How to protect the OT?

[daz]

Anyone who read my other threads may have noticed the issue i had turned out to be a shorted OT. Since everything else in the amp look fine and i was unable to find any reason this happened, i have to assume it was on the verge of shorting during the many months i spend trying everything in the book with what was my first build aside from a little 18 watter. So i may have overheated it during that time, i just don't know.

But what i DO know is since i have to buy another OT at around 100 bones with shipping, i want to protect it in case there is something i missed that still exists in the amp. I recall reading ways of protecting the OT at a forum somewhere, but i can't find it now. Can anyone give me advice on a good way to do this? I seem to recall a way with diodes.

[Steve Conner]

There are two ways to kill an OT: overcurrent from runaway tubes, and overvoltage from a mismatched or missing speaker load. And three ways (that I know of) to protect one:

1) A B+ fuse, hopefully blows before a runaway tube toasts the OT primary, but this is by no means guaranteed.

2) "Flyback diodes", high voltage diodes connected across each power tube. These allow the OT's magnetizing energy to return to the B+ filter caps in an emergency, such as forgetting to connect a speaker. The diode anode goes to ground, and the cathode to the power tube's plate. The diode rating must be at least twice the B+ voltage, hence you often see a few 1N4007s in series here. But don't overdo it, as the diodes can also "protect" your OT by failing short from overvoltage spikes and blowing your B+ fuse.

3) A resistor of a few hundred ohms across the OT secondary. This provides some sort of load if the speaker gets disconnected.

[txstrat]

As far as I know the wiring of most of the Fender amps is "sort of" a (little) protection. The OTs secondaries are shorted while no speaker is connected.
The OT can be damaged way more easily when the secondaries are "open" rather than when they are shorted.
Although the best protection is a connected speaker or sufficient dummy load.

[daz]

Ok, which would be the best way to protect it from runaway tubes? i'm not worried about speaker being disconnected, and if i were i could always use a switching jack for the speaker with a 30 watt or so resistor. But things i have less control over like a runaway tube. also what about a scree resistor going open or having a bacd solder joint? Will that do it?

[pdf64]

I like quick blow fuses in series with each power tube cathode, F250 or F330mA. Add a 100k resistor cathode to ground to stop it floating too high if the fuse blows
Slow blow fuse on the B+ transformer winding, T500mA.
Flyback diodes.
Shorting speaker socket.
Heater winding ground referenced via 100R 1/4 watt resistor/s. Peter.

[Steve Conner]

I always think shorting speaker jacks are pretty useless. You can just as easily forget to plug the other end of the speaker cord into the cabinet.

I investigated the missing speaker thing many years ago. I found that if an amp has NFB, and is stable with no speaker load, then it can stand being turned on with no speaker load, and even played. As long as the power amp isn't overdriven, the NFB acts to keep the speaker voltage down to a safe level, and hence the OT primary voltage too.

Once the power amp is overdriven, the NFB loses its grip and dangerous voltage spikes start to appear, but the measures I outlined in 2 and 3 above help to mitigate them. However, the mismatched load also causes the screen grids to overheat and melt when you crank it, and none of the measures helps with that. If the molten pieces of screen cause an internal short in the tube, this could finally blow the OT from overcurrent.

Also the first two "ifs" are by no means true for all guitar amps: some like the AC30 don't have feedback, and some may be unstable into an open circuit.

My investigations were done with a hi-fi grade OT, and a guitar one may be easier to damage.

If a screen resistor goes open, it will just turn that tube off, which shouldn't damage the OT. Unless the resistor were arcing internally and spiking the tube on and off, I suppose.

[Enzo]

And a shorting jack does nothing once a speaker cord is plugged into it. In intermittant cord is just as dangerous as no load at all.

[daz]

I'm gonna go with the flyback diodes. Two 1N4007's on each side of the PT is what i used for the rectifier, so i assume theres no reason to go with more than two. i think one is 1000v, no? Anyways, will that be pretty good protection for the OT in any given sitch? Thanks all !

[scole]

so if you think you killed your ot, how do you check it to be sure?

[Steve Conner]

http://www.geofex.com/ampdbug/outtrans.htm

There's also a trick you can do with a 9v battery and a neon lamp, a kind of ghetto version of the ringdown method used to test TV flybacks, though I'm not sure of the exact details. I think you connect the neon across the primary, and if the OT is good, tapping the secondary wires on the battery terminals should make the lamp flash. If it has shorted turns, they'll damp the kickback and the lamp won't light.

Could be wrong, though, maybe someone else has an opinion on this.

[martin manning]

Re the neon bulb test, I've seen this various places, and RG Keen wrote about it in Premier Guitar a while ago.

Basically the test is this: A NE-2 bulb in series with a 100K resistor is clipped across one winding, with ALL OTHER WINDINGS OPEN CIRCUITED.

Then, a battery is connected momentarily across the same or another winding. As the battery connection is broken the collapsing field produces an EMF that causes the bulb to flash, but only if no winding is shorted. So, a flash means a good transformer.

MPM

[pdf64]

I meant to do that test with the damaged OT I had. But couldn't find a neon, got impatient and cut the windings open. The overheated area was easy to spot.
The fault started out in mine as intermittant, it was fine at domestic volumes but would start to cut out when it got cranked in a band situation, so it needed enough signal voltage to break through the damaged insulation. Eventually it stayed in the cut out state.
But I'm certain that what started it off and damaged the insulation in the first place was that tube red plating. When that happened, the slow blow fuse on the B+ went eventually, but by then then damage must have been done. The only way I can see to prevent the high current from a red plating tube causing OT damage is to specifically target it with a quick blow fuse in the cathode circuit.
The sure test for a bad OT is to sub in a known good one.
I put at least 2kV worth of diode rating for flyback protection, ie 2 x 1N4007 in series, as the last thing you want is for the flyback voltage to kill the diodes and cause them to short out.
Better not to put them in at all than to put too low of a voltage rating one in there.
2kV as that's what I've seen the OT insulation is rated at/tested to.
The OT can much higher voltages on the primary than the B+ puts out. Think of how the flyback works to create enough voltage (from a battery) to light a neon in the geofex test above. Peter.

[MWJB]

Steve wrote: "I investigated the missing speaker thing many years ago. I found that if an amp has NFB, and is stable with no speaker load, then it can stand being turned on with no speaker load, and even played." - I'd be loathe to test this out with my own amp, I've had amps with NFB and no shorting jack blow the tubes immediately when a speaker has opened up.

Conversely, I've witnessed folks power up & try and play through Fenders (with a shorting jack), before realising no cab was connected and no problems have followed (though I'm not recommending that you deliberately try it). As Enzo says the shorting jack can't protect you in the event of a broken speaker lead however.

A good way to protect the OT in normal use (which is the scenario Daz tells when the old one blew), is to use a heavier duty part.

Three 1N4007 in series are usually recommended for the flyback diodes.

[daz]

three eh? Done deal, thanks. I have one question tho. long ago i modded my classic 30 to cathode bias using a mod file i found at blue guitar. There were 2 of them and i recall Enzo had written one of them. Anyways, the mod showed to remove the flyback diodes the classic 30 has. My question is, why was this done? My current amp is cathode biased too, so is there a reason i shouldn't use flyback diodes with it?

[Hardtailed]

So you guys recommend going with the safety resistor accross the OT secondary instead of the shorting jacks?

Me I was all happy that I had devised a way to have 2 speaker jacks with different impedance yet have both short to ground unless one of them was connected

I'm unsure which way to go now. I've often heard that it's better to have too low of an impedance than too high, the tubes will simply refuse to put out this much current.
I've personnally never had a bad speaker lead, I've also never turned on an amp without a load connected, but I guess it's cheap protection so why not?

As I can see, to not be wasting too much heat and power while in normal operation, you need a resistance high enough that it's almost an open circuit.
Say a 200ohms resistor on the 8 ohms tap, the load reflected to the primaries goes from 3.2K to 80K!!!

I hadn't thought about the flyback diodes though... thanks for the suggestion. I may go and buy some extra diodes then.
Still unsure how and why it works exactly.