Mini amp questions

[lazzer408]

I'm back into tinkering with tubes again now that the cold weather is here.

I found this schematic online and had a question about the output stage. It lacks a dedicated phase inverter and I don't see why this method of phase inversion was chosen over the option shown in the second image. Can someone explain? Any advantage to the design in the second photo?

[Printer2]

One less triode needed.

[lazzer408]

Originally Posted by Printer2

One less triode needed.

How so? There's 3 in both drawings.

[Miles Prower]

Originally Posted by lazzer408

I found this schematic online and had a question about the output stage. It lacks a dedicated phase inverter and I don't see why this method of phase inversion was chosen over the option shown in the second image. Can someone explain?

That first design is a self splitting LTP/PA in one. This particular implementation is horrible. A 440R resistor isn't nearly enough tail load, and a CCS would be a helluvalot better (though it'll require a negative rail). It also enforces Class A operation, since the limit of plate current swing will be between 0mA and Itail.

Any advantage to the design in the second photo?

It will probably perform better than a cathodyne if your DC rail voltage is limited. (In that case, you could always substitute a MOSFET for a triode and use that for your cathodyne (sourceodyne?) splitter instead.

[tubelab.com]

The phase splitter (first triode) in the first schematic has zero gain. The gain stage in the second schematic has a good deal of gain but the output tubes have less gain since they operate as an LTP with a shorter than usual tail. There are trade offs in both cases and the self split version (second schematic) will have a somewhat lower total power output and imperfect balance. This will cause it to sound somewhat like an SE amp, but allow the use of a P-P OPT.

I explored many versions of this circuit and settled on something quite like this for the 2 watt guitar amp in the Hundred Buck Amp Challenge thread. The schematic is in post # 1021.

The Hundred-Buck Amp Challenge

I now have a version using a mosfet phase splitter that can be disabled allowing the 2 watt "quasi SE sound" and 5 watts of full P-P pentode terror.

[lazzer408]

Originally Posted by Miles Prower

It will probably perform better than a cathodyne if your DC rail voltage is limited. (In that case, you could always substitute a MOSFET for a triode and use that for your cathodyne (sourceodyne?) splitter instead.

I thought the 2nd pic IS a cathodyne PI.

Originally Posted by tubelab.com

The phase splitter (first triode) in the first schematic has zero gain.

The first tube in the first drawing isnt a phase splitter. It's just a gain stage. Shouldn't a plate follower have some gain?

The second stage doesn't have a gain stage. It's just a cathodyne phase inverter (from what I understand).

I'm wondering why that designer chose to build the output stages the way he did. My thought was he could have made better use of that first triode as a PI. If he required it for gain, he would have been better off paralleling the triodes of V3 and wiring it for SE to keep his first triode gain stage.

[Svein_B]

The "Compact Amp" with the self splitting output tubes was published in Electronics World, June 1961: Compact Hi-Fi Power Amplifier by Melvin Leibowitz

As pointed out above, it is not a great concept. It has been elaborated in a number of steps published at: Simple EL84 poweramp/integrated

SB.

[tubelab.com]

The first tube in the first drawing isnt a phase splitter. It's just a gain stage. Shouldn't a plate follower have some gain?

You are right. The schematics arent visible while posting a reply and my brain goes fuzzy at 11PM, that's why I turn the bench power off at 10PM to avoid shocking moments. I got the schematics reversed in my post of last night.

The bottom schematic is a conventional design and the first stage phase splitter has no gain. The unbypassed cathodes in the output stages will cause a mild gain drop as long as the tubes remain in class A and a rather serious gain drop if driven into AB teritory. The output impedance increases in both cases. Add a cap if you need more gain.

The top schematic is a self split output stage that acts as a rather imperfect LTP. The input stage does indeed have gain. The major advantage here is a low parts count, since it doesn't do anything real well. It does allow the use of an ungapped P-P OPT in a circuit that provides a somewhat SE like sound. I am exploiting this property in a guitar amp that has $40 in total parts count. The extra output tube costs $1. The difference between a SE and a P-P OPT is $7. The design works but I can't get the THD below 2% at any power level and it rises with power output. It's good for a guitar amp, but I wouldn't try it for HiFi.

[Merlinb]

Originally Posted by lazzer408

I'm back into tinkering with tubes again now that the cold weather is here.

I found this schematic online and had a question about the output stage. It lacks a dedicated phase inverter and I don't see why this method of phase inversion was chosen over the option shown in the second image. Can someone explain?

Because it's a guitar amp. Not a hifi amp.

The whole point of this amp was to get very low output volume (power), and lots of gain. A 'proper' PI would sacrifice gain but result in more output volume than necessary, the exact opposite of the design intention.

Here's the full schem for the benefit of everyone. Notice how it's totally anti-fidelity...?

[Frank Berry]

Why not build a SE amp?
Your schematic will not give true push-pull operation. There will be much more 'push' than 'pull' which will severely limit the available output power (and increase the distortion).
Since you don't require a whole lot of power, it would seem that a SE design might be a better and less costly solution.