NES repair advice needed

[alank2]

I was having a problem with cartridges not working and they were real sticky trying to push them in or pull them out so I opened up an original Nintendo entertainment system to find it had a bit of what looks like dirt/mud that somehow got into it. I cleaned the contacts on the 72-pin connector, took it all apart, washed it out, rinsed the board carefully, and blasted it with compressed air and got it back together. This improved the sticky issue with the cartridge, but it didn't always work when inserting cartridges so I thought perhaps I needed to also lift the pins on the connector which I skipped because they did seem high enough to me already. Took it apart, lifted the pins, and then noticed I hadn't put the underside of the RF area metal case back on it the first time, at this point it is not plugged into anything and I go to put the small 2" x 2" plate back on and somehow it comes into contact with a solder joint and I see a small arc/short occur. There is a 2200uF cap on it which rectifies an 8 VAC loaded signal coming in from the wall adapter so I have to presume that it is one of its leads that shorted to ground. I finish putting it together and now here is no composite video output. I tried different cartridges and still no video. Sometimes the red light blinks which I think is some sort of protection thing, and sometimes not. I disabled the protection chip by cutting pin 4, still no video. I take it all apart and look at the video pin going to the RF module, it has very little voltage on it when powered on (somewhere between 0-.3 V), but no signal of any type on the scope. I pulled the transistor connected to the video line on the RF module thinking maybe it was zapped and loading down the video signal, but it tests okay when pulled so I put it back. Voltages seem mostly good, the voltage regulator outputs 5V when powered on and the IC's are getting it. Sometimes I see a sawtooth type pattern on it, but I don't think that is related to it not working, but maybe that the heat sink is getting too warm not being connected to any of the metal shields. I looked at pin 21 coming out of the non CPU IC because I saw a schematic where it showed this was video out, same thing, no signal.

I don't know if it is supposed to put out a video signal even with no cartridge inserted, so that is one problem, perhaps my 72 pin connector is really the problem and it just can't read a cartridge, but I wonder if the short I saw did something, maybe even allowed a negative voltage spike to ground which damaged an IC or something.

I saw a test point oscilloscope thing for the NES on a site, but it doesn't mentionen where the test points are and I don't see anything labeled on the PCB either.

I'm trying to desolder the RF module from the main board so I can get the other metal shield side off of it to look around for something damaged.

 

Any thoughts on what to look at next?

[emerson]

Here are some questions I have:

- How did you go about washing/rinsing the pcb?

- How long did you let the pcb dry before connecting power cord (regardless if the power switch was on or not)?

- Do you have valid a clock signal on pin 29 of the cpu?

- As for the video signal, is the picture black or grey?

- What kind of signals do you get on the audio and RF outputs?

 

Start by ensuring your system clock is present and at the correct frequency. At pin 29 of the cpu, my console measures just under 21.5MHz at 2.30Vpp, with around 0.875V DC bias. The onboard crystal states 21.47727MHz so I'd say that's good.

 

An unmodified NTSC NES with no cartridge will output a grey screen when powered on and the power light will blink. The video will cut in and out as the power light blinks. When the power light blinks on I get a standard composite signal which peaks around 640mV. When the power light blinks off I get a flat 640mV DC signal. For video measurements I am using a 75 ohm terminated adapter on my scope connected to the composite video output on the NES. If you're using an analog scope you may need to turn your brightness way up to see the video signal.

 

An NES with pin 4 of the security chip cut and no cartridge will output a constant grey screen with no power light blinking.

 

The blinking light indicates the security chip in the console did not receive the security code it assigned to the cartridge, thus putting the console in an endless reset loop while looking for the correct code. Assuming you're testing with a licensed Nintendo cart, the only symptoms which cause the blinking is a worn out/dirty pin connector or a dirty cartridge. The communication lines for these two chips are on the far right side of the pin connector (or left side of cartridges) and depending on your cleaning methods may not get as much attention as the middle pins. This is why sometimes a game will start to load but still reset itself. In your situation cutting pin 4 on the security chip only means you won't have to clean 4 of the 72 pins (or 8 of the 144 pins if you add the cartridge) as thoroughly as the others.

 

I would replace the 72 pin connector with a known working one.

[alank2]

Thanks for the help - I really appreciate it.  I'm familiar with washing PCBs as I manufacture some of them, in this case I gave enough water to try to toothbrush scrub some of the junk away followed by a little more rinse to rinse it off and finally an air compressor blast to dry it.  Then I waited a bit for it to dry further before reassembling.  I tested it after cleaning and it worked, but the game cartridges were not reading cleanly, sometimes they would work, sometimes not, sometimes with bad screen artifacts so I thought I would try to pull the bottom pins on the cartridge connector up a bit with a hook.  I thought they were high enough, but perhaps not.  This is when I saw the RF cover hadn't been replaced and went to put that back on and saw a small short occur within it.

 

I am not getting a constant grey screen even with the 72 pin completely unplugged and the video signal itself has no waveform activity, just a constant level.  I may be getting the flat 640mV DC signal all the time, but I will check it.

 

I will check the other things you suggest and report back!

 

[alank2]

It has a clock on pin 29 of the cpu IC, but not as strong as yours.  850 mVpp, 500 mVavg dc bias, but it is 21.475 Mhz.

 

Thanks for confirming the gray screen, I'm getting no signal at all on the video output or audio outputs.  RF modulator removed so I am sampling the two pins that connect to it.

 

Video is 1V dc bias when powered on, no waveform or signal.

Audio is 2.5V dc bias when powered on, again no waveform or signal.

 

I am feeding 5V directly into it from a DC bench power supply, it consumes anywhere from 1.1A to 1.3A when I power it on.

 

[alank2]

This inputs and outputs on the 7404 (U9) look good, one is in the middle with its outputs in the middle (pins 10/11).

 

The 74LS373 (U2) has no action going on with its output control pin 1 or its clock pin 11.  Do these come from the PPU?  I wonder if the PPU is what was zapped.

 

BTW, the center of the CPU and PPU get VERY hot and in burn your fingers if you touch them hot - is that normal?

[emerson]

8 hours ago, alank2 said:

I may be getting the flat 640mV DC signal all the time, but I will check it.

If you recall, I measured the 640mV from the composite video output of the rf modulator with a 75 ohm termination in circuit. I used the same output you would connect to the tv. If you measure the video signal off the main pcb you will almost certainly get different results than me. Moreso, you will likely get different measurements from the same test point with the rf modulator removed as well as attached.

 

4 hours ago, alank2 said:

This inputs and outputs on the 7404 (U9) look good, one is in the middle with its outputs in the middle (pins 10/11).

When you say it's in the middle, do you mean around 2.5V DC? It was common back then to use CMOS inverters as cheap linear audio amplifiers. That is what's happening here. You'll also find a similar configuration in the player 2 famicom controller.

 

4 hours ago, alank2 said:

I am feeding 5V directly into it from a DC bench power supply, it consumes anywhere from 1.1A to 1.3A when I power it on.

Where in the circuit are you connecting the 5V bench supply? Also, look at your current measurements. Both of those values are above the stated current of 850mA (check the bottom of the case) as well as the 7805 max current rating. This suggests you have a short somewhere. I installed one of those cheap digital voltage/current meters in an NES when I was designing my rumble controller and it measures 0.33A with no cartridge installed. My current is measured on the 5V line at the header connecting the rf modulator to the main pcb.

 

4 hours ago, alank2 said:

It has a clock on pin 29 of the cpu IC, but not as strong as yours.  850 mVpp, 500 mVavg dc bias, but it is 21.475 Mhz.

Your frequency seems good but as you said, your amplitude is lacking. Another indicator of a short.

 

4 hours ago, alank2 said:

BTW, the center of the CPU and PPU get VERY hot and in burn your fingers if you touch them hot - is that normal?

No, that's not good at all. It's possible one of these chips is the cause of your short. Depending on how good you are at desoldering you could remove one chip at a time (start with whichever one gets hotter faster I guess) and check your current draw. If possible, install sockets for any chips you remove. Do so at your own risk.

 

[alank2]

Using a thermal camera, only the CPU and PPU heat up, and even at that, the CPU heats up quite a bit faster/hotter.  It is probably the problem, but perhaps I can check all of its pins to see if anything is obviously shorted that should not be.

[alank2]

I ordered a replacement board off eBay for $35.  I may still try to fix this one by comparing the two of them if I can find cheap parts to try.