Dumb biasing question

[Lectroid]

Somehow I've always thought that cathode-biased amps were subject to the same rule of thumb as fixed bias amps, that biasing at or below 70% of max dissipation is a good safe target.   But today I read a statement on RobRob's site that jumped out:

For All Cathode Biased amps 100% of max dissipation is your safe limit.

Is RobRob correct?  Can cathode bias amps be biased much higher than 70% safely?  Maybe not all the way up to 100%, no, but how much higher?

[SEL49]

Rob is correct. It's not uncommon to see amps running at 110% or higher.

[shooter]

yep


I try n dial mine between 90-110%


Idle is the most the tube will be dissipating, once you "add" a signal the bias "self adjusts" typically lowering the tubes dissipation.


all that is subject to "how it sounds" since self biased amps are sorta fickle when you OD the PA tube, sometimes it sounds good, sometimes it sounds like bad death-metal

[pdf64]

Well regarded cathode biased push pull amps pretty much invariably operate much closer to class A than class B.
As an example, the cathode voltage of AC30 rises from 10V at idle to 12.5V at full (presumably clean) output power.
https://el34world.com/charts/Schematics/files/Vox/Vox_ac301960.pdf

[stratomaster]

I worked on a Budda Twinmaster 18 that failed spectacularly recently. After recovering much of the damage (2x EL84s and their cathode cap, the output transformer, a 12AX7 for V1, the V1A section cathode network--which caught fire due to the innards of the output stage cathode cap causing a short, and the V1B cathode network due to proximity to the flame) I found the EL84s were biased at about 170%. 

The cathode resistor was within spec at 100Ω. That amp left my shop with a 180Ω resistor in its place, and still at about 110% static dissipation.

[pdf64]

... I found the EL84s were biased at about 170%. 

The cathode resistor was within spec at 100Ω. That amp left my shop with a 180Ω resistor in its place, and still at about 110% static dissipation.

That approach pushes the operating point towards class B at high power outputs, generally doesn't sound as good.
A better approach is usually to identify the root cause of the overdissipation, and either buck the incoming mains supply voltage down, or reduce the screen grid HT supply node voltage. 

[Lectroid]

Thanks to everyone for the great information.

Rob is correct. It's not uncommon to see amps running at 110% or higher.

on a Budda Twinmaster 18--which caught fire... the EL84s were biased at about 170%.

yep...I try n dial mine between 90-110%...Idle is the most the tube will be dissipating, once you "add" a signal the bias "self adjusts" typically lowering the tubes dissipation.

As an example, the cathode voltage of AC30 rises from 10V at idle to 12.5V at full (presumably clean) output power.

Wow.  Amazing to learn this.  I've built 5 cathode and fixed bias amps, and never know I'd missed this.  It's good to know. 

What is the mechanism by which the bias 'self-adjusts?'  If the bias voltage increases under load, does that decrease the plate current, to produce the self-adjusting correction?

Anyway, I got my dumb question answered.  So much to learn.  Thanks to all!

[pdf64]

...
What is the mechanism by which the bias 'self-adjusts?'  If the bias voltage increases under load, does that decrease the plate current, to produce the self-adjusting correction?

That's it!

[shooter]

^^^ what he said


when you 1st "setup" the amp, I call it diminished returns....


I start with the "target Plate VDC", which is a math n guess problem for the "series PA tube circuit"


once i'm close, I measure the tubes dissipation using the cathode R value, VDC drop to get "I" - current.


if it's close, I play for 15 minutes trying to decide "tone/sound/loud, grit n all manner of guitarist happy points"


then begins the diminishing returns, say you bump Rk, that changes current, which changes V-drop, which hopefully "dials the sound" closer to what i'm trying for.  eventually you get to a point where you're up against the laws of physics n well those laws don't care about amp sound