Let's talk MOV's

[Randall]

I want to offer adding an MOV to vintage amps (as well as others) as a service. Seems like a quick and easy precaution that could protect older amps from flaky voltage spikes at gigs.... to a point. I think I understand that it won't be much good with lightning or other catastrophic failures, but is good for short spikes and other over voltages caused by kitchen appliances cycling on/off, etc.

I'm a little behind on the specs part of this, and I don't like a false sense of security. Say I'm in 120vac land. As I see it, I want a 130 - 135 vac rated MOV, with the highest joule rating giving the longest life due to ability to absorb and conduct with the least amount of degradation. Following this line of thinking, would 140 joules be good, bad or indifferent?

And I see a Panasonic ZNR product with the above ratings, but it seems to be a ceramic , not metal oxide. Does this matter?

[Chuck H]

And I see a Panasonic ZNR product with the above ratings, but it seems to be a ceramic , not metal oxide.

Well... If it's not metal oxide it can't possibly be a MOV then can it What does ZNR mean? Zener?

Sorry I can't say about the "joules" but I do know this circuit is standard on some amps. Most notably some Trainwreck's. I think it's also included in our resident member R.G. Keen's "immortal amp" mods and circuits.

[bob p]

There's a really long thread with a lot of theory, if you look for it. I asked some detailed questions about specs for this type of thing, and got some great responses.

[Jazz P Bass]

The thing about MOV's that I do not like is they can cause a fire when they get hit hard (ie: lightening).
And they degrade as they get used.
There is a spec that states the usefull life is when they are 10% below the clamp rating.
How do you test that? (without further degradation)
Paralleling them & utilizing a properly sized inline fuse helps.

I think you would be 'offering' trouble.

[Steve Conner]

Nowadays TVS diodes are a tempting alternative to MOVs:
1.5KE400CA TVS Diodes - Littelfuse

They are basically zeners with a very high pulse rating.

[bob p]

^ nice.

[salvarsan]

As the MOV energy rating goes up, so does its leakage current. They have a total energy lifetime that limits how long you can use them before they're junk.
...but you can stack the deck for their longevity by using several spike suppression methods in parallel.

An MOV rated at ~170 volts, an RC dV/dT snubber in across it, maybe a common mode inductor before them is commonplace.
Have a look at switching power supply inputs for some ideas.
Add in a gas discharge tube for the 500V+ spikes if you think its merited.

[trobbins]

I want to offer adding an MOV to vintage amps (as well as others) as a service. Seems like a quick and easy precaution that could protect older amps from flaky voltage spikes at gigs.... to a point. I think I understand that it won't be much good with lightning or other catastrophic failures, but is good for short spikes and other over voltages caused by kitchen appliances cycling on/off, etc.

I suggest your basic premise for using a MOV is poor, as incoming transients from the mains that would cause part failure in a valve amp and can be alleviated by a MOV would be extremely rare. If the equipment was going to be used in a bad mains environment which was of concern, then that really needs to be addressed in other ways than providing a nice path for surge currents to come in to the amp.

Imho MOVs have a place within the amp to protect the amp from it's own devils. For example, a MOV directly across the primary winding of the PT can suppress winding induced overvoltage when the power switch is turned off, and hence help protect primary winding parts such as the mains power switch. MOVs are normally covered by a sleeve of heatshrink to constrain the mess that can occur if they fail short and power is still applied (a technique used for PTC and sometimes cap parts). For that type of application, the MOV Joule rating would normally far exceed spike energy expected, so the MOV would have a service life in excess of some other parts in the amp.

[R.G.]

I suggest your basic premise for using a MOV is poor, as incoming transients from the mains that would cause part failure in a valve amp and can be alleviated by a MOV would be extremely rare.

Hmmm. Not sure about voltage spikes from the mains being rare. There is a reason the safety standards in most countries have ratcheted up from 1500V to 2000V to 4200volts.

If the equipment was going to be used in a bad mains environment which was of concern, then that really needs to be addressed in other ways than providing a nice path for surge currents to come in to the amp.

Hmmm. Not sure about that. It doesn't necessarily follow that putting a voltage clamp on the mains input will entice surge currents in. Surge currents happen when something lets them in. The mains setup won't let surge currents in until pushed by high voltage spikes. If the spikes break over a MOV or a TVS, or some other surge protector instead of the primary insulation of your very vintage, very expensive amplifier with its very expensive, irreplaceable power transformer, the actual surgi-ness of the current doesn't much matter. It eats the energy (voltage*current*time) that would otherwise puncture the insulation. Having a transient clamping device is not really like leaving the gate open.

Imho MOVs have a place within the amp to protect the amp from it's own devils.

IMHO MOVs and their ilk have a place within the amp to protect the amp from its own devils AND to protect it against primary side spikes. Clamping a voltage transient is a good thing, no matter which side of the wire to the MOV it comes from.

For example, a MOV directly across the primary winding of the PT can suppress winding induced overvoltage when the power switch is turned off, and hence help protect primary winding parts such as the mains power switch.

Yes, they do a great job of that. Also, similar spikes coming in from the AC power line are clamped and the therefore help protect primary windings, switches, etc. from those too. The difference is pretty much only in the available energy in the spike. PTs have limited energy they can discharge. The power mains have - well, a much bigger ability to make macho, burly spikes.

MOVs are normally covered by a sleeve of heatshrink to constrain the mess that can occur if they fail short and power is still applied (a technique used for PTC and sometimes cap parts).

MOVs will always fail either short or exploded. Exploded happens when they get so much energy that they vaporize. Otherwise, they fail short. The stock thing to do is to put a fuse between the MOV and the source of a possible continuous source of power, like the AC power line. Steve will update me on what TVS devices do, but MOVs lower their breakover voltage a tiny bit with each spike they eat. When they eat enough that the power line peak voltage starts tripping them over, the current skyrockets, and they fail shorted. If there is a fuse between them and the power line, they do the right thing and blow the fuse - again protecting the stuff after them, albeit unnecessarily in this case. And they may make a mess in dying. Lots of things do that. TVS devices may not have the driftdown, and may only explode or short without gradually drifting down.

[trobbins]

I'm happy with the use of MOV's for main transient suppression, but I thought it was a poor presumption that valve amp damage was a consequence (ie. that would be a rare occurrence), and a poor presumption given the rarity of an event where adding a MOV would be a good protective element to add to mitigate such damage - especially as the discussion had already indicated that robust transients such as from nearby lightning would likely damage a valve amp with or without an added MOV.

It is always best to suppress transients at their source. If someone's home is generating such transients as to damage valve amp equipment then going around adding MOV's to all household equipment is a poor remedial action by itself. Commercial environments are another matter, but again if there is a variety of electronic equipment being brought in then the best location for suppression is as close to the feed point as possible, and a commercial filter can be used. Without any other voltage clamping part in a building, or significant impedance, a sole MOV will be the first part to conduct current until insulation in another part somewhere else in the building breaks down - which is a concern to be noted and appreciated (similar to when a harmonic filter is added to address a local problem, but may end up pulling current in from elsewhere).

I agree with the failing short, and can add that if the resistive path through the failed MOV is somewhere between explosion and blowing an upstream fuse, then internal arcing and fuming of the internals over a second or two is sometimes observed.

[Alf]

Nowadays TVS diodes are a tempting alternative to MOVs:
1.5KE400CA TVS Diodes - Littelfuse

They are basically zeners with a very high pulse rating.

Have you used them yourself Steve ?

[km6xz]

I have used them, and they are not rare at all, 1000s of times faster, cheaper with higher energy dissipation than MOVs with much lower leakage. But they are not the universal answer. Their sustained clamp voltage is not too much higher than breakdown voltage for most models. So putting them across 125AC mains that have a potential of spikes of 800 volts will not conduct the rated current before failing. The most common TVS are back to back zeners for full AC cycle conduction of singles for single direction. Sure you can get them where they can withstand very high voltage but not with low enough practical breakdown voltages, say 20-30% higher than the mains operating voltage. Even the small SMD can be rated to 1.5kw.
Most of them in user gear are surface mount protecting low voltage systems so you probably have dozens of them in your home now.
They are cheap, small and reliable so why not use them but if there is potential of really high level spikes, they should be used with another type, or just two in parallel, one being for moderate spikes, selected for breakdown voltages just a little above the mains voltage, and another for protection for multiple cycles of really high level spikes with an initial breakdown voltage of 400-500 volts that can handle clamping voltages over 1 kv.

I run into them in guitar amps here. Yerasov puts them across ground and the 16 ohm OPT secondary and I have run into several that behaved badly, like breaking down at much lower voltage such as the last one, a 1.5kw device, that was breaking down at 9volts RMS, presenting a perfectly normal output until 9 volts was exceeded so the TVS presented a very low resistance short. Another one was found a few months ago that was just shorted. The owner assumed his buddy was right in predicting the transformer was bad so he installed a new transformer which had the same symptoms. It took just seeing the signal clamp precisely at 9 volts to identify what was going on and that the $0.08 diode was at fault, not the $100 transformer.

[Chuck H]

The owner assumed his buddy was right in predicting the transformer was bad so he installed a new transformer which had the same symptoms. It took just seeing the signal clamp precisely at 9 volts to identify what was going on and that the $0.08 diode was at fault, not the $100 transformer.

That's not as gross an error for a novice as we see here sometimes. And at least it was a well meaning friend rather than a repair tech. But didn't you know Stan? It's always the output transformer! Amp broken + novice tech = output transformer.

[km6xz]

Only problem was the "friend" kept the old transformer.

I remember the most common suggestion the girls at the counter would get from guys when bringing in their gear: "I could do it but I am too busy, it is probably just a capacitor". If the guy was too full of himself, they would smile and ask "thank you, that will be helpful, which capacitor?". I wanted girls on the counter because guys are always better behaved and more likely to cooperate, and never get angry at attractive girls. It makes them feel better by thinking they are impressing the girl, who is "just a girl" without realizing they know 50 times more about it and how to get the customer to recreate the symptoms on a small test counter. Installing that separate test counter was the most valuable asset in the shop, it saved thousands of hours of chasing symptoms never found because the specific combinations of actions was not told to us. Some guy coming in with a 32x8 mixer saying "just check it all" instead of "when sending a signal on Aux3 at +6, from channel 9 or 10, it distorts when sent to headphone 2 but not lower levels". The girls would make sure the symptoms were reproduced and watched taking notes as the customer recreated them. The tech had good odds of having actionable information for a definite fault resolution the first time it was looked at.

[Chuck H]

That's clever thinking. Straight up.