Symmetrical clipping for single-ended output stage?

[utervo]

Is symmetrical clipping possible with single-ended output stages?

How can one tune a simple 6V6 output stage for symmetrical clipping?

[jjman]

Yes.

Bias at the mid point and ensure the driving stage is strong enough.

May need to reduce plate voltage if the mid-point violates static dissipation max. A 6v6 would probably need to be less than 325 volts (plate to cathode) or even less, to achieve this in my guesstimation.

[Firestorm]

Yes.

Bias at the mid point and ensure the driving stage is strong enough.

May need to reduce plate voltage if the mid-point violates static dissipation max. A 6v6 would probably need to be less than 325 volts (plate to cathode) or even less, to achieve this in my guesstimation.

Mid point of what? Just asking.

[martin manning]

That would be the center of the load line, half-way between cut-off and grid current limiting. The thing is that for a given B+ and load impedance there is only one bias setting that will center the operating point, and you would like that point to also be just about 100% plate dissipation. Practically that means matching B+ and load values.

I'm wondering why the desire for symmetrical clipping? That implies more odd-order harmonics like a push-pull output section generates, and moving away from the quality that is most prized in SE output sections.

[utervo]

That would be the center of the load line, half-way between cut-off and grid current limiting. The thing is that for a given B+ and load impedance there is only one bias setting that will center the operating point, and you would like that point to also be just about 100% plate dissipation. Practically that means matching B+ and load values.

What would be the method to achieve this?

I'm wondering why the desire for symmetrical clipping? That implies more odd-order harmonics like a push-pull output section generates, and moving away from the quality that is most prized in SE output sections.

I'm transforming my 5F1-like circuit to a single-ended 5E3 (thread https://tubeamparchive.com/viewtopic.ph ... 2&start=15. I've noticed that many single-ended amps have "something" in the distortion tone that I don't like, and I have been thinking if this is to do with the asymmetrical clipping.

[surfsup]

look at a 6V6 graph. Find your bias point and draw existing load line. find where it intersects zero at which grid curve (negative voltage value). rebias to half that value.

[martin manning]

What would be the method to achieve this?

See here: http://www.valvewizard.co.uk/se.html And note especially for center bias: Z = Va^2 / Pa. Anode voltage is B+ minus ~15-20V for cathode voltage, and Pa for a 6V6 is 12W. If you have 360V for B+, then you need a load impedance Z of (360-20)^2/12 = 9600 Ohms. If you have a load impedance of 5000 Ohms, then you need an anode voltage of SQRT(5000 X 12) = 245V, so about 260V B+.

[jjman]

I have a 6v6 SE close to "mid biased" at 21.9v cathode, 354 plate (332 net plate) and 43ma. Static plate dissipation 13.3watts. Screen is 3ma and this is a UL audio stereo power amp, which is why I wanted max swing before clipping. Each cathode resistor is 500R.

Biasing in the "middle" allows for max output before clipping and the clipping is on the "top" and "bottom" as clipping begins. Biased in the "middle" between the 2 clipped endpoints on the signal output wave. I use my scope to see it.

This is SE talk. PP is usually different unless Class A, which it's usually not.

[utervo]

I'm still a bit lost, I need to read some books tomorrow. My B+ is around 366 and cathode voltage 20, so anode voltage is 346? Cathode resistor is stock 470r and output transformer is 5k to 8 Ohm speaker. What is the equation to count the correct cathode resistor or the midpoint biasing?

Thanks a lot, I wilö look into this tomirrow with a fresh mind.

[pdf64]

Mid point biasing can't be achieved with that combination of tube, VB+ and primary impedance.
Consider a 6L6GC, if your PT and OT can support the increased current and power.
Pete

[rdjones]

You could ground the cathode.
Bias with an adjustable negative grid bias supply.
This will eliminate any potential for bias shift through the resistor/capacitor.

rd

[Firestorm]

Somewhat related question: how low can the signal pull the plate in SE?

[pdf64]

You could ground the cathode. Bias with an adjustable negative grid bias supply

That would increase the effective plate voltage and so likely to require colder bias.
With cathode bias, the cathode voltage can be clamped by means of a suitable zener, to prevent it rising above the zener voltage.
A zener could also be used to lose some V from the B+.

This will eliminate any potential for bias shift through the resistor/capacitor

If center biased, it shouldn't shift much anyway over the linear range.
When signal level increases past saturation, grid current will force the bias more negative, whether it's fixed or cathode bias.

If the existing tube, OT and VB+ were to be used, another option would be to operate the OT with a higher load on the secondary, ie 16 ohm would give 10k primary load, which should be closer to mid point.
Pete

[rdjones]

You could ground the cathode. Bias with an adjustable negative grid bias supply

That would increase the effective plate voltage and so likely to require colder bias.
With cathode bias, the cathode voltage can be clamped by means of a suitable zener, to prevent it rising above the zener voltage.
A zener could also be used to lose some V from the B+.

This will eliminate any potential for bias shift through the resistor/capacitor

If center biased, it shouldn't shift much anyway over the linear range.
When signal level increases past saturation, grid current will force the bias more negative, whether it's fixed or cathode bias.

If the existing tube, OT and VB+ were to be used, another option would be to operate the OT with a higher load on the secondary, ie 16 ohm would give 10k primary load, which should be closer to mid point.
Pete

No, it shouldn't shift much at all over the linear range depending on the capacitor value.
The key here is what's happening at clipping, or more to the point, the difference between saturation and cutoff.

rd