Wavy Lines on NES-001 - but there's something more to it...

Hello all,

This is my first post, so, I apologize if it's in the wrong forum, but this looked like the appropriate place for this topic. All my research online can't seem to yield a workable solution for this problem.

This started about four or five days ago. I was playing my front-loading NES, model 001 from 1985 when I discovered horrible looking wavy lines pervading the screen. The actual game graphics were totally fine, but, it looked like there was a problem with the composite cables I was using. The screen basically looked like this:



I always have my NES plugged directly into my 46" Samsung LCD (as I do all my other game consoles). My first instinct was to try a different set of composite cables, but, it didn't solve the problem. Next, I tried it on an old CRT television with the NES RF Switch, but, it had the same issue. I tried it on another LCD, but, same deal.

After doing some searching online, I came to discover a lot of people suggesting it was a problem with the power supply. I always use the official NES AC adapter, but, it was plugged into a power bar, so, I tried plugging it directly into the wall outlet. The problem didn't go away. It was also suggested elsewhere that it's possible there was a problem with the capacitors. Being no expert in soldering and my understanding of how circuit boards and so on work, I figured the next best alternative was to swap out the motherboard from another NES with a non-functional 72-pin connector into the one I was having problems with since the pin connector on the one I always use is totally refurbished. After making the swap and using a different official NES adapter, the problem appeared fixed! The wavy lines were gone and everything was back to normal...

However, two days later, I went to play the NES again, and the wavy lines had returned! How can that be possible? After making the motherboard swap, which included the capacitors, it was basically a completely different NES. The swapped out motherboard was from a later model of the NES, specifically a Mattel Canada variant. The original was from launch in 1985. It seems very strange that two completely different motherboards and setups would decide to experience the exact same problem within a day of each other.

It was suggested on a YouTube video comment section that possibly swapping out the power/reset buttons could solve the problem. However, after I attempted that, it didn't help. The wavy lines were still present.

I'd like to know exactly what causes the issue to crop up, if possible. I've had my NES setup like this, basically unchanged for almost six years. Why would it suddenly happen like this? And, short of going to a top-loading NES, is there any other solution I can try or changes I can make to the environment to possibly remedy this situation? I'm not sure if this is related or not, but, when the problem happened again on the swapped-out motherboard, I tried the RF switch again on the CRT TV and it the sound was making a horrible buzzing noise that didn't happen before. Like I said, I don't know if it's related to this problem or not, but, it might be another clue.

I'm basically out of ideas. Any comments or suggestions would be appreciated.

Thanks!

The really weird thing is that it's not constant from scanline to scanline. The source of interference is somewhere around 1.35MHz or so... That strongly implies the interference is not from inside the NES, because everything inside is running from the same clock.

Did you move your NES physically around the room after those couple days?

No, I put it back where it was originally. By the way, that screencapture is NOT from my NES. It's from a YouTube video that was producing the same problem, however, the results are largely the same.

I see the same thing. Often, moving the video cable around fixes it. Is there an AM radio station on the ~1300 kHz frequency in your area?

Not that I know of... and if there were, it must have just sprung up overnight because I was playing the NES a day before the problem originally started and there was no indication of any wavy lines at all.

However, something environmental would account for the fact that the swapped out motherboard experienced the same problem inside such a short time.

The moving of the cable didn't do anything - like I said, the issue was prevalent with both the composite cables and RF switch connection.

EDIT: New clue: Holding the power button in affected the intensity of the waves... they were much more pronounced with the power button being held in.

That doesn't make any sense ... unless you're acting as an antenna and coupling to the system.

So, I've just tried a THIRD NES toaster and the same issue of wavy lines is appearing on that. So, it's got to either be a problem with the power adapter itself or something environmental. However, the Super NES, Nintendo 64 and Wii I have connected to the same outlet on the same television exhibit no abnormalities.

You guys ever hear of the power adapters themselves becoming defective?

An entirely different NES or just a third set of guts?

An entirely different NES, ungutted.

The NES is unique among that set because it takes 9VAC input, and rectifies and filters that inside, whereas the SNES and later ones use DC supplies (of varying sophistication).

A really simple test would be to use a stack of batteries to make >=7VDC and supply it to the NES (at the normal AC power jack) and see if this interference still appears. I have this hunch that it will, though.

(You could even use a normal 9V battery but it won't last long ... those things are only good for ~5-10m at the kind of power draw the NES has)

Another thing to try is a Sega Genesis model 1 power supply, which produces DC at a voltage that the NES's regulator accepts.

Is a ground loop between the NES Control Deck and the TV possible?

It depends - I'm no electrical genius, but, what would cause a ground loop?

Besides, it seems unlikely as I tried it on a completely different TV in a completely different room on a completely different outlet with the same result.

EDIT: Some new information - originally, I was only using a copy of Super Mario Bros. to test the screen, which produced the wavy lines. I then put in Dr. Mario and they were GONE - totally. I switched over to Metroid and again, gone. I put SMB back in and back, though diminished. I thought maybe it had something to do with the darker color schemes of Dr. Mario and Metroid affecting it, but, I also tried The Legend of Zelda, Family Feud and NES Open Tournament Golf and they were mostly clear. There were some waves on the title screens of Zelda and Golf, but, they dimished after about 10 seconds. Then, the strangest bit yet, I tried the Super Mario Bros./Duck Hunt cartridge and got no waves on the SMB title screen, switched back to just the SMB cart and had diminished ones!

What in the heck is happening here?

Bad caps in the PSU/Mobo is all.

But he tried 3 different NESes and they all exhibited the behavior.

It could very well be your TV, or interference from something near your TV. If not the power adaptor, that is.

Well, now I've tried a Sega Genesis adapter, and the waves were still there.

I really don't understand this. They seem to vary in intensity, sometimes they're very pronounced, sometimes they're hardly visible at all. There's just no rhyme or reason, that I can see. I even went so far as to unplug all my electronics and re-set them up, making sure they were all straight cables and no effect.

The last thing I can try to is to take the NES to another place entirely and see if it affects any difference.

As Drag suggested, if it were something near my TV, you'd think the problem would have eliminated when I moved the NES to another room entirely, but, it didn't... and I haven't added any new electronics to my set-up in quite a long time, so, I don't know what could be causing it. I'll report back after I've tried it at a friend's house, which is in the next town over.

So here's the itemised list of shit that's supposedly been tried:

- Swapped composite cables; no change
- Different LCD; no change
- Using a CRT television + RF output; no change
- Three (3) other NES control decks (original models ("toaster")); no change
- Using a Sega Genesis AC adapter; no change
- Some insanity about swapping cartridges/games and suddenly the problem being lessened

I'd say just get over it, I see the lines all the time on nearly every NES. It's just what it does, is it that bad? Play on an emulator if you want it perfect.

Changing cartridges changing severity ... vaguely implies lousy power regulation?

Oh sorry, I missed where you said you tried 3 different TVs.

I have no idea, unless it's some kind of house-wide issue with the power or some kind of global interference. I dunno what it is.

Haha, any chance there's a solar flare screwing around with EM interference the last week or so?

Hm. Since you have an open one, it'd be pretty easy to test for whether this is due to voltage rail noise or something else—tap a small capacitor (maybe a 100pF?) between Vcc and Vout (pins 40 and 21 on the PPU). If it gets worse, then the answer is probably ... if so, try then moving that capacitor between ground and Vcc (pins 20 and 40) .

A wire connected only at one end makes a rather effective antenna.

A bad ground pin on one end of the cartridge (perhaps it's just dirty?) leaves the cartridge's ground plane connected only on one end.

Is it possible that this is what's happening?

Maybe you could rule out EMI by putting your NES in a Faraday cage. I suppose a box lined with heavy-duty aluminum foil, with holes cut for controllers and A/V cables would do it. Not sure if it would need to be grounded, probably not.

It's possible, but, I would tend to say unlikely... when I cleaned my NES games, I used a very heavy duty oven cleaner to take off the 25 years of grime and tarnish. My carts are basically like brand new. I suppose it's possible, as I did this cleaning about four of five years ago, but, they appear fairly pristine.



I actually did try this, but, no luck.

EDIT: So, after giving a rest for a couple of days, I tried the system again. A few matches in Punch-Out!! and a couple of Dr. Mario rounds and the waves were still there. After about 10 minutes powered off, I tried the Genesis adapter again, just for laughs. First, with no cartridge in, just powered on. The waves were present on the purple screen of purgatory, so, I tried it with a game in. They were gone.



This is really confusing, considering that I tried the Genesis adapter a few days ago and it didn't help at all. I tried it with the official NES adapter again immediately after and they were still gone. I'm wondering if something in the system needs time to "warm up", which would cause the waves to dissipate over time? Does that seems logical and, if so, what component would it be and why would it suddenly happen now after years of reliable performance?

EDIT #2: After they had gone from the screen, I went up to press the POWER button. As soon as my finger touched the button, without even pressing it, the waves returned. I also noted a peculiar buzzing sound coming from the speakers when these waves were visible on-screen. I wonder if the power switch needs some kind of insulating?

It could be crosstalk between the A/V ports on the side of the NES. Mine are like that, and I have to wiggle them to get the best picture and to minimize the buzzing on the audio.

I don't know but most times I've run into something like this it was caused by capacitors. A few times the fix was a serious cleaning job.

The NES seems to be some magical exception but electrolytic capacitors just don't last this long. Especially if the board was used a lot or had dirty power. I'm amazed that I don't hear much about NES recapping. I've never seen an NES that needed a recap but I've had to recap a genesis and a ton of old 68k macs. The forums for the macs are all about recapping. Game gears, portable turbos, all sorts of old systems are notorious for it. I know the NES isn't from the capacitor plague era but good electrolytic capacitors still have a limited lifespan and we're way past it. Your NES hasn't been recapped has it? Might be worth a thought. There's nothing I trust less than a capacitor.

I've had a lot of luck washing boards too if you don't want to go that far. Had some pretty finicky carts that ended up being a lot more reliable after scrubbing lightly with a toothbrush and some alcohol, and then washing the whole board under the tap. Even though I already q-tipped them with alcohol. If you see evidence of capacitor leaks this has a really good chance of working though it's obviously temporary at that point. I've had good results with a clean looking board too though. I haven't washed an NES motherboard just games and mac boards.

With the macs a lot of people bathe the board in alcohol/water, water and soap, or just water. Even a really clean board with no signs of leaks that's acting up can come back to life with a serious cleaning. It sounds ridiculous but everyone seems to get away with it. I've even put stuff in the dishwasher (without hot water). Just put a fan on it after and leave it for a long long time to be sure it's completely dry.

So, give it a good cleaning, you say? How water tight is the RF modulator? I wouldn't want to get any drops in there. As far as the circuit board goes, just scrub it with a toothbrush and rubbing alcohol and rinse under the tap? I'm game for it as long there's no risk to the motherboard by doing this.

I'd sooner give that a try than have to worry about soldering and recapping - I'd rather save that as a last resort type scenario as I've never attempted it before.

Yeah... I was wondering about the RF part. I'm not sure if it would really be any worse to wash it but I don't know much about it.

I've definitely scrubbed the hell out of some boards and had them still work. I use a paintbrush if I can, though. You probably don't even need to scrub it much just a gentle pass to loosen up whatever might be on there.

Alcohol might take off some of the green on the board but nothing too bad unless you go crazy on it. Don't grind hard at it. I usually run water over it and scrub it a bit, then do a little alcohol and scrub that a little, then lots of water to rinse it all off. I scrub chips a little harder than traces. Some people use distilled water at the end to get the tap water minerals off but I've never bothered.

One place you can really go crazy with a toothbrush and alcohol is the cartridge slot btw.

Just make sure its dry. If I'm in a hurry I'll hang the board above a dehumidifier and it's dry as a bone in no time. That could be bad advice though.

Really all of this could be bad advice but I haven't had any problems with it. Made me feel like these boards are a little less fragile than you might think. Some boards are washed at the end of manufacturing I believe.

You can look for something like "washing circuit board" on youtube, there's a few videos of people's favourite way to do it and some of them talk about the risks.

Recapping isn't too bad. The thing that saves you a lot of grief is that you can clip the leads off the caps since you don't need them anymore (if they're not surface mount). Then just grab one lead at a time with pliers, heat it up and pull it out gently.

I tried the cleaning the motherboard method, very carefully, with Q-tips, rubbing alcohol and oven cleaner. Hit all the areas that looked dirty, sticky or corroded, including the main cap in the RF mod (which didn't look puffed up or dirty to begin with) and still no dice. Still catching waves!

Here's a silly question: What if you put an AM radio right next to the NES when you see the wavy bits? You might need to explore with what exact band it's tuned to. My NES doesn't produce those wavy bits, but the cartridge itself radiates weakly at 1340kHz.

When I see a high frequency, I try to relate it to known frequencies in the system. Here, 1340 kHz is one-fourth of the 5370 kHz dot clock of the NTSC NES and Super NES PPU. Are the wiggles spaced about 4 pixels apart?

Hmm. Did you really submerse it?

I think the only real reason fully covering it in water/alcohol can be a step above q-tipping it is because you can flush out all the little areas you can't get to, under chips and whatnot. I sort of feel like I'm just pushing dirt around with alcohol sometimes. I'd imagine if you brush the area next to a chip the alcohol will spread out under the chip and then evaporate, leaving some crud behind. No scientific evidence here though just a vibe. And dirt doesn't always cause a problem, it can be a shot in the dark.

I've seen non-working macs start working, no-sound macs suddenly have sound. But those have so many interference problems, some people get one and recap it before they even turn it on. Just a more complex board or something. You're more likely to get results when you've got so many problems on one board I bet. Your NES has 1 minor problem whereas the stuff I did this on had quite a few minor and major problems. So I guess your chances of success would be lower.

I did once have this sort of problem on a genesis and it ended up being mostly due to a bad voltage regulator. If I recall, it seemed to put out the right voltage when I checked it with a multimeter but I had no scope to look at the waveform. I changed it anyway and it did come out better. Though not perfect until I changed the caps.

If nothing else at least if you've gotten away with putting water on it. I sure felt better when I realized you can be a little rough with circuit boards.

Btw if you're thinking about recapping but you're still a bit scared to do it you can always find some board you don't need from somewhere and practice on that. The two things I watch out for when I remove caps is holding the iron way too hard on a via and then it just lifts off with the solder like it wasn't attached in the first place, or pulling too hard on a lead as I'm desoldering. If the solder hardens while you're pulling it out you can rip off a trace pretty bad that way. But if you're gentle and back off of a via when it's not melting for some reason rather than grinding the iron into it, you should be fine. My soldering iron is pretty bad but I've never ruined a board or even a single trace/via except when I'm trying to recover a chip with no regard for the board it's on.

I just have such a strong distrust of capacitors after all this mac business. I feel like it's hard to narrow these kinds of problems down if you haven't recapped. So many things could happen but bad capacitors eventually always happen. They're so likely to go that the second I see something like this I usually think about recapping the board. The NES, again, is some kind of magic for even working this long. I know in some cases capacitors are absolutely necessary and some are just extra insurance for clean power that the board will work without. Maybe some of them just aren't that important on the NES. I'd be interested to know why it's not such a problem on the NES if anybody knows.

Anyway, sorry for all the capacitor rants. haha.

Interesting! That looks about right to me (roughly 3-4 pixels) when looking at the initial screenshot (which isn't the OPs, it's just an example he found).

Caps do suck, even today. Most cheap PSU's buy cheap caps that dry QUICK because they're POS's, causing tons of failures. Any discount computer or PSU probably uses those caps, unlike my corsair 950 watt desktop PSU, which uses solid caps.

Yeah, they are. The problem is that they're not at a constant phase relative to the pixels, which they should be if it's noise directly generated by the PPU.

For a better point of reference, this is a screencap from my own system. This was from the first motherboard, before I swapped it with the second one, but, the results are largely the same.

Probs the TV PSU. Keep the NES on, turn the TV on and off a lot.

So, I'm finally going to bite the bullet and try replacing the main cap in the RF mod - the 2200 uF/25 volt one.

In case I can't find an exact match, is there any other spec I can use that's compatible?

Higher capacitance and/or voltage rating will be fine, maybe even preferable. Just don't go lower. Dimensions-wise, a newer capacitor of the same rating will be much smaller than that old one. If you have a choice, I would go with decent low-ESR caps like Nichicon HE series, or Panasonic FC series.

I'll see what I can find at RadioShack. I'm reading elsewhere that apparently the back of the RF mod just pops off without having to desolder it from the circuit board. Is this correct?

Yeah, the cover pops off. Desoldering the RF box itself is a pain, it takes a lot of heat.

Well, the only appropriate capacitor at RadioShack they had available had one the pins on either side, so, it's not exactly like the one that's in the NES. I don't know if this is even going to fit in... it certainly won't fit in that hole that's in the RF mod because it's going to lay horizontal instead of vertical and even if it can lay horizontal, I doubt the pins are long enough as they are. Can I still use this cap or should I order online something else?

In that case, I'd look for one with higher values that would fit better. The values of the capacitor aren't too important, as long as they're at least as much as the original 2200uF and 25V.

I thought the rule of thumb was same capacitance, same or higher voltage. Just in case a capacitor is filtering something specific only. I know you can generally get away with higher capacitance but is there no scenario where higher capacitance would mess with the original intended function of the capacitor? Like in the case of something like a high/low pass filter?

The huge capacitor there is a power supply filtering one. Larger is ok, as long as the equivalent series resistance (i.e. maximum instantaneous current) doesn't go up (down) too much. (It probably won't)

After some consideration, I'm going to try and hunt down a radial one instead, just so it fits into the slot in the RF modulator like it should. I feel like if I try to shape the pins on the axial one I have to fit, they're just going to snap anyway.

All right, I got my hands on a radial 2200 uF/25 volt capacitor, installed it and... no effect. Those damned oscillating waves are still over the screen.

This is ridiculous now. I've done pretty much everything I can to this thing to try and correct this problem and nothing. No dice. What else could this be?

If anyone else has a theory or a suggestion, I'm wide open. Heh.

I still think you report back on any of my suggestions

Unfortunately, I don't have an AM radio to try that experiment and...



is unfortunately completely over my head (as I mentioned, my knowledge of circuit boards and the shorthand lingo is minimal). Add to the fact that I'm in a small town in Canada where we don't have RadioShack or any place that readily sells capacitors anymore (which completely sucks) - I had to get the 2200uF from some backdoor alley place (quite literally) in a city 40 minutes away. If you'd care to elaborate on how I can try this test, I'd be willing to give it a shot.

What electronic parts do you have handy? I might be able to suggest a workaround.

Not very many. I have the original capacitor out of the NES that was removed to make room for the replacement and the unusable axial one. Both are 2200 uF/25 volt. Tons of assorted cables and so on, but, it's slim pickens beyond that.

This is kind of a doofusy question and it might've been answered already, but have you tried plugging in a different game console (like a SNES) and seeing if the lines are still there?

Yes, I tried the RF switch on the SNES and the lines were not there. It's definitely a problem indicative with the NES itself.

Bad caps is the problem with the console, the same as it was 10000 posts ago on this topic.

So. Picture shows a varying frequncy, so is some kind of FM. Frequency seems to vary from 1280kHz to 1240kHz, by counting pixels and comparing to the known pixel clock. Assuming output has been (incorrectly) deinterlaced by the television, two subsequent frames show almost identical timing and phase, except in the middle of the frame where the boxer's gloves are. So whatever signal's been modulated, it's basically only got multiples of 60Hz in it. (Specifically ... looks like 120 Hz plus some higher even multiple for the asymmetry) If it were a beat pattern against mains or NTSC frequencies, it would shift by a half scanline or 70% of a scanline every vblank—too small to see in this picture, I think.

Question: just how stable are these waves? do they move while the NES is on? How fast?

Assuming it is bad capacitors, and having already replaced the big on on the input stage to the 7805, you could try adding the one of those extra capacitors you have (the axial one should be easy, I'd hope) in parallel to one of the existing capacitors (Naively I'd put it right next to the PPU)

They move down slowly while the NES is on, as if it were beating between 60.1 Hz (NES and Super NES vertical frequency) and 60 Hz (fluorescent bulbs or PSUs) or 59.94 Hz (vertical frequency of other video equipment). I'll have to count the beats more precisely to see which.

EDIT: I just measured it, and it takes about 20 seconds for the wiggles to progress all the way down the screen and wrap around twice. This gives 10 seconds per loop, as if it were beating between 60 and 60.1 Hz.

See for yourself... http://youtu.be/dlMoNiIjrX4



So, you're saying literally touch the axial capacitor to the circuit board while it's running? I hope that's not dangerous.



Forgive me for being blunt, but, this comment doesn't help the problem at all, so, if you have nothing constructive to add, perhaps you should ignore this thread. As was mentioned earlier, the main 2200 cap has already been replaced with no effect. If you know another capacitor that could potentially be the culprit, suggest which one it is. There is very little information available on this problem online and most people who have encountered it on other boards that I've read either don't report the progress of any solutions or just forget about it. Since this seems to be a recurring problem with many systems, I would hope by the end of this, we'll have a working solution that other people can look to to get their systems operating correctly.

I put in an NES cartridge that shows a light gray test screen and did some things while the power was on:

While viewing through composite, the exact pattern of wiggles changes noticeably when I unplug the cable to the RF switch and changes back when I replug the cable.

While viewing through RF, the pattern changes slightly when I unplug the yellow cable and changes back when I replug the yellow cable.

To be blunt, filling the forum with tons of posts that aren't relevant because they don't help us get farther either is why my comment exists. We told you what the problem is, confirm it when you can, otherwise why fill itwith how many combinations and whatever obviously will work and not, we already know that. My comment is as useful as all others after we posted what the problem is.

Did I not do that by informing everyone that replacing the main capacitor did not help matters?



I'm sorry, but... what?



Fine, you say it's a capacitor problem? Which one? There's a few of them on the board and we already know it's not the 2200 one. So, where to now?

No, not while it's running. Just clip it in and see how the NES runs while you're going. Ideal places to add it would include "across the PPU" or "across the PPU's bypass capacitor". Here's a picture:
Ok, that's pretty clearly power supply crappiness.

Well there's more than one capacitor in there.

If any of them are doing a shitty job of filtering power (that's their job, and they retire at age 15) then that will be making it through to the video signal. Nobody replaces one capacitor and calls it quits. I wouldn't really expect replacing one capacitor would make a difference to be honest. Would you replace one spark plug in your car? Btw I think if there's any capacitors I've NOT replaced, it's the big ones. They seem to hang in there longer and I don't usually have them on hand. All the little ones have to go.

I think that's what 3gen is talking about here. Until you replace them all there's not much point in looking deeper.

edit: This assumes your power supply is putting out nice, clean power. New capacitors won't fix a terrible power supply.

Fair enough. As I've mentioned many times in this thread, I'm relatively new at this level of repair work. I've replaced the pin-connectors before, that's it. As far as capacitors and modulators and circuit boards and so on go, I'm very inexperienced beyond what I've picked up as I've investigated this problem. Since it was suggested that the main 2200 cap tends to be the culprit, I started there.

I'll search around online and see if I can find a pack that has all the new caps in it for an NES. Being in Canada, it seems that you can only get stuff like this online, anyway.



I think we can safely rule out the power adapter itself because using a Genesis one didn't change anything... unless using a bad power adapter can damage the NES?

Nah, switching between different unregulated power supplies won't change things much.

An obvious workaround (although it's suboptimal for other reasons) is to use a newer-style switching DC power supply of 7-10V or so. They won't emit any 120Hz (from the NES's AC transformer and full-wave rectifier) and so should make the waves invisible (because the switching frequency will be in range of 10s to 100s of kilohertz)

I had an NES that only gave a grey screen and in the process of trouble shooting I decided to just recap the RF module. I used the right type and everything, got them from Game-Tech website. Before I did the recap I had a solid grey screen but afterward I now see these waves as pictured and discussed in this thread. Could it be a bad capacitor? This was my first soldering job so maybe I did not have strong connections or cross connections? What could I try to fix this issue given that I actually created it as explained?

Check for cold solder joints—they have a higher resistance than well-flowed ones. If you're using lead-free solder they're easier to create.

To clarify... lead-free solder is easy to get well-flowed solder joints? So you recommend that I use lead-free?

Er, no. It's easier to create a cold solder joint when using lead-free solder.

Not dramatically so, though.

Have you tried replacing the voltage regulator? I had the same issue with a Super Nintendo. I bought a new, high rated 2 amp 7805 from console5.com and when I soldered it in place it made the video crystal clear. I also recapped it first, but the problem wasn't solved until I had put a new voltage regulator after putting all new caps.

I am experiencing the same exact problem with my NES-001. My toploader does not experience this problem at all.

It may or may not help, but some time ago I had a similar problem with a Mega Drive.
It had only slight walking bars, but had a terrible noise of circa 60hz in the sound.
I tried lots of things, but in the end it was the power switch that somehow has inducting that noise.
Maybe you could try a different power switch and see if it helps.

Power switch or power supply?