Trying to Fix my SNES

[RIP-Felix]

Part 1: Not leaving well enough alone.

     Okay, after the success of my thread on fixing an Atari 2600, I've decided to try my luck at the Super Nintendo. Recently I became the proud owner of my very first SNES!!! This is such a great console. I wish I could have owned one as a kid. Instead we got a Genesis, which I still think was probably the better choice. Many of the same games released on both systems and the Genesis versions were often the definitive version...well...at least that's what I tell myself. I digress...

     Well the SNES I bought was guaranteed to work or my money back...and it indeed does work just fine. "But wait, I thought this thread is about fixing a broken SNES"? Yes, yes, don't get your cords in a bunch. I'm getting to that. While yes, the SNES I bought works fine and I have been having a bunch of fun playing my small collection of real carts, what I didn't know when I bought it was not all Super Nintendo's are alike. And there's one important difference in particular that I'm referring to - Video Quality.

     I also recently bought an Open Source Scan Converter (OSSC). However, there's something amiss when I hook my SNES up the the OSSC via RGB SCART. It's underwhelming, a bit softer than I was hoping. Much more so than my Genesis via RGB. So I went on a search to find out why. The guys over at My Life in Gaming did a very nice video which explains the issue:

Spoiler

 

 

With a nagging sense of buyers remorse I opened my SNES and removed the RF shielding covering the CPU chip-sets:

Okay, that's not a big deal. I can just buy an RGB or ultraHDMI like mod for it, right? Sadly, no! There is no mod currently available for my console. I resigned myself to the fact I would just have to be okay with a softer image. But, you know me.

I couldn't leave it alone. I became obsessed. I...may have...gone a little overboard...again...and bought another SNES. Well, actually, it was two.

The first is a 1-chip missing parts, sold as is and not working. It looks to have been worked on previously, Given the desirability of 1-chip consoles, I'm probably taking a bit of a gamble here. But that's the point of this thread - to challenge myself to fix a broken SNES. The second SNES I bought is just for parts, so I can piece together any missing from the 1-chip and not have to pilfer any from my current working console. The biggest issue it looks to have right off the bat is some ugly rework of the cart slot. I'm really hoping the contact pads weren't worn away by someone who doens't know how to use a braided solder wick properly. It looks bad in the photo, but Until I get it in hand, I can't  know for sure. My hope is that nothing is fried or corrupted, and I can just rework the solder joints.

So that's where I'm at with this project thus far. My goals for this project are:

1.  Learn about the SNES and have fun diagnosing and solving problems.
2. Refine my technique using an oscilloscope, multimeter, soldering iron, and reflow station.
3. Learn more about electronics and schematics.
4. Compare board revisions, discover interesting anecdotes and asides in the SNES relm.
5. Eventually have a working 1-chip SNES main board I can drop into my current SNES. Then compare the picture quality.

So far I only stand to loose $60. But it's a descent price to pay for the fun I'll have trying to fix the hardware, and I may just end up with 3 working SNES consoles. That would be GREAT!

To be continued...

[hansolo77]

  A neat new project.  We forget how primitive technology from then is compared to today.  Me, I'd take a converter and make the RF cable go to HDMI and be done with it.  But, as you said, the quality is "meh".  I noticed this too in another fashion.  I found out last year I could install a hard drive in my "fat" PS2 and then use a "hack" to install games directly to the drive without needing the disc.  I thought that was great because my copy of ICO never played right (it was a CD-ROM and not a DVD-ROM, so it used the other laser which wasn't strong enough on the earlier model PS2's).  Anyway, I got the "hack" to work and all is good there.  The problem is the quality of the video.  The way I had to connect it up, since I no longer own any CRT tv's, was to plug it into the RCA red/white/yellow A/V ports on the front of my Home Theater amp.  It's really soft, and the audio doesn't even convert to surround even though the game says "Dolby Surround" in the opening videos.  I'm really pissed about it.  I also don't like how I can tell the PS2 to run in 16x9 mode, but all it does is make a letterboxed 4x3.  It doesn't stretch to fill my screen!  UHG!  Maybe one day I'll figure out a solution.

[RIP-Felix]

HD Retrovision PS2/3 Component Cables and RetroTink2x is the combo I'd recommend. I bought an OSSC to solve the problem you are describing, but the RetroTink is more affordable and has composite inputs, which the OSSC doesn't (although it still looks bad). Of course the best possible video output is the completely lossless RGB. You can buy cables, but need a device to accept SCART (a european video standard that never made it to the US). That is what the OSSC is good for! Watch this, it will explain in more detail:

Spoiler

 

 

[RIP-Felix]

Part 2: Developing a Strategy for Repair

So before I even recieve the broken SNES's, there is the work to be done. First I need to do my research. Below are some service videos I was able to dredge up from 1993. This kind of thing was very helpful when I was troubleshooting my Atari 2600. So I thought this would be a good place to start again. My preferred idea was to find a service manual, like I was able to find for the A2600. However, I've not found one yet. If anyone comes across one, drop a link. Videos can lead you astray, technical manuals are much preferred if they exist.

Super Nintendo Repair Videos:

Spoiler

 

So from reading and watching videos, I've started to come up with an approach. Please note this is not a guide, it's just me documenting my progress for the purpose of my own repair. It is not intended to be a comprehensive troubleshooting flowchart and I'm sure there are errors. If you follow my steps and brick your system, then that's on you.

4 Step Troubleshooting Procedure:

1) Disassemble, Clean, and Inspect:

Spoiler

• I will remove the PCB and all attached peripherals. Then I always thoroughly clean the entire board with 90% or higher Isopropyl alcohol using a tooth brush and lint free cloth, no cotton swabs or paper towels which leave fibers behind. I repeat the scrub down twice.
• I will thoroughly inspect all contact pads and surface mount components for corrosion and electrolyte leakage, also left over flux, grime and gunk. Of course, since this is a reworked board, I need to make sure all the components are there. The cartridge slot cover and its contact pins need special attention, as this is the most common source of blank screens, corrupted sprites, or inconsistent results when powering on. A trick I came across I think I'll try is using WD-40. It's a solvent that cuts through grease/grime and corrosion. Working that in with the tooth brush and cleaning off thoroughly with Isopropyl alcohol should be quite effective.
• Once the board and all the peripherals are all clean and thoroughly dry, all missing or defective components replaced, a test is in order. Most of the time, that's all that's needed to get things working again.

2) Check the Power circuit:

Spoiler

• Is the LED indicator Lit when the system is powered on? If so, I can rule out the voltage regulator, power supply, 1.5A pico Fuse, Capacitor health, and pretty much the entire power circuit. That will make trouble shooting much easier. Now there could still be some issues related to the power circuit, but getting power to the entire board is half the battle.
• Check the Power switch slider with a multimeter. Its contacts could be grimed up and not making a connection.
• Is the AC Adapter Good? From what I've learned in the A2600 repair, a 9V adapter can produce 14-16V without Power draw, that's fine. Once in circuit that should drop to 8-12V. The voltage regulator will pull that down to 5V for the main logic circuit. So the higher voltage is perfectly fine. Now there can be some graphical issues caused by insufficient Voltage and noise from unconditioned power. A good AC adapter and power supply will prevent that. Note that the SNES uses a weird 10V 850mA adapter, but any 12V 1A adapter should substitute fine. Something else to check is that the 2 prongs on the adapter have a resistance of 50-65 ohms. If it's infinity, the coils are burned out; The adapter is dead and not worth repairing. Next I need to check the barrel jack connector reads open. The resistance between the positive and negative contacts should be infinite (1). If not, then the problem probably lies in the DC rectification circuit inside the adapter. I can check that the cord isn't broken by reading continuity of the barrel jack and PCB contact Pad for both the positive and negative lines. No continuity, means the cord is broken somewhere (usually near the strain relief). Lastly I can check the capacitor and diodes on the PCB for health too. If there is any scoring or burning, it's a goner and probably unsafe to try and repair. It's easier to replace this than go too far into repairing. However, if I did have a bad adapter I'd keep the barrel jack. If I snip it off, I can mod any 12V 1A DC output adapter with that special barrel jack. I currently have a cheap and very light Chinese 9v adapter that was sold as a NES/SNES adapter and has 2 barrel jacks for each console coming out of it. It tests as 9.52 volts without load. I suspect some of the graphical noise is caused by this shallow voltage, it's probably noisy unfiltered power too. So I will likely buy a different adapter.
• The SNES uses a 7805 Voltage regulator just like the Atari 2600 does, pretty cool! So I can follow the same test procedure I used in that thread. It should maintain  5V +/- 0.1V on the output, and an input voltage of between 8-12V DC. If I get that, I can rule out the Voltage regulator.
• EDIT: On PAL SNES, there is also a 1000uF 25V filtering capacitor that needs to be checked for health. It's not suposed to populated on NTSC boards, but when I originally made this post, I didn't know that yet.. That's a very important cap and should be pretty obvious if it has leaked or blown.
• There is a 1.5A fuse that provides circuit protection from surges. A simple continuity test will confirm if it has blown or not. It's an easy and cheap replacement.

3) Reflow Solder Joints and replace blown caps:

Spoiler

• I'll Reflow any solder joints that look corroded or bubbled first. Applying a new flux and solder to the joint will renew the connection and rule it out as a potential error.
• Blown caps or those in poor health should be replaced. My multimeter should be able to tell me if they are good or not, but only if desoldered and pulled out of circuit. And if I have to do that, I might as well replace them. New caps should help clean up noise in the picture.
• The chipset could have a bad connection on any one pin and be causing all kings of issues. If I'm seeing graphical artifacts the first step is to reflow all the pins with flux and a little new solder. Then I know they are not causing the issue. If there is a grapfical error due to a bad PPU, CPU or what-have-U, then everything leading up to this step should rule out other potential causes. I'll know I should try tracking the error down to the specific chip that has gone bad. That's where the oscilloscope will come in handy. But lets hope it doesn't get that far.

4) Replace Defective Chips:

Spoiler

• Use a oscilloscope to narrow down which chip is at fault and replace with a good one. More on the specifics to be edited in later, if necessary.

EDIT:

Ever wonder why Nintendo used a propriatary power plug for the SNES? Here's an interesting anecdote I discovered while researching electronics for this project. I noticed that the 7805 Voltage regulator used in the power circuit of the SNES is the same as the A2600 repair I did awhile back. Apparently its also compatible with Atari 400/800, Atari 2600, Atari 5200, Atari 7800, Commodore 64, Emerson Arcadia, Fairchild Channel F System II, Intellivision INTV System III, Magnavox Odyssey2, Mattel Intellivision, Mattel Intellivision II, Mattel Intellivision v1, NEC Turbo Duo, NEC TurboGrafx CD Dock, NEC TurboGrafx-16, Nintendo NES, Nintendo NES Front Loader, Nintendo NES Top Loader, Nintendo SNES, Nintendo SNES SNS-CPU-GPM, Nintendo SNES SNS-CPU-RGB, Nintendo SNES SNS-CPU-1CHIP, Nintendo SNES SNN-CPU, RCA Studio II, Sega Genesis, Sega Genesis v1, Sega Genesis v2, Sega Master System, Sega Master System v1, Sega Master System v2, Tandyvision, and Vectrex. Wow, that little thing powers alot of retro consoles! Well, the Genesis power adapter is 10V 850mA and would be a one to one replacement for the SNES adapter if it were not for Nintendo's decision to us a proprietary plug. I guess that kept customers coming back to them for replacements and reduced their returns form people using the wrong adapter. I get it, it just seems type-a controlling. Sega went with a standard plug, and they did just fine with the Genesis.

EDIT:

If I look closely at one of the pictures in the e-bay listing, there is a something a bit shady. ***Edit of this Edit: It's not supposedt to be populated on NTSC boards, so this is fine. I didn't know that yet, but figured it out later.***

It's obvious to me that I will need a recap kit. So I just got a full SMD cap kit from Console5.com. I also picked up a DC barrel jack adapter for the snes, that will allow me to use my Sega Genesis wall wort to avoid e-bay scalpers charging $16-25 for an OEM Nintendo adapter. I picked up two extra 7805 voltage regulators (handy to have) and a 1.5A pico fuse while I was there. That should cover the electronics that I may need. The total was $15.52, so now I'm in $75 for this project thus far.

To be continued...

[RIP-Felix]

Part 3: 1-chip SNES Arrived Today!

Well that was fast. I went ahead and performed my test procedure:

1) Disassemble, Clean, and Inspect: 

I had a look at Console5's cap kit documentation:

A SNS-CPU-1CHIP-01 is not supposed to have a Capacitor 58 installed for north American board revisions. So the lack of it is a non issue.

2) Check the Power circuit:

The 1.5A pico fuse has continuity, so it has not blown and is good to go. The real issue I can see off the bat are that most of the capacitors have been installed incorrect. It seems someone has previously tried a recap and didn't pay close enough attention to polarity. I went ahead and de-soldered them. After confirming they were good (out of circuit) and of the correct spec, I soldered them back in with the correct polarity. However, I did make a fairly big mistake:

I lifted a solder pad off the board! After an oh $hizzzzz...moment, I realized it was the ground pad. So I just scoured the ground plane next to the pad a bit and soldered to that. I've checked with the multimeter that has continuity with the ground plane, so all is good. It was just a scary few minutes while I felt sorry for myself and came up with a plan.

Another thing to point out is that the removable portion of the 62-pin connector is missing. I hoped I could just swap the one from my working snes to test, but it turns out that some of these are 2 parts and others are one piece soldered in place. My working snes is the latter, which is harder to clean when the time comes. So when the other broken SNES arrives I'll check to see which kind it has. Hopefully the former. If not I'll need to buy one somewhere.

I decided to end testing there for now. I'd rather not plug it into power until the cart slot arrives. This way I can see the effects on the screen. Also, I kinda want to keep my Good snes intact. I'd rather build my Frankenstein monster from the other dead snes, once it arrives.

***EDIT***

I checked e-bay and the other snes wont arrive for at least 3 days. Well, I'm too impatient to wait that long with my curiosity. So I removed the power switch and controller panel from my working SNES to test the 1-chip. I can't load a cart without the pin connector, but the Power LED did come on and the 7805 voltage regulator checked out. So that completes Step 2, which is good news. I already completed most of step  3 (Reflow Solder Joints and replace blown caps) during step 2. So that only leaves the 62-pin connector.

If the replacement SNES doesn't have a detachable 62-pin connector I'll have to desolder it and replace with a one piece, which can be had new for cheap. If after that I get graphical garbage, I know it's due to the chipset (CPU, PPU or SRAM). That would be the hardest thing to fix, because I'd need narrow it down to the chip, find a working replacement, and rework fine pitch SMD components with a hot air station. That adds considerable difficulty to the repair. It is however, the next step in my education working with electronics. I haven't done it before and need the practice. I'd probably practice on some an old scrap motherboard first, but I'm hoping it won't come to that.

To be continued... 

[RIP-Felix]

Part 4: Second SNES Arrived Today!

Well the second SNES arrived today. This one was a much older unit and in serious need of Retrobrite. Good thing I'm only interested in what's inside. It's a complete unit, everything that was missing from the 1-chip that I needed was inside. Even the 2 piece 62-pin cartridge connector! It looks to be in good shape even, which was lucky. So I went ahead and Frankenstein'ed the hell out of it and placed my monster on my reanimation table (test bench). After some thunder and lightning I threw the switch and....

Nothing...

Well, crap. I double checked that everything was hooked up right and even hooked up my working SNES to confirm. The LED is on and it should be displaying and sending audio out, but I get neither. I hoped it could be that simple, but had a feeling it wouldn't be. It's no shocker, given the desirability of these 1-chip consoles. I'm not the first to try and revive this monster.

The next thing to try is reflowing all the solder pads on the chips just to make sure they have a good connection. I'll check continuity of the 62-pin connector and the solder point on the board, to be sure there are no loose connections. That can cause this blank screen.

If that fails I'm going to have to get out the oscilloscope and tackle 4) Replace Defective Chips:

• I need to find a pin that shows ROM access if the game is running, but I can't see it. That way I can either confirm is the process is getting that far but the video signal isn't getting out, or that it isn't. The answer will narrow down the possibilities.
• I'll need to check the +5V VCC going to each chip, to make sure they are getting power, and grounds to be sure current can flow through the chips.
• I need to find a good way of checking each chip in the chain, some kind of diagnostic technique that will allow me to narrow the fault down to one chip. 

So now is the time to do homework, I need to learn how to do the above...

To be continued...

[RIP-Felix]

Part 5: IT'S ALIVE!

I was watching some an old YouTube live stream with Bob from RetroRGB and Lord Voultar. They were comparing 1-chip consoles with and without the bypass board. It's a 3 hour stream, so I had it on in the background while I was poking around the board's vias with my oscilloscope (It's an I-pod nano that has been converted into a cheap 20MHz scope). Anyway, I was poking around the CPU and noticed that sometimes there was some kind of activity and other times none. I tried power on/off and hitting the reset button and noticed that a waveform would periodically come on screen. I noticed that it happened only when I hit the reset button a few seconds after turning the power switch on. The Wave form looked like a game was running, repeating non linear pulses. So I plugged into the TV, flipped on the power and...

Nothing...just like before. But unlike before I waited a few seconds and hit the reset button...after a few attempts the familiar intro to super mario world came on.

So it's working. I just need to figure out whats wrong with the initial power cycle. I wonder why it needs to wait then reset.

***EDIT***

It's like it needs time to "pre-charge" or "warm up" before the reset trick will work. It takes about ten seconds after PWR ON before the reset button will begin working. On my OSSC LCD screen I notice that after a few initial resets it will show the 15KHz sync then fail and say no sync, after that it only takes one or two more attempts to sync and be stable. Once it has been on and running for a while, if I turn it off then back on withing 4 seconds, it will fire up no problem, but not after 5 or more seconds. After that I have to hit the reset button a couple of times to get it back.

***EDIT***

It's an odd one, but I do now have a working, if finicky, 1-chip SNES. I plugged it into the OSSC via RGB scart and it looks much better than my SNS-CPU-RGB-02. That one had a sort of diagonal checker board pattern of jail-bars, Looked like strands of rope or yarn in large patches of solid color (see below). The 1-chip clears that up and looks sharp:

Wow, no wonder they are so sought after. Almost emulator quality picture. The pictures were taken with my phone, since I don't have a capture card. Consequently,the quality looks bad zoomed in like this. but that's the best I can do atm. It's much better in person. Both are through RGBs SCART to my VIZIO P55-F1 HDMI5 via the OSSC in Line triple mode (720p). That is the highest line multiple the TV will accept.

Now I just need to figure out that power on issue. Why does it fail to start immediately? Why do I have to hit reset a few times to get it to work? That's that last hurtle.

***EDIT***

I replaced the 7805 voltage regulator with a 78S05 and placed a 470uF cap across the output/ground, to prevent the vertical bar in the center of the screen sometimes seen. I also went ahead and recapped with the Console5 SMD kit, being sure to get everything placed right and checking the capacitor health. Everything checked out, including the previous caps from the last recap. That did not resolve the issue, and the system exhibits the same behavior. I think that definitively rules out the main power circuit.

I read that 9v is sufficient to work with the 7805, but not quite high enough to prevent a rolling bar of static fuzz that crawls down the screen every 10 seconds. I was using a cheap 9v aftermarket adapter purported to work with both nes & snes. I used my CRT before getting the OSSC, which is why I didn't notice it before. Now that I have it hooked up to the OSSC on my big screen, it's there just like they said. So I hacked up a 12v 2A AC/DC adapter I had laying around. I reversed the polarity so that the center pin is negative, to match the SNES requirement (this is opposite to most adapters). After checking it with the multimeter and feeling for any heat in the wire, which would indicate a short, I tried it out. Works fine to remove the static bar. No change to the reset/sync issue.

I was hoping it might fix my power/reset issue, which I'm starting to think is a problem with SYNC. I could bypass the RGB sync altogether with Voultar's 1CHIP/Mini THS7374 RGB Bypass Kit. It produces a clean, properly attenuated sync. If that's the problem, this might fix it. However, I'd like to find some diagnostic procedure to confirm before buying his kit.

To be concluded...

[RIP-Felix]

And now for a short Anecdote:

I had a bit of a start when I was scoping the SRAM next to to the 1-chip CPU and getting some high frequency patterns. I had my CRT on and set to channel 3, because I was checking to see if the RF output was getting picture. Anyway, when I poked a line there was a sizzling sound that sounded like I was frying the board! It took me a moment to realize the sound was coming from the CRTs speakers. I had to remove the RF shielding over the chipset to access them. Also, I used alligator clips to ground my probe. I realize now, that acts like an antenna. I was broadcasting static! What I was hearing was that being picked up by the CRT.

Gets the blood pumping!

[RIP-Felix]

PART 6: Rejected by the Gods

Here's where I'm stuck:

On powerup it black screens. After a few seconds I can hit reset a few times and It will startup. After the initial power fail, then reset trick, every game I've tried runs perfect. No graphical glitches that would indicate a bad PPU, It just takes about ten seconds after PWR ON before the reset button will begin working. On my OSSC LCD screen I notice that after a few resets it will show the 15KHz sync. Once it has been on and running for a while, if I turn it off then back on withing 4 seconds, it will fire up no problem, but not after 5 or more seconds. After that I have to hit the reset button a couple of times to get it back.

I have reached the end of my ability's. I started a topic on the Shmups forums to see if they can help me find a way forward. I'll update with news if we find a solution...

[RIP-Felix]

PART 7: Cheat

Well, it's fixed!

Of course the solution I came up with was to buy a working 1-chip motherboard of e-bay (just the motherboard, to save a little bit). It still cost me $75. It came in and works perfectly. A quick swap into my case and pesto! It's fixed!

Okay, okay...I know! That's cheating, but I'm not really done here. I have 2 reasons for doing this:

1. I'll have a working SNES to play with while I try to fix the other motherboard. That allows me to take risks with I wouldn't otherwise. If i do end up fixing it, I can always sell it.
2. If the working one goes bad in the future I have the semi-working board as a backup to use for parts, even if I never figure out what it's problem is.

Anyway, this is where I'm at. The shmups forum was helpful, but no way forward has materialized yet. I have an idea or two to try, but I'm waiting on a Hot air rework station to arrive before I can replace a chip on the board involved in the reset circuit.

[tthurman]

Success is success, congrats!

 

[hansolo77]

I wonder if maybe you have a cold solder joint somewhere.  I had a similar issue with something I had done and it fixed by re-soldering it.  Kind of like the connection isn't secure until it's heated up a little.  Regardless, glad you got a working system!

[RIP-Felix]

Yeah, I thought that too at first. I went over the entire board thoroughly. I reflowed the 62-pin cart connector, all the fine pitch chips, and the clock crystals. I haven't tried reflowing or replacing the bazillion tiny SMD resistors/capacitors/diodes and etc on the bottom side of the board. That's a bit much! However, I did replace the electrolytic capacitors as well as the voltage regulator on the topside.

At this point I'm confident the issue is a fault on one of the chips, rather than a bad connection. There's a 3 leg device with designation U11 involved in the reset/power circuit that I'm going to swap out. I have a few other working boards with that part. They are non 1-chip boards, so sacrificing one of them to restore a 1-chip makes sense. It's just I've been waiting on a hot air rework station to arrive that will allow me to lift chips without damage. It arrived today, so maybe this weekend I'll give it a try. I think I'll try it out on an old PC motherboard first, just for practice before the real thing.

If nothing else this project has greatly improved my soldering skill. I feel comfortable taking on projects that I used to call it quits on. Those fine pitch legs on a CPU don't scare me anymore!

[hansolo77]

More power to you man.  I still question my skills with an iron.  From installing modchips to simple wiring for RetroPie builds, I'm still really iffy.  I don't even feel confident with stuff like reattaching the negative battery spring on a double D-Size battery flashlight!

[RIP-Felix]

22 hours ago, hansolo77 said:

More power to you man.  I still question my skills with an iron.  From installing modchips to simple wiring for RetroPie builds, I'm still really iffy.  I don't even feel confident with stuff like reattaching the negative battery spring on a double D-Size battery flashlight!

The secret is FLUX and a temperature controlled soldering station, not those 15-30W irons you get $10. They get too hot and damage things. Without flux your solder oxidizes into a brittle mass that wont melt. 340-Degrees C, a good tip and some flux is what you need to get off on the right foot. From there it's all technique. That's what YouTube is for, among other things.

As a quick aside, I've seen many traces on my boards that were damaged by people using those cheap soldering irons. The copper amalgamates with the solder if you go much above 350C. Or they scrub with soldering braid too hot, which wipes/scratchs the trace off. Once it's dammaged, it's hard to get a good connection. The heat can melt the solder resist that keeps the solder on the trace, not on the ground plane around the trace. Once it's gone, It's easy to make a short. Overall, a $100 investment will get you a very decent (if cheap) rework station with a temperature selectable ceramic tip soldering iron, hot air gun, and even a DC power supply. This would make soldering MUCH easier.

This is a must have if you're going to attempt to install a Dreamcast DCHDMI or N64 UltraHDMI mod. Anything more complicated than simply joining 2 wires will greatly benefit.

[hansolo77]

I enjoy the tinkering, but can't support spending a lot of money on a big station when all I do is tinker.  I'll admit I just use a cheap on/off iron, but I did invest in a good sharp tip and a much finer diameter solder.  I had been using the stuff usually included as a freebie with the iron, which is thick.  Took me years to realize the problem was the solder.  I've been getting a lot better results with the finer solder.  I think i might get a better iron one of these days.  I also got some flux but I don't quite get the benefit of it yet.  The stuff I initially bought was in a white plastic "jar", and was a goopy brownish yellow gel.  After one use (using a toothpick to apply), the container became sealed from the flux and was unable to be opened again.  I then bought some stuff in a syringe but it clogged and became unusable too.  I also have a problem where my tip oxidizes too fast and requires a lot more solder and frequent cleaning.  I still consider myself a 1st grade novice, and no where near a high school senior yet.

[ClassicGMR]

3 hours ago, RIP-Felix said:

The secret is FLUX and a temperature controlled soldering station, not those 15-30W irons you get $10. They get too hot and damage things. Without flux your solder oxidizes into a brittle mass that wont melt. 340-Degrees C, a good tip and some flux is what you need to get off on the right foot. From there it's all technique. That's what YouTube is for, among other things.

I learned that a long time ago working for Zebra. I was afraid to take the job at first because I thought my soldering skills sucked. Turned out all anyone really needs is the RIGHT solder tools. Flux, de-solder ribbon, temperature controlled solder irons - it all makes a huge difference.

 

EDIT: Also WELL VENTILATED WORK STATIONS! Whew is that necessary!

[RIP-Felix]

On 6/15/2019 at 3:33 PM, ClassicGMR said:

...Also WELL VENTILATED WORK STATIONS! Whew is that necessary!

Good point!

Solder fumes are toxic, and should not be breathed in. Problem is, unless you like to solder in 100 degree F heat, solder fumes will be drawn to your face by convection. It's caused by your body heat warming the air in contact with your skin, which rises just like heat from a fire. We're walking thermals. That's why on a still night the smoke from a campfire seems to follow you - because it is. Especially on cold nights.

Spoiler

 

On 6/15/2019 at 1:33 PM, hansolo77 said:

I enjoy the tinkering, but can't support spending a lot of money on a big station when all I do is tinker.  I'll admit I just use a cheap on/off iron, but I did invest in a good sharp tip and a much finer diameter solder.  I had been using the stuff usually included as a freebie with the iron, which is thick.  Took me years to realize the problem was the solder.  I've been getting a lot better results with the finer solder.  I think i might get a better iron one of these days.  I also got some flux but I don't quite get the benefit of it yet.  The stuff I initially bought was in a white plastic "jar", and was a goopy brownish yellow gel.  After one use (using a toothpick to apply), the container became sealed from the flux and was unable to be opened again.  I then bought some stuff in a syringe but it clogged and became unusable too.  I also have a problem where my tip oxidizes too fast and requires a lot more solder and frequent cleaning.  I still consider myself a 1st grade novice, and no where near a high school senior yet.

Yeah, all that's what I'm talking about. I too was using one of those for the longest time. However, when I was preparing to install the UltraHDMI mod in my N64, I decided to get a proper station. It was only a $50 Chinese cheapo, not a name brand. But it's really made all the difference. The one I just bought was $75 and has hot air, and a DC power supply as well. So that's got me covered for the modwork I've gotten into.

Flux improves attraction and cohesion of the solder to the copper trace. You can reheat cold solder joints with fresh flux to renew them. That's what I did to my SNES. It saves me the trouble of removing the old solder with a wick and replacing with fresh solder. It makes it much easier to get the solder to stick and creates a stronger bond.

 

[RIP-Felix]

PART 8: Digging in...stubbornly

I just got a new rework station with an  iron, hot air, and a DC power supply. I went ahead and swapped U11 (a 3-prong SMD IC  rama suggested to me on SHMUPS) with one from a working SNS-CPU-RGB-02. Went smooth as butter with hot air, way easier than I thought. SMD doesn't scare me anymore! Anyway, the same behavior was persistent on the 1-chip. Black screen on power up, reset button causes the game to start as normal. Also, the donor board still works fine. So U11 definitely wasn't defective...dang it!

A word of caution though. If your working on a SNES always keep at least 1 screw in the power switch, to keep it secure. Or use some electrical tape on the bottom side to cover the pins. Like an idiot, I let it dangle and then I knocked it off the standoffs while testing. It must have shorted on the RF shielding. When I went to turn it on...nothing (no PWR LED). After a bit of self pity (I took a break), I confirmed I'd blown the pico fuse then replaced it. Easy fix, but it gets the mind racing. "Will I have to Recap again? Did I just send 12V strait to my CPU and blow it up? Did I just F#$K up my 1-chip." It's best to take a break and come back with a multimeter after the panic subsides and reason returns. If nothing else, it get the heart pumping!

From the available schematics, datasheets, and from Fullsnes documentation I've pieced together the following:

• Note: I'm using "x" to indicate components I've already ruled out by either replacing with working units or I have confirmed operation with a multimeter, reflowing, swaping out, and etc. "-" means I don't know or haven't ruled out yet. Also, I'm not listing every capacitor or resistor. Just the major components. I did perform a recap, so that's already ruled out.

Everything in the Power Circuit:
x 10V AC/DC Adapter (Center pin -)
x Barrel Jack Conn
x 1.5A Pico Fuse
x PWR Switch
x 7805 Voltage Regulator

Everything in the Reset Circuit:
x Reset Switch
- S-APU ("combines S-SMP, S-DSP, 64Kx8 Sound RAM, and possibly the NEC UPD6376 D/A Converter")
x X2 (24.576MHz ceramic resonator that feeds into the S-APU. I replaced with one from a working donor board.)
- S-CPUN A (combines CPU, PPU1, PPU2, S-CLK)
- S-WRAM B
x T529D (U11), "for resetting the system after detecting the voltage at the time of powering on".
x Cartridge Slot pin 26
-Expansion Slot pin 19

- EDIT:  I forgot about the SuperCIC chip (F411B). This is Nintendo's security chip that checks for another CIC in the cartridge to see if they match. If not it refuses to boot. This ultimately ended up being the issue (See part 9 below). I had "assumed" that if it were bad, it would refuse to allow the game to start regardless of the reset button being pressed. So I prematurely ruled it out. Turns out that a cold solder joint can cause this behavior. I must have missed that when I reflowed the chips on the board. Too bad, it really threw me for a loop.

I may be missing stuff, but that's what I was able to piece together.

I've been wondering about the initial PWR cycle. What I'd like to know is how it works, but there is little to no information out there. Just from observation, when you start any SNES you'll notice a flicker (I assume it's the Reset pulse), followed by a short pause before the game starts up. Mine doesn't flicker at all when I turn on the power switch. Just still black, no sync on my OSSC is displayed. The Sync doesn't happen at all until I manually press reset. Then it flickers and the game comes up like normal. So it's like there's no reset pulse on power on. So rama's suggestion on the SHMUPS forum sounded dead on point.

That T529D (U11) is very suspect. Like rama said, it's job is to generate the reset signal after it "detects" power on (4.2v). It's DataSheet says that it's used as a "measure against erroneous issues at power on/off". I mean, bingo! That sounds like the right thing to check, but I've already confirmed replacing it doesn't fix the issue. So that suggests the CPU isn't ready to receive the reset pulse. In this thread rama points out that there is another reset signal generated by the S-APU that the CPU is "waiting for". I've read a bad APU can cause the system to hang (black screen), but I've not heard of the reset button fixing it. I did swap X2, 24.576MHz ceramic resonator that feeds into the S-APU, to rule it out. I also ran the Burn in Test Cartridge and it passed on everything, even though it sent my OSSC for a ride (kept dropping sync).

Perhaps my APU is going bad, but it's not dead yet? But then, why would the test cart give it a pass? Same if the CPU or S-WRAM had issues? If my only way forward would be to try swapping the APU, then the SWRAM, and finally the CPU, perhaps it would be better to keep this board as a donor for when my working 1-chip goes bad. If I were to buy yet another for parts board off e-bay, it'll probably be an easier fix than this one. I would like to try my hand at replacing SM IC's, for the practice (that was the point of this purchase), but it's actually easier to get working a 1-chip off e-bay than it is to find donor boards (let alone specific chips). Go figure...

I'm still not ready to surrender, but I may not have a choice. I'm willing to use hot air to remove the S-APU, S-WRAM, and CPU if I have to, but not if it means I have to destroy and easier to fix board in the process. I'm all for being persistant, but destroying a better unit is...wrong.

[hansolo77]

You lost me, but it's fun reading your progress.  At this point you could probably build your own console..

 

[RIP-Felix]

7 hours ago, hansolo77 said:

You lost me, but it's fun reading your progress.  At this point you could probably build your own console..

Lol...Fat chance! No no, I'm just a novice playing around at this stuff for fun. I do have mad respect for electrical engineers now, however. Especially reverse engineers! Holy $h!z it's much harder when you don't have schematics and service manuals to go off of. Nintendont vs Atari? Hands down the Atari 2600 is easier to troubleshoot thanks to service manuals and available schematics. The SNES? Nothing! If you had an issue they had you call a number and they did everything in house. Then they must have killed all the workers, because no one ever talked! Seriously, all the schematics on this I've found were done by enthusiasts. Nintendo keeps their secrets secret. The closest thing I've found is a Burn in Test Cart rom. Useful sure, but not a servicing flowchart complete with test diagnostics and helpful insights.

[RIP-Felix]

PART 9: Never give up, never surrender...It's VE Day B!t@hes!

I am happy to report victory has been achieved!

Thanks to a post on the SHMUPS forum I swapped the SuperCICs from my broken 1-chip and a working SNS-CPU-RGB-02. I mentioned that one of my goals was to use hot air to work off a chip with many fine pitch pins. This was my first try and it went well, thanks to a proper rework station, good flux, and solid pair of course. Well, the result was a proper boot up! No reset button needed. the game started up first try! To confirm the CIC was bad, I then tried the donor board. Weird thing was it too booted up fine. So the only explanation is a cold solder joint. I thought I reflowed all the chips on the board, but I must have missed that one. The CIC was fine.

So that concludes it. I learned a lot about the SNES and had a bunch of fun trying to figure this one out. Ultimately it proved beyond me and I had to consult those with more experience than me. They came through and I'm the better for it. I pushed my soldering and reworking ability. It was close, but I'm glad I didn't give up. I feel nice bit of pride for seeing this one to the end. Moreover, the world has one fewer 1-chip SNES's going to the landfill. Actually, I bought 2 broken SNES's and both work now. That's awesome to me. That an otherwise novice can resurrect a busted console and keep it going on into the future. Yes! To put it simply...

[hansolo77]

Called it.  Congrats man.

[Hayesb]

Nice, was reading thru your fix. Sadly I have A LOT of SNES system with black screens and it's super hard to diagnose snes chips. SO I'll be working on those since I've fix what was easy now comes the time and studying!

[RIP-Felix]

Glad I could help contribute to the available information online, sadly there isn't enough of it out there for would be SNES repairman. Nintendo sure doesn't help in that department. I wish you the best of luck on you adventure!. May the landfill never claim another console!

The CIC was one of my first suspects and I reflowed the chip just with a hot iron. That was ineffective. Took me a long time to circle back to the CIC after I thought it was clear. Don't be like me and clear a component until it's been properly checked and reflowed.