How would you design transformers in such a way so that you can run a tube amp without a load ?

[Cathode Ray]

Open the feedback loop, if there is one. The AC30, for instance, does not have output stage feedback.

Not an expert, but I'm pretty sure that won't allow you to run a tube amp with no load.

[maxkracht]

R.G. wrote: ↑Wed Sep 06, 2023 8:46 am
Open the feedback loop, if there is one. The AC30, for instance, does not have output stage feedback.
Not an expert, but I'm pretty sure that won't allow you to run a tube amp with no load.

R.G. IS an expert...

Anything connected to the output of the power tubes is a load. So, if there is no load on the output of the amp, the power tubes aren't being used to produce that sound. He could still have an amp that reconnects the power tubes when it's not in the setting in question and operate like a normal amp. He probably needs the power tubes plugged in to keep preamp voltages at a reasonable level, so they are being "used" even if they aren't passing audio.

Any of these points could be incorrect in relation to his amp because in the few minutes I watched, that guy was using technical terms as jargon instead of their actual meaning. So something like "load" might not mean anything. Only way to know for sure what is going on is to buy his amp and reverse engineer.

[Cathode Ray]

R.G. IS an expert...

Anything connected to the output of the power tubes is a load. So, if there is no load on the output of the amp, the power tubes aren't being used to produce that sound. He could still have an amp that reconnects the power tubes when it's not in the setting in question and operate like a normal amp. He probably needs the power tubes plugged in to keep preamp voltages at a reasonable level, so they are being "used" even if they aren't passing audio.

Any of these points could be incorrect in relation to his amp because in the few minutes I watched, that guy was using technical terms as jargon instead of their actual meaning. So something like "load" might not mean anything. Only way to know for sure what is going on is to buy his amp and reverse engineer.

If we take him at his word, he is passing signal through the power tubes.

So.. that's my quandary.

[TUBEDUDE]

You can run power tubes like preamp tubes. Disconnect the O.T. primary and run the plates with high value resistors to the B+. Then couple to the balanced line level xfmr I mentioned earlier.

[Cathode Ray]

You can run power tubes like preamp tubes. Disconnect the O.T. primary and run the plates with high value resistors to the B+. Then couple to the balanced line level xfmr I mentioned earlier.

He answers questions pretty reliably, if not in a "cagey" manner.

I'm going to see if I can get him to be a little more forthcoming about how he's running the amp.

He's saying the pre-amp section of the amp is a stock Marshall 1959 circuit.
The output section, however.. is not.
Also, he is using "custom wound transformers", and he also states there is no resistive or reactive load built inside the amp.

The power tubes are passing signal, and is key to the tone he's getting, and this amp will not drive a speaker - it is only equipped with a line out.

[TUBEDUDE]

As I just described.

[R.G.]

I pretty sure that I'm not the last word in tube amp expertise, but I do know that it is very easy for a tube output stage to break into gain-phase oscillation.

It may well be that opening the feedback loop that most modern output stages use won't absolutely allow it to be non-destructive to run without a load impedance for ALL amps. But it is certain that for some amps, the path to death from no-load conditions is through oscillation.

Tube output stages nearly always have an output transformer. OTs have leakage inductance, primary inductance, interwinding capacitance, winding resistance, all the things that make designing OTs tedious. The reactive components introduce phase shift, and phase shift makes for gain-phase oscillation in feedback amps. Three time-constants ("poles" in control theory jargon) are enough to cause oscillation with feedback as long as the open-loop gain is high enough. Tube amps have their output amps' gain limited specifically so they can use feedback to linearize, but not so much that they go into gain-phase oscillation. The OT itself is a two-time-constant device, so when you add a capacitance into the output tube grids, or a frequency dependent feedack path (presence?) or a capacitor to the PI, you have all three time constants. The only good cure for this kind of thing is to tinker the pole frequencies or decrease gain, or do fancy feedback stuff to compensate. Tube amps generally keep gain down.

There's a saying in control theory that everything oscillates eventually if you keep raising the gain. The load on output tubes is the reflected load resistance/impedance. This is generally down in the several-k region. Opening the output changes the plate load to being the primary inductance plus the two half-primary leakage inductances. The inductive impedance increases with frequency, and the voltage gain of an output tube can be modelled as its transconductance times the load impedance, which becomes, as they say, "large". Voltage gain increases, and with any feedback, a conditionally stable output stage becomes un-stable. The oscillation is generally high frequency and may or may not be audible. Big (for tubes) currents flow in the tubes, and the tubes try to reverse the current in the un-clampable leakage inductances, so they do their V = L* di/dt dance and make high voltages.

This can puncture the OT insulation, arc over the tubes, or arc around the socket pins. And it gets done at RF. There are other ways to kill stuff, but this is a good one.

Removing the feedback loop makes it much, much harder for oscillation to start.

[Cathode Ray]

It may well be that opening the feedback loop that most modern output stages use won't absolutely allow it to be non-destructive to run without a load impedance for ALL amps. But it is certain that for some amps, the path to death from no-load conditions is through oscillation.

Connecting this with something else the guy said in another video..

You're making a whole lot of sense here.

[jabguit]

A guy in Australia is doing this with amps he demonstrates on his YouTube channel.

that's Ossie - right here in Raleigh. known him for years

[Cathode Ray]

that's Ossie - right here in Raleigh. known him for years

Any chance you can shed some light on the design of his "Marshmellow" amp series ?

How's he running a 100w tube amp with "no load", as he states he's doing ?

[jabguit]

wish I could! haven't spoken with him in over a year and "cagey" is quite an accurate adjective....

cheers,

[Cathode Ray]

wish I could! haven't spoken with him in over a year and "cagey" is quite an accurate adjective....

cheers,

Yeah, don't get me wrong, I like what he's doing.

It's got me intrigued.. He's doing his own thing there (whatever it is) and I completely understand not wanting to hand out schematics on the Internets.

I would just like a high-level understanding of how he's running a 100w Marshall with no load.

[Cathode Ray]

He's saying the amp has a full power section, 4 EL-34s, an OT a PT, But he's taking signal off the OT.

timestamp 13:00

[teemuk]

IMHO, the feedback loop is pretty much disabled when the output tubes are.

So, like mentioned earlier, this suspiciously sounds like just disabling the power tube section (e.g. you can bias them to cutoff or just open the cathode circuit) and taking the signal from the PI. So, probably just a very "boring" solution and answer to the problem.

Because output tubes are not connected you can run the thing "open load", and simultaneously the feedback loop becomes "open" too. Not that it matters because there is no amplifier to oscillate to begin with. ...and which is BTW about what Blackstar amps do when you plug in headphones so nothing new under the sun.

Of course it will not react or behave even remotedly like a real tube power amp but that has always seemed to be more inconsequental in practice than fancy catchwords like "custom wound transformers".

[R.G.]

This is reminding me of my friend's comments on his gluten-free hand soap dispenser, or his noticing that some product is a floor wax AND a dessert topping...