Sharp NES freezing up

[Cornelius]

I'm having problems with my Sharp NES system freezing up. Some games seem completely fine, my kids played Bubble Bobble for a really long time no problem, for instance. SMB doesn't seem to have any problems. SMB3 seems okay, but I didn't play it for long enough to really know. However, other games I've tested with freeze up right away, like Contra, which always froze on the first level. Dr. Mario seems to work okay for like 30 minutes and then it freezes, sometimes having some glitches for a little while before that (e.g. pills coming while something is still falling, or pieces not falling when they lose support).

These games work fine on separate NES systems.

Games load easily on the SharpNES, but I also cleaned the pins and have installed a BLW, neither of which helped at all.

I looked around and people with similar problems have been told to replace the voltage regulator. Thought I was good after doing that, because I played around 5 levels of Contra without issue, but then it froze again. Still freezing on Dr. Mario. Next up are capacitors, so I replaced the three on the power supply, and 3 of 4 on the main board. The 4th one isn't labeled with polarity, so I'm not sure what to do about that one, just left it for now. With that all done, no change on the freezing up front.

Any suggestions? I could go after that last capacitor and just put it in the way it would be on a regular front-loader. I'm thinking I need to start looking at chips, but I can't see any way to do that. There is certainly no easy way to power the thing with it open/accessible.

Here's the board and cap in question pointed out:

[Cornelius]

Here's the two boards next to each other. I was thinking maybe I could do a swap or something, but that seems well beyond my capabilities.

[jb143]

I could go after that last capacitor and just put it in the way it would be on a regular front-loader.

Unless they drastically changed the board design(and it doesn't look like they did) then they should be the same. There is also such a thing as an non-polarized electrolytic called a bipolar capacitor...but I would lean more toward the first option.

[Cornelius]

I'm thinking that's what it is, actually. It has "B P" prominently printed below the specs. Thanks for pointing me that way. Now, since I don't have that cap, can I replace it with a polarized one like it is on the front-loader? I'll do some reading.

Unless they drastically changed the board design(and it doesn't look like they did) then they should be the same. There is also such a thing as an non-polarized electrolytic called a bipolar capacitor...but I would lean more toward the first option.

[jb143]

It's weird that the regular front loader calls for a standard electrolytic and yours calls out a bipolar. But if it's on the board then they must have had a reason. Typically those caps are used if the design calls for an unpolaraized cap but the capacitance is too large for ceramic caps to be available/cheap enough
You can make an unpolaraized by tying 2 electrolytics in series (+ together I believe, you'll want to verify that). You'll use the series calculation to get your values(which is the same as the resistor parallel calculation). It's likely to be some value that's hard to find and may not fix your issue. At any rate, digikey and mouser should have an exact replacement.

That cap looks to be part of the lockout chip. Can lockout issues cause freezing like you're seeing or just the blinkies?

[Cornelius]

It's weird that the regular front loader calls for a standard electrolytic and yours calls out a bipolar. But if it's on the board then they must have had a reason. Typically those caps are used if the design calls for an unpolaraized cap but the capacitance is too large for ceramic caps to be available/cheap enough
You can make an unpolaraized by tying 2 electrolytics in series (+ together I believe, you'll want to verify that). You'll use the series calculation to get your values(which is the same as the resistor parallel calculation). It's likely to be some value that's hard to find and may not fix your issue. At any rate, digikey and mouser should have an exact replacement.

That cap looks to be part of the lockout chip. Can lockout issues cause freezing like you're seeing or just the blinkies?

I saw that trick with the two caps in series and may do that, you just get a component with half the capacitance. But I'm suspecting bad ram or other chip. I may try to hook it up so I can feel the chips during operation to see if on is getting hot. Bad idea?

[jb143]

How is power being fed into the board. I would assume a transformer replacing the AC adapter and everything else is more or less the same? Are you able to take measurements when it's running? To make sure it has a steady 5 volts before and during the freezing? Notice any smells when it's been running for a while? When you replaced the regulator did you add any thermal grease? I'd also search for tips on checking for a bad CPU.

Sorry I can't be of more help, I'm not familiar with the Sharp NES TV at all. If nothing else, I can give you ideas to try until someone who knows what they're talking about shows up

[Cornelius]

Yeah, that's about right, though I don't really know what's in the silver box where power comes into a regular NES. Picture of the power supply:



Running it while being able to test stuff is much easier said than done. I'll have to pull wires out of their routing completely and arrange things carefully. I'm sure it can be done, but a hurdle I haven't gotten around to yet. No smells. Used some Arctic Silver I had, then read not to use conductive compound, so I cleaned that off and used some other stuff.

Hah, yeah, I don't mind the help at all, I'm pretty much a novice at this stuff.

How is power being fed into the board. I would assume a transformer replacing the AC adapter and everything else is more or less the same? Are you able to take measurements when it's running? To make sure it has a steady 5 volts before and during the freezing? Notice any smells when it's been running for a while? When you replaced the regulator did you add any thermal grease? I'd also search for tips on checking for a bad CPU.

Sorry I can't be of more help, I'm not familiar with the Sharp NES TV at all. If nothing else, I can give you ideas to try until someone who knows what they're talking about shows up

[jb143]

What you could try to do is solder wires to points of interest on the board, such as the regulator and CPU power pins and run them outside the enclosure so you can take measurements while it's running. Just make sure you label everything and don't short anything out obviously. Overheating parts usually make a distinctive smell, that's why I asked about that.

I'm mostly going by what I remember being issues with a regular NES, and I'm pretty sure most freezing issues are either due to bad connections or power problems. It certainly wouldn't hurt to clean every connector again just to be sure. If it was a chip going bad, swapping one out from a working NES would be a pain but doable. You would just have to be careful not to pull up any pads or traces.

Are you talking about the RF modulator on the original? I think the power jack is inside of there. Yours shouldn't need one, at least I certainly hope not. Are they running composite to the TV? Like I said, I'm not too familiar with these...other than it's a cool collectors piece that would be a shame to have not work anymore.

[Cornelius]

Are you talking about the RF modulator on the original? I think the power jack is inside of there. Yours shouldn't need one, at least I certainly hope not. Are they running composite to the TV? Like I said, I'm not too familiar with these...other than it's a cool collectors piece that would be a shame to have not work anymore.

Yes, that's what I was referring to. You asked how power was being supplied, and I was just saying I didn't really know how it compares to what is in a regular NES. I think the rectifier(?) and other power stuff is all in the box with the RF modulator, but haven't looked in there to compare.

[Cornelius]

Okay, so testing with it powered on:

voltage regulator is being fed 10.5V AC, outputs 10.1 VDC. Right off this seems like something is wrong to me.

There are three wires feeding to the NES board, Red White White. The Red is 10.1V, White 22.3, White 0.

I don't really know what to be checking for on the chips, but they are all seeing 10v at pin 1(? - top left).

[jb143]

Ah, now we're getting somewhere. Those chips should only be getting 5V so I wouldn't plug the cable that connects it to the power board until you get the voltage straightened out. Can you take the power board out and power it separately? That might make things easier. Make sure the regulator is well grounded because that can cause high voltage output. Reflow solder, check continuity, etc... Also, check the pinout for the exact part number just to make sure it matches what it's supposed to. There are instances where a letter difference in the part number can mean a different pin out. A regular NES can use a 7805, I wouldn't want to assume the TV does too so I'll let you verify that. Also check the caps on the power board too.

I'm not sure what the relay is for. Turning on and off the NES portion? Either way, if all else fails, it's just a 5V power supply. It should be easy to fix, or even rebuild if you had to.

[Cornelius]

Ah, now we're getting somewhere. Those chips should only be getting 5V so I wouldn't plug the cable that connects it to the power board until you get the voltage straightened out. Can you take the power board out and power it separately? That might make things easier. Make sure the regulator is well grounded because that can cause high voltage output. Reflow solder, check continuity, etc... Also, check the pinout for the exact part number just to make sure it matches what it's supposed to. There are instances where a letter difference in the part number can mean a different pin out. A regular NES can use a 7805, I wouldn't want to assume the TV does too so I'll let you verify that. Also check the caps on the power board too.

I'm not sure what the relay is for. Turning on and off the NES portion? Either way, if all else fails, it's just a 5V power supply. It should be easy to fix, or even rebuild if you had to.

Yeah, before I realized it was getting 10v, I had the nes running. Oops, hopefully nothing is damaged. I am able to power the supply without plugging it into the NES. The regulator seems grounded to me, I get continuity of the ground pin to points in multiple places on the board. I reflowed solder pretty much everywhere on the board. This is the part I put in: https://www.mouser.com/ProductDetail/863-MC7805ACTG. This is the part I took out:

I already replaced the electrolytic caps on the power supply.

Yes, I think the relay is for turning the NES on/off, since that is a separate switch from the main TV power.

No change to to voltages I'm seeing, though I was getting 12.2 v into the regulator this time. Still 10.1 out. From the reading I've done, a faulty ground would make sense for causing the wrong output voltage, but I don't know how to check or ensure that ground more than it is now.

[jb143]

Yeah, that regulator should have the same pinout.

I forgot to mention that the output would probably read a little higher than 5v without a load. You can put a resistor across the output and ground to get accurate measurements. 1K should be fine. Just use ohms law to make sure the resistors current value is high enough.

- of the rectifier, center pin of the regulator, - of the caps and ground to the NES should all read 0 ohms.

Did you replace or test the rectifier? I know I've seen them go bad before. I was trying to think of a way that could cause a higher voltage out of the regulator but now that I think about it, a shorted diode in it could potentially cause the regulator to flake out or die... especially if there's no fuse I the circuit. I think you have to pull it out of circuit to test (diode test).

[Cornelius]

Yeah, that regulator should have the same pinout.

I forgot to mention that the output would probably read a little higher than 5v without a load. You can put a resistor across the output and ground to get accurate measurements. 1K should be fine. Just use ohms law to make sure the resistors current value is high enough.

- of the rectifier, center pin of the regulator, - of the caps and ground to the NES should all read 0 ohms.

Did you replace or test the rectifier? I know I've seen them go bad before. I was trying to think of a way that could cause a higher voltage out of the regulator but now that I think about it, a shorted diode in it could potentially cause the regulator to flake out or die... especially if there's no fuse I the circuit. I think you have to pull it out of circuit to test (diode test).

No, I haven't done anything with the rectifier. I only have a vague notion what a rectifier is.

... a shorted diode...? ... within the voltage regulator? Or on the power supply board?

I'm mostly lost with your post, I'm afraid. I'll do some reading in the morning and try to get up to speed, but I'm a newb at this stuff and definitely have to go slow.

[jb143]

The square thing with 4 leads is a bridge rectifier. It's basically just 4 diodes configured so that when AC goes in 2 of the leads then the other 2 are positive and negative. Essentially flipping the negative part of the AC sine wave positive. So you get a series of positive "bumps". It's not DC yet but it's all positive voltage. The big capacitor after the rectifier smooths this signal out. It's still ripply at the top but the voltage never drops below what the regulator needs.

If one of the diodes shorts out or opens up, then it can cause problems for the rest of the power supply.

I don't know if this is your issue or not but definitely something to check out. You'd have to test the rectifier. I do know regulator diodes can go bad though. I fixed a DVD with a similar issue once, only there was a fuse to blow so no strange voltage issues. After narrowing down the problem it was a quick fix.

[Cornelius]

I just have a freebie multimeter, but assuming the speaker looking thing is diode test, the rectifier checks out (open one way, ~540 the other). Though I am reading that they can check fine like this, but still fail in use.

[Cornelius]

Who's the biggest idiot of them all? This guy. I had the damn multimeter on AC when I was checking the voltage regulator and ICs on the NES board before. That's embarrassing. So I take it all back, the voltage seems to be fine. That was quite the pointless detour.

[Cornelius]

So looking at the service manual it seems to think IC5 or IC6 may be overheating if "Abnormality appears after playing the game a long time", so it says to heat one up with a hair dryer to see if it freezes. Guess that's what I'll do!

edit: but wouldn't those chips be hot? I mean, I can now set it all up and play till it freezes, and the cpu feels a little warm, but definitely not hot. IC5 doesn't even really feel noticeably warm.

[jb143]

Heh, I'm sure you learned some new stuff at least.

Typically the normal operating temperature range for ICs start to get into the too hot to touch range. Though a chip can still go bad and not get hot. Unfortunately I'm running out of ideas.

I worked as an electronics technician for a few years fixing circuit boards but it was always just assembled boards. Manufacturing errors... backwards caps, solder under chips, etc... Any previously working, now broken troubleshooting has all been fixing my own stuff. Just to say, I kind of know what I'm talking about but by no means an expert.