Tube Amps

[travelin travis]

why is it bad to power up a tube amp without being connected to a speaker/speaker cab?

 

[boingoman]

why is it bad to power up a tube amp without being connected to a speaker/speaker cab?

VERY SIMPLIFIED EXPLANATION

It causes high voltages in the power tubes and output transformer, which can lead to rapid failure.

A LITTLE LESS SIMPLE

Ohm's law. Amps use voltage, current, and a load (speaker) to make power, and therefore sound.

Voltage=CurrentxResistance. For this discussion, resistance and impedance will be considered to be the same thing.

The tubes need to see a certain impedance to function properly. This impedance is set by the speaker load.

From a purely math point of view, as either current or resistance increases, voltage must also increase.
RxC=V
4x2=8
8x4=32

No load on a tube amp is essentially infinite resistance. Therefore, voltages in the output tranformer and tubes rise very high and can cause arcing across the tube components and transformer windings, destroying them both.

From that, you can see the result of too low a load as well. As resistance drops (like putting a 2ohm speaker on an 8ohm amp) current increases. Go too far, the transformer can't handle the extra current, and fries.

C=V/R
C=32/8=4amps
C=32/4=8amps

The reality is a bit more complex in terms of how V, C, and R interact, as they can affect each other in a tube amp in other ways aside from being static values plugged into a formula, but that is basically the deal.

 

[ggunn]

The reality is a bit more complex in terms of how V, C, and R interact, as they can affect each other in a tube amp in other ways aside from being static values plugged into a formula, but that is basically the deal.

I'm not challenging you on this, but can you point me to the comprehensive explanation, like a web site or something? I am degreed in EE, so I'm pretty sure I can get it. I have read some tube books (they don't teach tubes in EE any more) and rebuilt a couple of amps, and I don't remember any of the texts mentioning or explaining this effect.

Thanks,

 

[boingoman]

I'm not challenging you on this, but can you point me to the comprehensive explanation, like a web site or something? I am degreed in EE, so I'm pretty sure I can get it. I have read some tube books (they don't teach tubes in EE any more) and rebuilt a couple of amps, and I don't remember any of the texts mentioning or explaining this effect.

Thanks,

Maybe that was a little clumsily worded. I don't even know what I meant, though I have an idea swimming around. Maybe you've got some input.

There is a relationship between voltage and current that I kind of left out, as it didn't seem relevant, and could be confusing to the issue at hand.

I guess at the base I mean this: I left out the fact that with a very high load, in addition to a voltage rise, there is a current decrease. With a very low load, in addition to a current rise, there is a voltage decrease. This would seem to balance out in ohm's law, resulting in the same power output into a wide range of loads, but the tube amp has inherent voltage and current limitations. The voltage limit is reached first with too high a load. The current limit is reached first with too low of a load.

So the load, at the extremes, causes too much voltage or allows too much current flow in the amp's attempt to follow ohm's law by trying to put out sufficient current or voltage to generate a given power. It was designed to put so many watts into so many ohms, and adjusts C and V to compensate.
I don't know how that sounds to an EE, but that's how I see it.

Inside a certain range, those laws are followed, but from a different perspective. That's what I'm getting at as well. Within a certain range, voltage is independent of the load in an amp. A lower load means more current, same voltage, more power. A higher load means less current, same voltage, less power.

At the extremes, C and V are determined by R.
Within a certain range, C is determined by R and V.
It's kind of a chicken and the egg, which came first kind of thing. A perspective issue.

I am kind of trying to work it out for myself a bit (maybe I should take some EE classes, I'm a street-educated bum. )

So I though it was more important to stick to the main reasons tube amps fail with incorrect loads: too much current with a low load, and too much voltage with a high load.

 

[ggunn]

Maybe that was a little clumsily worded. I don't even know what I meant, though I have an idea swimming around. Maybe you've got some input.

[...]

So I though it was more important to stick to the main reasons tube amps fail with incorrect loads: too much current with a low load, and too much voltage with a high load.

I'm hip to low resistance/impedance (high load) => high current => component failure, but for a given voltage source, high impedance (low load) => low current. The AC voltage swings inside a tube are small compared to the DC plate voltages, and I don't see why decreasing the load (increasing the resistance) would cause these voltages to become extreme. Your invocation of Ohm where the voltage increases with resistance only holds true if current remains constant, and current rather than voltage is usually determined by resistance.

Anyway, I'd really like to see a concise engineering explanation of this effect. Again, I am not claiming that you are wrong; I'd just like to see the mechanics of how this failure occurrs.

 

[Toki987]

You are two sharpkats. I hope you continue this discussion for the rest of us that don`t know squat about electron wrestling. All I know is that the smell of an overloaded tube amp's transformer stays with you for a lifetime...

 

[boingoman]

I'm hip to low resistance/impedance (high load) => high current => component failure, but for a given voltage source, high impedance (low load) => low current. The AC voltage swings inside a tube are small compared to the DC plate voltages, and I don't see why decreasing the load (increasing the resistance) would cause these voltages to become extreme. Your invocation of Ohm where the voltage increases with resistance only holds true if current remains constant, and current rather than voltage is usually determined by resistance.

Anyway, I'd really like to see a concise engineering explanation of this effect. Again, I am not claiming that you are wrong; I'd just like to see the mechanics of how this failure occurrs.

Thanks for the benefit of the doubt. I will find out for sure, but here are some further ponderings.

What you said about ohm's law is true, but we are talking about a circuit designed to produce power into a specific load.

Making it about power changes a few things. Maybe that's what I meant when I said thing were a little more complex.

With the addition of a specified power, voltage and current have an inverse relationship as the resistance changes.

100W@ 8ohms = 28.3 volts, 3.5 amps
100W@ 16ohms= 40 volts, 2.5 amps
100W@ 1000 ohms= 316 volts, .316 amps

As current goes down due to higher resistance, voltage has to increase to keep the specified power output.

In addition, like you said, the AC voltage swings in the tube are small. But those voltages are determined by the rest of the circuit including the speaker, not the tube itself. The tube has no built-in limits or operating range. It will produce higher voltages or draw more current to make power until it or something else fails. The rest of the circuit, including the speaker, determines the tube's operating range, keeps it safe, and basically lets it produce X amount of output voltage across a specific load and draw Y amount of current to produce Z amount of power for a given input signal.

Putting too high a resistance on the amp affects the whole output circuit, and allows the tube to operate outside the safety limits imposed by the original design, as if the amp is "trying" to put it's rated power into whatever load you throw at it, though it can only do so safely into the load it was designed to handle.

If you put a 100ohm speaker on a 100W amp designed for an 8 ohm load, instead of putting out 28.3 volts and 3.5amps, the tube sees less current flow and puts out a higher voltage. It "thinks" it is still hooked to the correct load, and is "looking" to make a certain amount of power, it seems. It compensates with higher V to make up for less C.

With a 1000ohm load, current flow is down further, and voltage goes even higher.

At some point, either the tubes or the transformer give up the game, and your amp is smoked.

Am I getting closer?

 

[ggunn]

Thanks for the benefit of the doubt. I will find out for sure, but here are some further ponderings.

What you said about ohm's law is true, but we are talking about a circuit designed to produce power into a specific load.

Making it about power changes a few things. Maybe that's what I meant when I said thing were a little more complex.

With the addition of a specified power, voltage and current have an inverse relationship as the resistance changes.

100W@ 8ohms = 28.3 volts, 3.5 amps
100W@ 16ohms= 40 volts, 2.5 amps
100W@ 1000 ohms= 316 volts, .316 amps

As current goes down due to higher resistance, voltage has to increase to keep the specified power output.

In addition, like you said, the AC voltage swings in the tube are small. But those voltages are determined by the rest of the circuit including the speaker, not the tube itself. The tube has no built-in limits or operating range. It will produce higher voltages or draw more current to make power until it or something else fails. The rest of the circuit, including the speaker, determines the tube's operating range, keeps it safe, and basically lets it produce X amount of output voltage across a specific load and draw Y amount of current to produce Z amount of power for a given input signal.

Putting too high a resistance on the amp affects the whole output circuit, and allows the tube to operate outside the safety limits imposed by the original design, as if the amp is "trying" to put it's rated power into whatever load you throw at it, though it can only do so safely into the load it was designed to handle.

If you put a 100ohm speaker on a 100W amp designed for an 8 ohm load, instead of putting out 28.3 volts and 3.5amps, the tube sees less current flow and puts out a higher voltage. It "thinks" it is still hooked to the correct load, and is "looking" to make a certain amount of power, it seems. It compensates with higher V to make up for less C.

With a 1000ohm load, current flow is down further, and voltage goes even higher.

At some point, either the tubes or the transformer give up the game, and your amp is smoked.

Am I getting closer?

I don't know. If you look at power amp output ratings, they roughly double as resistance halves, so I am not sure that your assumption that your theoretical amplifier puts out 100 watts into whatever load it is presented with is accurate. Also, the voltage output of a tube is limited by its bias voltage; it cannot create energy, it can only gate it. That's why they are referred to as "valves"

I must google some more...

 

[boingoman]

I don't know. If you look at power amp output ratings, they roughly double as resistance halves, so I am not sure that your assumption that your theoretical amplifier puts out 100 watts into whatever load it is presented with is accurate. Also, the voltage output of a tube is limited by its bias voltage; it cannot create energy, it can only gate it. That's why they are referred to as "valves"

I must google some more...

You are right about the 100watt thing- I was anthropromorphising a bit.

I just got a serious tube head to explain it to me. I didn't understand some of it, but here is what I got. Perhaps it will point you somewhere.

The load helps keep the tubes in their proper operating range. The proper load is in part determined by the plate voltages. The plates need to see the proper plate-to-plate impedance to generate their output voltage.

If the load is too low, the tubes cannot swing their max voltage. If it is too high, the voltage in one tube spikes.

There is DC voltage on the primary. There is also transformer action on the primary, as it has a tap, the center tap that provides current to the tubes.

With no load in a push-pull pair, the tube that would normally be shut off by the bias voltage does not shut off, as due to the high impedance, it cannot draw current. Bias voltage sets idle or shut-off current.

The transformer wants to maintain voltage balance. With one tube at say 200 volts, the other tube, freed from the restraint of idle current, does not shut off and is compelled to pass the remaining voltage.

As one side of the transformer goes to zero, the other has to go to the full voltage. There is part of transformer action I didn't understand.

This is a part I kind of got lost on, as I didn't think to ask if the output voltage swings very high, or the plate voltage does. I think it is the plate voltage, which is in part determined by the proper plate-to-plate impedance.

Due to the transformer action btween the two ends of the primary caused by the center tap, the voltage potential available to the tubes can be on the order of three times the DC on the primary, which is usually about the same as the plate voltage.

So the plates and transformer primary can see upwards of twelve hundred volts in a lot of amps, if there is no load present.

In addition, the lack of load leads to high-frequency oscillation, as the reactances in the transformer, ignored when the amp is properly loaded, are suddenly treated as a high-strength input signal. Essentially the amp can go instantly to full power output, even with no input signal.

 

[cellardweller]

I was anthropromorphising a bit.

Huuuuhhh?!?

Anyway... Not to subtract from the subject at hand, but how hard can it actually be to bias your own tubes?

I mean if you have a nodding acquaintance with a volt-meter, and have access to one... what more does it entail other than adjusting a pot or something until the outputs are equal???

Great read this thread... Carry on!

*EDIT*Any tutorials out there that you know of about this???

 

[travelin travis]

Man I gotta learn more about electronics, amps. If only someone would write "Tube Amp Building for Dummies".

 

[boingoman]

Man I gotta learn more about electronics, amps. If only someone would write "Tube Amp Building for Dummies".

Building them and fixing them isn't that hard. I do a fair amount of amp repairs. I can diagnose and fix a lot of common problems, a few uncommon ones, and have built a couple of amps. Even many preamp mods are pretty easy.

Understanding exactly what happens inside them is a bit more involved, apparently.

Luckily, you don't have to, a lot of times. It's like being a mechanic or any other kind of tech. A deep understanding of some of the more intricate and involved things isn't necessary to get the job done.

The Desktop Reference of Hip Vintage Guitar Amps by Gerald Weber
Inside Tube Amps by Dan Torres
Dave Funk's Tube Amp Book

All good places to start, especially Weber's book. A lot of people talk about Aspen Pittman's Groove Tubes book, but I though it was pretty worthless as far as learning anything.

 

[cellardweller]

Thanks for the references boingo. I assume this would explain the how/why of biasing, and then some?

 

[travelin travis]

yea, thanks for the references.

 

[ggunn]

Yeah, this stuff is starting to make some sense. Here's a web page that illustrates the effect without explaining it too deeply, other than to say that the extreme voltages are due to "energy storage" in an output transformer with no load on the secondary:

http://www.w8ji.com/demonstation.htm

BTW, I appreciate the exchange of ideas without rancor. Many times my quest for information in some fora has drawn responses that contain more posturing and vitriol than any genuine effort to get to the essence of a subject. Thank you; I am learning a lot in this discussion.

Oh, and I have Weber's book (it was a lot of help to me in "blackfacing" my '68 Super reverb) and another one about the same size that may be Funk's.