on/off switch pop when turned off

My amp I built pops quite loudly when I turn the power switch off. I've read to eliminate that put a cap across the switch. I did that, a 0.01uF 600v and no change.

Any suggestions?


  • Schematic.jpg
    211 KB · Views: 448
And as a power switch bypass, .01 uf @ 600 v is not enough voltage. Refrigerators and air conditioners can put 1000 v spikes on western power lines. Old TV's always had 1000 v rated caps, and that is what I installed in my ST70 and PAS2 tube amps. I got the idea from old Sam's photofacts schematics, decades before the internet. If you believe your supplier might lie on his cap voltage ratings, you might buy UL rated caps which are rated for the power supply they go on- typically 200 to 300 Vac for US market appliances. I believe they are described as x2 or y1 caps or something, although I am vague which rating you would use for a power off switch. The .01 uf @ 1000v cleaned up the turn off pop on my dynaco equipment, and the only failure in 35 years both vaporized the cap and arced over the turned off power switch, so I call that a lightning strike (particularly based on the weather that day).
Last edited:
Caps across mains or mains switches should always be AC mains rated and fail-safe: X or Y. The DC rating can be ignored.
Hundreds of thousands of 1000 VDC disc caps were installed in televisions across the power switch. Most of these did not burst into flame, and went to the dump sedately intact. Are all modern capacitor suppliers such liars now that they have to have UL/CSA/VDE/CE auditing their QA process? If you're going for a UL rating for your product UL etc. will insist on it, but otherwise?
Lying is not the issue. The issue is dielectric behaviour with constant AC stress, and fail-safe behaviour.

Many DC rated caps survived across the mains. Many did not. If the user was lucky the cap merely exploded. If unlucky, the cap caught fire or shorted, which then welded the switch contacts and started heating the supply cable. Modern X caps are guaranteed to behave safely, even under severe overvoltage. Given that they do not cost much, there is no excuse for DIYers to use ordinary caps across the mains except through ignorance of the issues.
Yeah, finding them is the big problem. I have supply of salvage X2s from PCAT power supplies, .47 uf not originally across the power switch, but if you want to buy an X2 or Y1 you have to call in or buy 2000 or something. x or y rating is not a column of the newark.com cap selector table. Newark catalog pages are not viewable by my computer, however convenient their $6 one day surface shipping (sometimes 2 days) is. Mouser is three days away surface UPS and $8 freight minimum; I buy from them occasionally, but in the index of the paper catalog, their is no listing for "safety caps" or "X" or "Y" caps, either under general or under film or ceramic . Everybody seems to love digitech, but it has been 4 years before somebody would tell me their warehouse is in Milwaukee, and their "purchaser rating" of the part, I spit on that. I remember when digitech was a little surplus flyer from Ohio (1975), and I already have enough surplus. I won't be buying anything from E-bay either, it is not worth keeping a balance in a Pay-Pal account to lose bids at the last millisecond to some hotshot with a supercomputer. Nor is it worth paying $10 freight ($12 from some vendors in CA) for one stupid part. Geography is an E-bay blank area if the vendor doesn't spell it out.
I'm just sitting here regretting I bought the TO220 On semi triacs without the built in snubbers: and wondering if I have to have x2 or y2 caps for switching 120 vac straight (1 amp fuse) with the triac I already have. The pass A75 triac schematic I copied from here has .047 uf in series with 5.1 ohm 1W as snubbers, and a .047 uf across the hot and neutral. The .47 uf x2's I have salvage are 10X too big.
Last edited:
It is 4 AM in Sweden. Read wikipedia.org, ceramic capacitor, about 5/8 of the way down the article it dives into x2 and y capacitors. There is a little picture of the 3 AC lines and where they go. I found an article on wikipedia once on x & y capacitors that went into great detail, but I can't find it now. The only thing I know for sure, when you cut up a dead PCAT switching supply, right after the fuse they have a .47 uf x2 cap between hot and neutral. They don't have a cap across the power switch, because nobody actually uses the power switch on a PCAT power supply, the computer switches the triac internally.
If I was buying a whole $300 amp I wouldn't worry much about $12 freight from CA, but I do when I'm buying $60 of electronics parts in one $6 box, instead of buying every $0.25 part in its own possibly $12 box.
Last edited:
Osvaldo, What about a MOV? I looked it up on Wikipedia and it seems to be the closest thing to what I am trying to achieve. But there is no mention of them in the application I am looking for.

Indianajo, you don't have to answer this if you are in bed... But, about the 0.01uf 1000v cap that worked for your Dynaco: Is the higher voltage rating the reason my 600v does not help? And, I don't think the X or Y type caps are what I want unless I can find info on them keeping the pop from being so loud. They are described as RF and EMF suppressors, so I don't know how they would work to eliminate the loud pop when you turn off a guitar amp.

I am finding really hard to believe I am the only one who has asked this question and there are no definitive answers for it, other than turn the standby switch off first?? I am not willing to settle for a solution that is not really a solution unless it just can't be done. If that is the case please state that.

Just because other amps are like that is not good enough.
Last edited:
Commercial MOV's are typically put across the L1 to neutral to eliminate spikes, so they have ratings like 300 VAC or 540 VAC. This means they don't begin to break down until 1.4*300=520 VDC or 1.4*540=700? vdc. I put them on my hifi amps between hot and neutral after the fuse, because there is a lot of lightning here, and the transformer wire is probably rated 600 VDC. I was lucky when the lightning hit my PAS2 preamp, probably the size of the lightning surge was damped out by the furnace motor or refrigerator motor, so it didn't burn up the power transformer. I salvage MOV's from dead variable frequency motor drives from my former factory job. They are a switching power supply, and they have them because the DC electrolytic caps the energy is stored in are only rated for 400 VDC. Same with the PCAT power supplies, the input L1 goes to a rectifier then to a 250 v rated electrolytic capacitors, so they have an MOV near the input, after the fuse.
Your guitar amp probably has a gain of 5 to 20 anyway, so 50 mv that gets to the input stage will cause your speaker to pop. My dynaco hifi had a gain of 50 in the preamp and another 10 in the power amp. You want a capacitor because they have resistance that gets lower and lower as the frequency goes up, and pop is a high frequency sound.
The capacitor acts like a short for high frequencies, the resistance is 1/sqrt((2*pi*frequency*C)^2) so the higher the frequency the lower the impedance (resistance). I put my .01 @ 1000 v right across the power switch in my dynaco PAS2. Some other people put a .047 uf or something across the input winding of a power transformer, to do sort of the same thing. You might try this with a non-x2 or non-y cap, just to see what value you want to buy in an x2. No pop, that is the right value. Then buy an X2 that value, or a Y if that is more correct. After reading wikipedia, I like the idea of a flameproof coating on the cap (in the UL rated ones). You may be in less trouble by getting the x versus y wrong, although lawyers can make a million off of any mistake.
That 1 uf panasonic cap you linked to, that is pretty extremely large to have powered up all the time across the power switch. Might be okay, but I've never seen one like that. I got the idea for the .01 from old television schematics, from the fifties, that had a hard copper power switch instead of electronic like these days.
My comment about the time was about DF96, he could be just about waking up now if he sleeps 12AM-6am like a normal person.
Last edited: