What happen if you use the wrong impedance output tap of the OT?

[Alan0354]

What happen if you use the wrong impedance tap of the output transformer in the tube amp? It should lower the power due to mismatch, but what else? Wound it change the sound? I don't believe it will hurt anything.

[Austin]

It changes the sound, try and see which you like better. 4 ohm speaker on 8 ohm tap makes it have less volume but more pleasing kind of distortion to my ears. 8 ohm speaker on 4 ohm tap has better bass response and a more punchy sound but also more brittle and not as easy to listen to.. That's just me though.

[Chuck H]

Indeed, the tonal qualities of a mismatch is a subjective thing. Exactly what changes are happening sort of becomes moot in the face of other parameters. Since different transformers have different properties and different speakers have dofferent impedance curves. Even general opinions vary. Some people even tout the tonal differences of parallel vs series speaker wiring.

On another subject...

I don't believe it will hurt anything.

Why do you think this?

It depends on how much mismatch you have. WRT tubes, a dead short is very hard on them. possibly causing an overdissapation condition that could indeed cause a failure. The wrong tap on an OT isn't a dead short. But if, for example, you have an amp set up to play into a 16 ohm load and you plug into a 4 ohm load, you are moving closer to a dead short and incurring some of the detrimental effects on the tubes. An open load can be death to the OT. If, for example, you have an amp set up for a 4 ohm load and you plug into a 16 ohm cabinet you are getting closer to that open load and incurring some of the detrimental effects. It's generally acceptible to mismatch by 100%. That is, a 4 ohm load or a 16 ohm load on the 8 ohm tap. A 2 ohm load or an 8 ohm load on the 4 ohm tap, etc. Going outside the proper load is at your own risk. Going outside the 100% mismatch has proven to cause failures many times.

[Alan0354]

Indeed, the tonal qualities of a mismatch is a subjective thing. Exactly what changes are happening sort of becomes moot in the face of other parameters. Since different transformers have different properties and different speakers have dofferent impedance curves. Even general opinions vary. Some people even tout the tonal differences of parallel vs series speaker wiring.

On another subject...



Why do you think this?

It depends on how much mismatch you have. WRT tubes, a dead short is very hard on them. possibly causing an overdissapation condition that could indeed cause a failure. The wrong tap on an OT isn't a dead short. But if, for example, you have an amp set up to play into a 16 ohm load and you plug into a 4 ohm load, you are moving closer to a dead short and incurring some of the detrimental effects on the tubes. An open load can be death to the OT. If, for example, you have an amp set up for a 4 ohm load and you plug into a 16 ohm cabinet you are getting closer to that open load and incurring some of the detrimental effects. It's generally acceptible to mismatch by 100%. That is, a 4 ohm load or a 16 ohm load on the 8 ohm tap. A 2 ohm load or an 8 ohm load on the 4 ohm tap, etc. Going outside the proper load is at your own risk. Going outside the 100% mismatch has proven to cause failures many times.

I am not referring to open of short condition, bad things happen with that. But say if you put a 4 ohm speaker to 16 ohm tap, you reflect much lower impedance to the plate of the power tube, so the voltage swing is smaller with a given current. Do you mean when the load impedance is small, the plate cannot swing low enough, therefore more voltage develop across the tube with the given current, so more power dissipation?

But if you connect a 16 ohm speaker to a 4 ohm tap, then the impedance presented to the plate of the tube is higher than design, the tube swing in to saturation easier, I don't see why it would be detrimental. Am I missing anything?

[Chuck H]

A load that is too low will increase current. Sometimes increased current is bad. The amp is working very hard at a much lower efficiency. Checking current under operating will probably reveal the tube overdissapating even if things look kosher at idle. A load that is too high will fail to damp voltage spikes that can punch through the insulation on the OT winding. And it really doesn't take as gross a mismatch as you may think. A guitar speaker may only be at it's rated impedance in a very narrow frequency band and much higher everywhere else. Even over 100 ohms at frequencies generated by guitar amps. Consider that voltage spikes in a tube amp running into a proper load can be over twice the Vp. Now what happens when you increase the impedance 2X and then 2X again! With much lower damping it's easy to reach voltages that can damage an OT. This wouldn't happen with, for example, a 4 ohm tap on a 20ohm resistive load. There is still enough damping at all frequencies. But if you increase, say, a 100 ohm impedance at 4k to a 200 ohm and then a 400 ohms... Well now your deadly close to a relatively open load at signals 4k and above with lightning flashing around inside your OT.

EDIT: Incidental, but a load that's too low is probably not going to hurt the amp. The tubes will just wear out a bit faster. The tubes do self limit on how much current they can provide. But if a low load has the tube at max diss all the time that bill has to be paid eventually. In a properly adjusted and loaded amp the power tubes get enough of a break to rest between demands. Take away the breaks and they wear out out faster. Just ask any blue collar worker about this principal.

[Alan0354]

Thanks for the reply. It is the voltage spike when the tube turn off, the primary of the transformer flyback to +ve voltage that can potentially damage the transformer if it fly too high.

Let me confirm this. Say if the tube idle at 50mA. Say if in perfect match, the primary impedance is 4K, so when the tube turn off and the plate current goes from 50mA to 0, the flyback voltage will be 50mAX4000=+200V. Say B+ is 400V, then the terminal of the primary flies to +600V. But say if I have a big mismatch of 4 times, then the impedance become 16K. So when flyback, voltage will be 50mAX16000=+800V. So if B+=+400V, the totally swing will be +1200V. This is when the damage happen. Am I correct?

thanks

Alan

[Chuck H]

I'm vague on the exact principals, but experienced in the phenomenon. Your interpretation seems similar enough to other theories I've read. IMHE the actual flyback voltage spikes can go well over even what is predicted on paper (in print, whatever). I discovered this when fine tuning a zobel filter on an OT primary wind. The actual parts spec needed to be higher than the predicted spec in order to prevent repeated failure. If such is the case then voltages capable of jumping winding insulation are a legitimate consideration with a high mismatch. As I illuded to, a low load probably won't kill most amps. In fact... There was one thread here a few years ago where a guy was running his BF Bassman head with NO load plugged into the "speaker" jack. On that amp the "speaker" jack is a shorting jack, so, a dead short. He would then plug the "extension" output into another amp. Basically using his cranked, shorted Bassman head as a preamp. The small amount of bleed signal on the "extension" jack was enough for his purposes it seems. He did have worn out power tubes but chose not to change his MO in favor of the tone he was getting. So he replaced the power tubes and just kept running the amp into a dead short... So there you go.

[Austin]

If you connect a 16 ohm speaker to a 4 ohm tap the voltage on the plate of the power tube will fluctuate much more wildly because the transformer primary is an inductor, and inductors try to maintain the same current going through them and can swing up to twice the voltage to do it. So the 600 volts normally on the plate could swing up to 1200 volts and cause internal arcing in the tube or damage to the output transformer primary insulation. Probably unlikely but it is possible.

You would also of course get less power, but more odd order harmonic distortion too as a result. A look at the tube data sheets shows that as the load resistance goes higher, so does the odd order harmonics. To get the nice sweet sound that tubes are famous for, try going the other way and use less resistance than optimal for a more squished waveform and hence more even order harmonics. Like a 4 ohm speaker on a 16 ohm tap. Try it and see. It is technically asking the tube to work harder this way but it's probably okay, tubes are pretty tough...

[Alan0354]

I'm vague on the exact principals, but experienced in the phenomenon. Your interpretation seems similar enough to other theories I've read. IMHE the actual flyback voltage spikes can go well over even what is predicted on paper (in print, whatever). I discovered this when fine tuning a zobel filter on an OT primary wind. The actual parts spec needed to be higher than the predicted spec in order to prevent repeated failure. If such is the case then voltages capable of jumping winding insulation are a legitimate consideration with a high mismatch. As I illuded to, a low load probably won't kill most amps. In fact... There was one thread here a few years ago where a guy was running his BF Bassman head with NO load plugged into the "speaker" jack. On that amp the "speaker" jack is a shorting jack, so, a dead short. He would then plug the "extension" output into another amp. Basically using his cranked, shorted Bassman head as a preamp. The small amount of bleed signal on the "extension" jack was enough for his purposes it seems. He did have worn out power tubes but chose not to change his MO in favor of the tone he was getting. So he replaced the power tubes and just kept running the amp into a dead short... So there you go.

Short out the main jack?!!! That's cold!!!!

Thanks for the knowledge.

Alan

[Enzo]

That is why Fender mounts shorting jacks on the speaker output. A shorted output is far more desirable than an open output. If someone fails to connect a load, the short saves the amp.

[Chuck H]

That is why Fender mounts shorting jacks on the speaker output. A shorted output is far more desirable than an open output. If someone fails to connect a load, the short saves the amp.

I suspect this is why vintage Marshalls have rep for fried OT's. It's an open jack on a head. So, an accidental open load is much more probable.

[Alan0354]

I'm vague on the exact principals, but experienced in the phenomenon. Your interpretation seems similar enough to other theories I've read. IMHE the actual flyback voltage spikes can go well over even what is predicted on paper (in print, whatever). I discovered this when fine tuning a zobel filter on an OT primary wind. The actual parts spec needed to be higher than the predicted spec in order to prevent repeated failure. If such is the case then voltages capable of jumping winding insulation are a legitimate consideration with a high mismatch. As I illuded to, a low load probably won't kill most amps. In fact... There was one thread here a few years ago where a guy was running his BF Bassman head with NO load plugged into the "speaker" jack. On that amp the "speaker" jack is a shorting jack, so, a dead short. He would then plug the "extension" output into another amp. Basically using his cranked, shorted Bassman head as a preamp. The small amount of bleed signal on the "extension" jack was enough for his purposes it seems. He did have worn out power tubes but chose not to change his MO in favor of the tone he was getting. So he replaced the power tubes and just kept running the amp into a dead short... So there you go.

I think I figure out why OT burn when the secondary is open. When the secondary is open, the OT is more like an inductor than a transformer. When the output tube swing positive, the current decrease or even turn off, the inductor fly back. Think of ignition coil in car, you run a current through a coil at low voltage, then you cut off the current abruptly. The coil fly up to over 10KV to create a spark!!!! The better the transformer, the better the inductor it becomes and the higher it flies and faster it burns!!!

But back to a 16 ohm speaker connects to 4 ohm tap. Far as I remember, the voltage increase is only the square root of the resistance. A 4 times jump in load impedance only cause double the voltage. The transformer should be able to handle that. I work years in high voltage circuit of 10KV and over. Insulation is about 400V per mil ( 0.001"). Even if you use more conservative numbers, `150V/mil is quite safe. Unless the transformer is scattered wound that you have one turn from the beginning of the winding over lap the end of the winding, you don't see the full voltage between the adjacent wire. At that, say the coating of the wire is 1mil, you can at least stand 300V safely.( two coating). But if it is layered winding, you don't get much voltage between adjacent wire. The max voltage is between the two layer. But you still don't get full voltage and there is a layer of paper in between. I don't think putting a 16 ohm speaker on 4 ohm tap will likely hurt the OT.

Anyone actually burn an OT in the case of 16 ohm speaker connect to 4 ohm tap?

[Steve Conner]

If you run too high a load impedance (say 16 ohm speaker on a 4 ohm tap), you'll exceed the screen grid dissipation rating of your power tubes. The first big power chord you hit, the screen grids will melt and short the tubes out. Or if you're lucky, the screen resistors will fail open circuit.

As far as I know, the voltage spikes that cause burnt OTs and tube sockets are the result of a really serious mismatch, an open circuit even.

I did some experiments years ago with deliberate mismatches, to try the various protection circuits that amp makers used. The 200 ohm resistor across the speaker jack was good at stopping the spikes, but it couldn't save the screens.

[Chuck H]

Ok... So the general consensus is that a greater than 100% mismatch is best avoided. And valid reasons have been given. But Alan0354 keeps questioning the replies or reinventing the question in an attempt to find the favorable answer he wants. So, as my final statement on the matter, Alan0354, do as you like. It's your amp. It will probably be fine working with impedances from 4 ohm to 16 ohm load or tap provided your amp doesn't already stress the power tube or OT maximum tolerances. Many players do this every day without issue. The ones that do have a problem have places to go where amps get fixed. It all works out.

[Enzo]

SOmething I tell my trainees: Getting away with something doesn't make it a good idea.