NES Controllers

[Moriambar]

Hi,
recently a friend has found a used NES which works now, apart from the controllers which seem not to work (eg: only some keys work, and they depend on which controller is used and not which port they're inserted).
I currently haven't the NES with me but I have the controllers. I opened them (well, the first one) and all of the cables show continuity from the board to the connector. Now, is there a possibility that it's just some bad contacts between the carbon pads and the pcb? Can I clean them, eg with IPA? Or should I check the mn4021b chip onboard?

Thanks for your help,
cheers
edit: if it helps I can attach photos in the morning

[Radioman23]

The first thing I would try is to carefully disassemble the controller and give the pads& circuit board a cleaning with IPA. There is very little harm in trying this first. The rubber pads could be worn out( they don't spring back up after being pressed) but you'd have to order a new set of button pads from either "Repair box" or the other common brand that is seen on ebay. If that fails, and you have tried a new set of button pads, Caig makes a few products that might restore the contacts. I love their liquid red Deoxit cleaner for general contact cleaning, after IPA& such. Caig also makes a product that may help the contacts themselves: CaiKote 44.  Failing that,I assume you checked the cables for breaks so... I'd put the chip& circuit board under a microscope or a jewelers loupe to look for clues. If you have partial functionality, then I'm not 100% sure the mn4021bchip is the issue.

A quick look says the mn4021b might be available , with mistubishi & Mitsumi stamps on the two sellers I found. From experience, I know Mitsumi made a lot of parts for Nintendo over the years, so that'd be my first choice. Its 3 AM locally, so I might be a bit groggy.
CORRECTION: Its Mitsumi, not Mitutoyo. Mitutoyo makes calipers& machine tools, not Nintendo parts.

[T3sl4co1l]

And similarly, some contact cleaner on the plug wouldn't hurt; note this is unlikely to be a direct problem, as it works at all.

Same for the cartridge connector of course.  Which has the added 'feature' of the contacts tending to get bent out of the way, at least for the front loader (the Famicom I think is better behaved?).  It's possible to pull the springs gently back into place with tweezers.

Tim

[Radioman23]

Unlike the N64, there seem to be plenty of new made connectors for the NES/ Famicom series.

Heck, even new made cartridge PCB's. Not that I have vetted these new makes personally as I type this, but they might be usable. I know that trying to salvage the connectors from the " Kingworld" made N64 connectors is a waste of time, as the original  connector bodies use a slightly different gold plated spring contact form. Again, not an issue for OP, BUT worth noting.

[Moriambar]

thanks for all fo your replies. I'll try to clean the contacts, and I am attaching here some pictures of the connector and board.
Unfortunately I'm unfamiliar with the brands and products you're mentioning (except ipa), being in Italy there's no deoxit here or similar (I have to kinda order them from germany and they're not cheap, even before shipping!) I'm quite sure there could be other brand/names available here, just don't know what to search for.

Apart from that, the whole NES, which is still at my friend's, works perfectly, cartridge loading etc. And the cable for the pads, as I mentioned, show continuity to the port.

Apart from cleaning is there any way to test this (eg, with a psu and a scope) without the NES attached?
Thanks

[T3sl4co1l]

Sure -- with an ohmmeter, probe across a pair of pads.  It should read high or open (depending on which polarity you're using, there may be a small current through the logic chip, or the pullup resistors -- the straight carbon tracks).  Pull the respective rubber piece from the top shell and try to position it over the board, and test the action.  Whatever resistance it goes down to, is what it's doing.

You probably need about four hands to probe this, so take the time to do up connections carefully, or make a setup or something.  Like, micro-clips could probe at the IC pins, hands-free; if you don't have any, you could solder jumpers onto respective pins, just for testing.

The resistance should be, Idunno, below a kohm or so?  I forget what button pads usually do.  In any case, it must be small in relation to the respective pullup resistance.

Board looks pristine, so I'd guess it's just button pad wear.

Tim

[Radioman23]

Looking at those pads, I'd replace them with a new set. I just don't see anything too bad on the PCB.The traces look good, the solder joints, from the image, look shiny and good. I assume the chip has no obvious visible signs of damage?. Failing all that (though new button pads can only help)I'd get he data sheet for the chip (I don't have it on hand as I type this) and look there for ideas. In my experience, which is subject to change, things like this are best tested with the NES there. It is certainly possible to test without it, but I'd prefer to have the console there. If you haven't already, give the contacts in the plug a good cleaning. I thought the Euro-zone would make inter-European shipping simple, but I am woefully ignorant on that at the moment.

[Moriambar]

Looking at those pads, I'd replace them with a new set. I just don't see anything too bad on the PCB.The traces look good, the solder joints, from the image, look shiny and good. I assume the chip has no obvious visible signs of damage?. Failing all that (though new button pads can only help)I'd get he data sheet for the chip (I don't have it on hand as I type this) and look there for ideas. In my experience, which is subject to change, things like this are best tested with the NES there. It is certainly possible to test without it, but I'd prefer to have the console there. If you haven't already, give the contacts in the plug a good cleaning.

thanks I'll try and do what's possible. Btw no damage anywhere to be seen and

I thought the Euro-zone would make inter-European shipping simple, but I am woefully ignorant on that at the moment.

easy? kinda. Inexpensive? No. I mean: no customs but it's usually way more expensive than finding something equivalent sold in Italy and getting it from amazon or other shop

[Moriambar]

Sure -- with an ohmmeter, probe across a pair of pads.  It should read high or open (depending on which polarity you're using, there may be a small current through the logic chip, or the pullup resistors -- the straight carbon tracks).  Pull the respective rubber piece from the top shell and try to position it over the board, and test the action.  Whatever resistance it goes down to, is what it's doing.

On the board? well they're open. The ones under the buttons are actually quite dirty and worn though

You probably need about four hands to probe this, so take the time to do up connections carefully, or make a setup or something.  Like, micro-clips could probe at the IC pins, hands-free; if you don't have any, you could solder jumpers onto respective pins, just for testing.

The resistance should be, Idunno, below a kohm or so?  I forget what button pads usually do.  In any case, it must be small in relation to the respective pullup resistance.

Board looks pristine, so I'd guess it's just button pad wear.

Tim

thanks, I'll try this next!

[Moriambar]

ok follow up.
Board and pads cleaned with IPA, now when I press the A button the resistance on the corresponding pins on the chip drops from Megaohms to ~500 ohms. But the B button drops it to ~5k and the SELECT button to around 40k (each measured on their corresponding pins.

Are these "button problems" or should I investigate more?

Cheers

[T3sl4co1l]

What are the pullups, like 50k?  5k is pretty marginal, less is fine.  Also more to the point, whether it's consistent or not.

Popping mine open, I measure around 100 ohms for a good contact (D pad), varies somewhat with pressure.  Mine are probably not nearly as well used as yours...

FYI, chip is a CD4021 or similar -- just a basic shift register.  So it's a classic serializer, and basically used as an SPI IO expander.  And despite the 2-3m cable, no supply bypass cap -- to be fair, at 5V it's slow enough it likely doesn't need one. Not sure offhand what drives it, from the main board; that would be a major source of noise (EMI emissions) -- I mean, it passed, back in the day at least, so it can't be too bad.

Tim

[Moriambar]

What are the pullups, like 50k?  5k is pretty marginal, less is fine.  Also more to the point, whether it's consistent or not.

Popping mine open, I measure around 100 ohms for a good contact (D pad), varies somewhat with pressure.  Mine are probably not nearly as well used as yours...

FYI, chip is a CD4021 or similar -- just a basic shift register.  So it's a classic serializer, and basically used as an SPI IO expander.  And despite the 2-3m cable, no supply bypass cap -- to be fair, at 5V it's slow enough it likely doesn't need one. Not sure offhand what drives it, from the main board; that would be a major source of noise (EMI emissions) -- I mean, it passed, back in the day at least, so it can't be too bad.

Tim

The pullups measure 3.5k. The keys when pressed measure:
~300-500 ohms for the A button
~1k-2k for the B button
~200-400k for the select button with intermittent contact.

I am quite sure this controller has those pads well gnawed (at a slant too!) so it has to be substituted.
I guess there's no in-home way to fix that, am I right?

Anyhow I'll be back with more from the second controller, but I guess it will be just the same

EDIT: yeah, basically the non working buttons show ~40k when pressed unlike the ~100-400R of the good ones