Pentode vs. triodes tremolo comparison

[tubenit]

My favorite tremolo in any amp I've played was in my Dano Centurian which is almost identical to the Silvertone 1482.  It has a real slow deep vibrato/tremolo  sound to it & I always left it on when I played the amp.  Just a very cool tone, IMO.

It sounded a lot like the vibrato in this tune by Paul Black and the Flip Kings at 27 seconds in:
https://www.youtube.com/watch?v=NnU1mZg2nH0

So I keep wondering what exactly it was about that tremolo that made it sound so great to me?  Maybe it was the Dano amp and the speaker as a whole ........... OR maybe the 6AU6 pentode ............ OR maybe the tremolo circuit values???

Thing is that I don't know enough about pentodes to know how that would compare to a triode with a mosfet?

Any thoughts on which would have that deeper slow tremolo tone in the Tweed BluezMeister circuit?   

And I also wondered about changing out the tradition Fender tremolo circuit cap and resistor and pot values to match the Dano's?

I am thinking about maybe going the pentode route and using a 5879,  or a  6AU6 or a 5654 (as Sluckey loaned me one once and it was great in a tremolo circuit).

Thoughts??    With respect, Tubenit

[eleventeen]

All I can do is to offer an untested theory. Both circuits are phase-shift oscillators and thus they should produce sine wave outputs. So, why should sine wave "A" as a modulating signal sound different than sine wave "B"? It shouldn't.


However...(I have not tested this ) I don't know whether whether ALL phase shift oscillators output pure sine waves. For the most part, when you look at PS osc schematics in the idealized form, one generally sees a: triodes, and b: three identical R-C networks, the caps and the R's all equal. I can readily believe that would output a clean, symmetrical sine wave.


What I *can* tell you from direct experience is that a low-frequency phase-shift oscillator that isn't running a high-gain (12AX7) tube but instead is running on a low-gain 12AU7 can take a remarkable amount of time to build up to its ultimate state of self-sustaining oscillation. I know this because I have breadboarded such a circuit. I am not talking about how much time it takes for a dead-cold tube to light up and start working, I'm talking about a tube shut off in such a circuit, then B+ applied again.....with the filament still hot, though not lit up. It can take 3-4 seconds for such a ckt to start up upon application of fresh B+ with an already-hot filament.


With possibly-leaky old caps (like I used because I was just goofing around with junkbox parts) and a weak old 12AU7, it might not start at all! Same circuit with a good 12AX7 will start right up.  It's quite noticeable, not subtle at all.


Bottom line: With a much higher gain pentode in place of a triode, it would not surprise me to find a different output waveform with a more aggressive slope, and certainly a more rapid and reliable startup of the oscillator.


I think, if my curiosity was tweaked as you are suggesting yours is, I would breadboard just the trem oscillator in a way that I could scope it (not that easy to see very low freqs on a scope) and lash it into an existing amp to see how I liked it.


Me? I never use tremolo, LOL. 

[moonbird]

tubenit -


Hope you make this ampp -- looks cool. What is the circuity on the lower right of the diag? thanks.

[HotBluePlates]

I would consider simply replacing the 100kΩ resistor and 250kΩ pot in the stock circuit with the 1MΩ for each as in the Dano.

I'm not sure simply using a pentode as your oscillator will change the action of the circuit. But if you do opt for the pentode, notice the Dano oscillator has a 0.5uF cap for the screen bypass cap, rather than the typical 0.1uF.

That helps ensure the tube has good gain to the low frequency the pentode will be oscillating at. That screen bypass has an effect somewhat like a cathode bypass cap, in that if it is made "small enough" the tube's gain doesn't extend as low as it does with a bigger cap.

[Willabe]

Yeah man, I dig it!                (Paul Black and the flip kings)
   
Thanks for posting that.


               Brad     

[Backwoods Joe]

If you like slow tremolo, just remember to use larger cap & resistor values in the phase shift circuit. That will make the most difference. The gain of the tube shouldn't affect the speed of the oscillator.

[Heinz]

The tremolo in your schematic wiggles the bias of the output tubes. By doing so it moves the operating point of the output stage up and down which mainly modulates the volume but will also have some more subtle side effects. Pulling the bias voltage down will "starve" the output stage and will lower the point of "break-up" so it will add some distortion while the volume fades. Pulling the bias voltage up will increase the current and will raise the point of "break-up". This will clean up the tone as it is getting louder.

I assume that nearly everything in and around the output stage has an influence on these side effects, even the (master-) volume setting. If you hit the sweet spot you may perceive the sound of the tremolo as particularly pleasant.

Other types of tremolo like the Fender "roach" tremolo only modulate the amplitude of the sound signal without introducing extra distortion.

[Fresh_Start]

I'm guessing that Heinz may have hit on the magic here. This circuit is power amp bias vary tremolo with cathode bias. IOW it's not just the output of the tremolo section but also the interaction between the tremolo "signal" and the power amp as a whole.

I know that biasing a Princeton Reverb, you need to find the sweet spot where the tremolo sounds best and that's not necessarily the same bias point you might choose without the tremolo. Adding cathode bias to the power amp makes it even more interesting because the power tubes are always trying to self-bias.

Cheers,

Chip

[jjasilli]

A geofex article shows how to add SS tremolo to a tube amp - for an amp with no tremolo; or to free-up a tube for some other use.  The article takes the position that because tremolo is not in the signal path, it doesn't matter how it's generated.  But I think that different styles of tremolo do sound different, some more lush or musical than others.  But as Heinz points out, this might not be (entirely) because of the tremolo circuit itself, but in how it interacts with other aspects of the amp's topology.

[Heinz]

Thinking about your initial question something else came to my mind. While in theory the type of device working in the LFO should not matter there is yet another side effect if the LFO signal is high impedance (which pentodes happen to be).

If the modulating LFO signal has a large amplitude (high tremolo intensity) it will have a large voltage up and down swing. Swinging down will lower the output tubes' bias voltage to cut-off, silencing the amp. But when swinging up it will raise the bias voltage to the point where the output tubes' grids start drawing current. This effect starts gradually a few volts below cathode voltage and will deform the upper half of the LFO signal. So the modulating signal will not be a perfect sine wave but will have flattened positive peaks. This effect will be more prominent if the LFO signal is high impedance. A low impedance signal combined with a large coupling cap will be able to provide some grid current and will more or less maintain a sine wave shape. The effect will be audible as the absence of a well-defined volume peak. You can play with the values of the coupling resistor and coupling cap to modify the intensity of this effect.

As an alternative, here's yet another type of tremolo that I have built. If you feel like experimenting you can try it. It is neither a bias vary tremolo nor one of the "roach" tremolos. Instead, it uses a pentode as a variable gain stage by wiggling G2 voltage. I have never seen anything like this in a commercial amp (maybe I'm the inventor  ) but I can confirm that it works well. It has its own subtle type of sound. Again, G2 voltage does not only control gain of the stage but varies the operating point of the stage which has side effects.
The tube used here is a PCF802/ECF802 which is equivalent to a 6JW8. The input signal for the pentode must not be too large so the position in the signal chain must be chosen carefully (e.g. after the tone stack). If the input signal is too large the pentode will distort and distortion will be modulated by the LFO. While this may sound like a cool effect it sounds pretty bad at least with this circuit.

[kagliostro]

Talking about odd schematics

give a look here





and here





K

[Fresh_Start]

Holy cow Heinz!  That circuit has all kinds of interesting aspects. Tremolo on the screen grid of a pentode gain stage plus a split anode resistor to reduce output. Thanks for sharing.

Cheers,
Chip

[Stankfut]

Is that 6BE6 in the first drawing posted by kagliostro a heptode?  Never seen one of those before....

[Heinz]

Dammit! They stole my design 

Yes, that's a heptode. I have never seen one in a guitar amp either. But as I suggested, the best location in the signal path is right after the tone stack. The signal level is low enough for this circuit to prevent it from distorting.

I used the split load resistor to get gain of the stage down to 1 with tremolo switched off so I would be able to insert it anywhere without changing the gain structure. Of course, using a single load resistor is possible if the extra gain is desired.