly unecessary, you just need to put engough solder on top of it to bridge it.

Now that you have broken the right pad, you should scrap the coating around the pad. I think it connects to two traces. You should connect those spots to the other pad with thin wire. Alternatively you can inspect the board and find where the traces connect and route longer wires.

here is the link its from a youtube video from Satoshi Matrix https://www.youtube.com/watch?v=8qMbtjUBeO8

[MaxWar]

here is the link its from a youtube video from Satoshi Matrix https://www.youtube.com/watch?v=8qMbtjUBeO8

When does he say to remove the resistor?

[AlmostOriginal]

6:40 - Remove it and bridge the connection with solder.

[MaxWar]

Then no need to remove r83, just bridge over the resistor. It has the advantage of being both easier to install and reversible.

[MaxWar]

w00t!!! The Genesis VDP noise Is a Fixed!! Thanks and props to Ace for the Idea of the Diode, it totally works.

I managed to fix my dead SR3 with some spare wire and could get back to testing.
I then scavenged a 1N4005 and putting it in series with the Cart 5v did the trick.
Not sure why it did not work with the other diode the other day. Either the diode was bad, not suitable or my alligator clip was set wrong *facepalm*

This is much better than using a coil. Coil removed maybe 45% of the noise but with the diode it is more like 95%.

In the meantime I also lost a SMD resistor for the chroma output. *facepalm*
Next installment I will try to figure out a proper way to insert the Diode into the PCB.

In the meantime here are some pics of my Ghetto Trio.

[Ace]

Wow, what a setup. XD

Nice work.

Also, Sean, I have Satoshi's Super Retro Trio with me and I must say that bridging the resistor isn't really the best thing to do. On my TVs, with the resistor removed, the video from the NES side was overly bright. If anything, putting back a 75ohm resistor where R83 is made the video look correct on my TVs. Perhaps some middle-ground needs to be found between no resistance and 75 ohms of resistance.

[bustedstr8]

When trying to find the right level for an analog signal like that, your best bet is a small trimmer. Give that pcb a small shot of Drakon love and remote mount a 100 ohm trimmer in a good spot.

[MaxWar]

I was wondering what this whole "Nes picture too dark issue" was all about as i did not feel anything was wrong on my side here.

So I just measured R83 on my SR3 on it is indeed 75 Ohm already and the picture looks just fine. A little research found me a post by Satoshi where he claims it is 750 on his console.

Sounds to me like someone at the Retro Trio factory made a typo when ordering parts -_-
Might be the same person who ordered 3.3v ram for the SNES board instead of 5v ram -_-

Following that logic, how many more random flaws might we find on these consoles?

[Bibin]

VOH on a CMOS chip like everything today is 99% VCC or pretty damn close on an unloaded condition.

In case of an 5V ASIC IO connected to a 3.3V powered RAM what will happen is that the ASIC VOH will drop to 4V-ish, because there's an ESD diode inside the ram that will drain the IO line to the RAM's lesser VCC (3.3V + 0.6/0.7V diode forward voltage drop). This diode will be under stress when the IO is high, and when it blows, the ram IO will survive, but not much longer as there's nothing to drain the excess IO voltage and that will breakdown the oxides on the input. A series resistor in here protects the diode by limiting it's current quite a bit, letting the ESD diode survive (and the IO thereafter)

This is what I was alluding to earlier, thank you for the clear explanation. This is why I am concerned about so many 3.3V based bootleg cartridges and systems and all of that which will eventually damage the irreplaceable ASICs on these systems. It will be big problem for the Neo-Geo MVS boards when everyone is using the 161-in-1 bootlegs that are so common.