What's a good rating for a MOV suppressor?

[leadfootdriver]

Not only Boogie has them. I don't know how often they explode but since it was mentioned I would assume they were not talking about a couple of cases only.
The other thing about NTC surge thermistors was they work as specified only if your device has a constant current draw which is not the case with a guitar amp so at this point I'm not really convinced I should use both.

I was reading the TVS file jazz p bass posted, and it did say that they could explode but only if the max spec is exceeded. That would be 4500V for what I bought. But I never heard of it until you mentioned it.

I'm not interested in putting them on my OT secondaries tho. I haven't seem a schematic that uses that.

[R.G.]

I read somewhere that MOVs are known to explode and set equipment on fire which doesn't sound like something that I need in my equipment...

I find that when I start writing "I read somewhere..." that I need to go check that out.

MOVs are known to explode and set equipment afire; this is true. It's also true of capacitors, resistors, diodes, and wires (see "exploding wire" on google; they're used as blasting caps to set off explosives). All these things can be made to explode and start fires in exactly the same way as a MOV: you put in massively more electrical energy than it can take, and it heats up to the temperature where the materials inside melt and vaporize. You can explode nearly anything that can conduct this way.

The reason MOVs are known to explode and/or catch fire is because they're connected to electricity. They won't go off sitting on the shelf. You can pretty much guarantee your equipment won't explode or catch fire by disconnecting the AC power cord. But that's not all that useful, either.

Like everything used for electricity, applying MOVs requires understanding what they do and thinking about what failures would do. MOVs get connected to high-power electrical supplies as part of their normal job, so it makes sense that if you get it wrong, lots of electricity can get funneled into them, with various bad results. On the other hand, they are intended to absorb damaging transients which could cause other damage. The trick is to use them to protect against transients without adding problems. This is possible; otherwise, MOVs would not be commercially available. For instance, those "power line conditioners" you buy in power strips? Several MOVs (or tranzorbs, similar thing) in each one.

The MOV across the Output transformer half-primary is effectively a shunt capacitor - so the MOV size should be kept as small as possible to reduce the capacitance (eg. 7mm disk size) - and preferably a resistor is placed in series with each MOV to make an RC zobel network (eg. resistance similar to OT primary loading).

I put MOVs across the primary of the Workhorse amps for transient protection. They do have some capacitance. But I found that I could not detect any change in the frequency response of the amp in the audio band with them there or not. For amps with overall negative feedback, this may or may not be an issue, but my conclusion was that they were something that could be ignored as far as the sound of the amp goes. They're a lot more reliable than the string of high voltage diodes on each side of the primary.

The other thing about NTC surge thermistors was they work as specified only if your device has a constant current draw which is not the case with a guitar amp so at this point I'm not really convinced I should use both.

They do different things. MOVs and similar devices do essentially nothing unless there's a voltage transient, which they eat, converting the transient energy into heat.

NTC surge thermistors are specified for much more than constant current use. They are a resistor that decreases in resistance as it gets hot. As long as the amp pulls X current, they insert a negligible amount of resistance in series with the power line, because there is enough current to get their resistance low enough to be negligible. When the amp is cold, they have a bigger resistance. The trick in designing with NTC thermistors is to learn from the datasheet what their resistance-versus-temperature and resistance-versus-current behavior is, and use the fact that when the amp is off (and presumably the NTC is cold) the NTC will have a high resistance for a few AC power cycles before heating up and becoming negligible in the power circuit. You also have to provide for the fact that the NTC stays hot as long as the amp is on.

They do different things. You may or may not need both, depending on what abuses you want your amp to be immune to.

You mentioned capacitance in the OT mod. But what about just the AC lines? Does the capacitance of the MOV have a snubbing cap effect for the whole amp?

A MOV on the incoming AC power line affects only the AC power in. There are some quite large caps (i.e. the power filters) between the AC power line and the audio circuits that completely swamp any capcitance of MOVs on the power line.

And as I mentioned, I found no change in the audio response from MOV capacitance on an OT. It depends on the capacitance and the impedances of the circuit it's attached to. If the self-capacitance of the OT primary or the wiring/tubes/secondary reflected load/etc connected to it is large compared to the actual MOV capacitance, the MOV does not materially affect the audio response.

Can one horse run faster than another one? Absolutely. But WHICH ONE? Details matter.

[leadfootdriver]

Great post! Thanks for that.

[GainFreak]

I find that when I start writing "I read somewhere..." that I need to go check that out.

It looks like it was the Wiki. I know it's not reliable but it's a hint to investigate further:

Varistor - Wikipedia, the free encyclopedia

It turns out that thermally protected MOVs (TMOV and iTMOV) are available that will take care of the "exploding" problem:

US - Electronic Components Distributor | Newark.com

I'm aware that MOVs and NTCs will handle different issues. This looks like a good article about inrush current limiting:

Inrush Current

[trobbins]

I'm not interested in putting them on my OT secondaries tho. I haven't seem a schematic that uses that.

Your incentive may depend on whether you can get a replacement OT, how expensive a replacement is, and whether you've 'lost' an OT in an amp before and suffered the pain. This type of OT failure is not prevalent, but it is 'common', and amp designers/restorers are including OT protection, with the main OEM form being a pre-load resistor on the speaker side, as that is the simplest and lowest cost option to statistically reducing failure rates. Some older amps also used just C, or RC networks, across the half-windings, or from anode-to-anode, both to tailor high-frequency response and to provide OT protection.

I know of one PA amp manufacturer using plate-to-ground valve diodes for such protection back in the early 60's - it's not a simple addition to squeeze in two 6AL3 under the chassis! Peavey alert their owners to the risk as well (• View topic - Just a question....).

Hi-fi amps would notice the added capacitance of a MOV, and inadvertantly using a 'large' MOV in say a 10k PP is likely to start pulling in the treble end of even a guitar amp. For hi-fi'ers the aim would be to use the MOV-R to dampen say the main transformer self-resonance, but be negligible in the audio range.

Ciao, Tim

[leadfootdriver]

I suppose I'd be biased to do it if I had a meltdown!

It would be easy to do since I have a perfboard style turret board coming. I bought this style of board so I could have freedom of construction.

The schematic I'm building from specs a B+1 of around 395V. We'll see what it ends up as when I get it running. But assuming that's what I'll end up with, should I spec a MOV the same way as the AC line; giving myself a safety margin for degradation?

I found 429VDC MOV that clamps at 650V.

TMOV14RP250E Littelfuse Varistors

The next MOV in the radial style is a 473VDC that clamps at 710V.

TMOV20RP275E Littelfuse Varistors

[trobbins]

For a PP OT half-winding MOV, the 1mA DC voltage, at minimum tolerance end needs to be safely above the high tolerance end of your B+ (eg. mains high, no load) - so you should be looking for probably a 320E model or higher for starters (eg. 459VDC 1mA rating). Those parts are 14mm/20mm disks - they have pretty high capacitance - I'd suggest trying to stay within 100-200pF. The joule rating for this application is way over the top no matter what size common MOV you choose.

I think the 1mA VDC rating is the best design guide to use. The clamping voltage is not particularly useful, as it is based on a particular test circuit and is likely to be fine for the transformer as the main aim is to keep the primary winding from stressing voltage insulation levels.

Smaller diameter MOVs may not be stocked in the voltage rating you want, so one workaround is to use two or three MOVs in series. It's best to place the MOVs or MOV-R parts as electrically close to the relevant primary half-winding as possible - as that is the root cause of any damaging transient.

[leadfootdriver]

Here's a 505vdc. It's the 14mm/20mm you speak of. It has a capacitance of only 40pf.

VDRS05C385BSE Vishay/BC Components Varistors

[trobbins]

That part number is a 7mm diameter part - which should be fine - and yes it has low/negligible capacitance. The 1mA 620VDC rating is +/- 10%, so would be fine for amps with B+ max up to about 500-540VDC, which may be equivalent to say a nominal 450VDC B+ at idle.

I have a batch of 7mm 330VDCmin movs, which have come in handy for a wide range of amp restorations, including up to 3 in series for 800V B+ amps.

[leadfootdriver]

Right! 7mm... I was trying to make sense of the data sheet, and somehow thought it was the larger ones.

Thanks!

[R.G.]

should I spec a MOV the same way as the AC line; giving myself a safety margin for degradation?

Yes. Even more, give yourself some margin for AC power line variation.

What you're trying to protect against is a transient spike puncturing the insulation inside the OT. These things are caused by things like sudden interruptions of current flow in the inductance of a speaker load (somebody yanked the cable out), a power tube trying to instantly stop the current flowing in the leakage inductance on the primary of the OT, and perhaps even a (VERY!) stout electrostatic discharge on the secondary wiring. All of these have in common a sudden voltage rise between one of the OT windings and another winding or ground.

We're not too interesting in putting in just enough voltage resistance. The layer insulation in most OTs is probably good for 2kV. I say "probably" because no OT maker is going to either test it or tell you the results if they did test it.

On the primary side, the "on" power tube pulls down on its half-primary, forcing it to within 50-100V of ground if it's really cooking, or honestly to ground if the tube shorts. Since the primary of the OT is an autotransformer, pulling one side down causes the other side to go up above B+ as much as the active side went down. So each half of the primary has to be insulated to at least twice B+ just to work normally. And B+ varies. It's reasonably constant at any given operating condition, but the AC line varies +/- 10% at least, so "normal" B+ may be 10% high all the time. And it could rise even more in fault conditions where a tube opens up.

All that's a way of saying that I think a transformer's internal insulation is good for much more than 110% of two times B+. You want MOVs to protect the insulation, so the voltage on the MOV breakover needs to be bigger than the 2.2*B+ and smaller than the breakover voltage of the insulation. So far, so good; but we have no idea what the insulation breakover really is.

So add in some padding. Make the MOV voltage be ... um... three times B+. 3.241 times B+. Pick a number. But don't pick it too low.

The window of acceptable MOV breakovers is easy to hit in this application because OT insulation is usually very good. You just want something to keep V = L* di/dt from going to infinity, punching an insulation hole along the way. T.R.'s right - the energy rating of any MOV you're going to put in there will be massively more than needed to clamp the transient energy. And that series connecting lower voltage MOVs is a good way. I had to use two in series on my amp line.

[trobbins]

R.G. - your 2.2 and 3.3 ratios are for full primary winding protection :-) Half-winding MOV configuration is half.

[R.G.]

Yep. One could certainly do this same kind of calculation to get half-primary values, then use two of them. It's probably not a good idea to only MOV-protect one half. I was mostly doing the calcs to show what things to consider. In all cases, you want a MOV value that's comfortably above the voltages that happen in normal operation, but below the level at which damage happens. This is only tricky when those two voltage levels are close together.

[Enzo]

When I was a kid, and we are talking 50 some years ago, I have a scope tube given to me, and i was trying to build a scope around it. I needed something like 1500v for the third anode - or whatever number the side of the tube was in a 5UP1. Having enough knowledge to be dangerous to myself, I figures, gee, if I send 120VAC into the 12v winding of a 120/12v transformer, I ought to get about 1200v out the primary. I could then rectify that for my tube. By golly it worked. The little transformer lasted even a couple months befor it self-destructed. Litle arcs showing through the fish paper.


But I got a trace on my screen and some deflection. In retrospect, pretty impressive the abuse that little transformer took before it croaked.

[trobbins]

I recall powering up a 5UP1 in high school physics - ahh they don't make 'green' like that anymore!