Cathode bias shift

[kg]

the Vgk(avg) increases as the input signal increases, as it should. this is "cathode bias shift."

however, unless the amp is class A, the Ip(avg) will go up too. that's what generating the additional Vgk(avg).

simply, in most amps, the value of Rk is too low to offset the increase in Ip(avg) and what results is additional power dissipation.

in the class A case the bias shift is very slight, because it's only due to even harmonic distortion, and not tube conduction angle. in that case you will see a drop in tube power dissipation due to the AC plate current component being delivered to the load.

[MWJB]

Kg - is that while the amp is operating within it's class A power rating, or irrespective of how hard it is driven?

So the issue that EFK raised is more pertinent to class than to "cathode bias" per se?

[Chuck H]

So the issue that EFK raised is more pertinent to class than to "cathode bias" per se?

Well, if an amp is fixes bias there's no potential for a bias shift.

[Chuck H]

Ok... Simply stated please (and thank you). When the voltage on a cathode bias amp rises is the grid biased more negative WRT the cathode and therefor more prone to cutoff on large swings than if the bias were fixed?

[loudthud]

Well, if an amp is fixes bias there's no potential for a bias shift.

Dynamic bias shifts when the power tubes are driven into grid conduction. Additional charge accumulates on the coupling capacitors which moves the bias towards crossover distortion. If the drive is high enough, you will actually get a dead zone in the crossover region (both tubes are not conducting for an instant) and inductive spikes will appear on the output.

[Merlinb]

There are three mechanisms that lead to bias shift. One is blocking, which moves the grid more negative, which therefore tends to increase bias and push things more towards cut off.

The second is grid-current clipping. This also shifts the average grid voltage more negative, although it is a separate mechanism from blocking.

The third is cut-off clipping, which shifts the average anode current more positive and is equivalent to reducing bias (even in a fixed bias amp).

If the amp is centre biased the second two mechanisms more-or-less cancel each other out and the bias remains stable, leaving you with blocking to think about. If you want to cancel that out too then you actually have to bias *colder*, so that more cut-off clipping cancels out the grid blocking.

[tubeswell]

Ok... Simply stated please (and thank you). When the voltage on a cathode bias amp rises is the grid biased more negative WRT the cathode and therefor more prone to cutoff on large swings than if the bias were fixed?

Yes the grid is -ve w.r.t. cathode, but it in a cathode biased stage, it is actually the Vk which is increasing, so the Va-k is less than what it might be in a fixed bias stage all other things (HT etc, being equal). Since the Va-k is the bit which is driving the load, then there must be marginal additional energy that is being dissipated by the tube as heat in a cathode biased stage (compared to a fixed bias stage), which is the gist of what I understood loudthud to be saying.

[Chuck H]

Thank you guys. Though I'm sure you know that I was speaking as to how it relates to the thread and playing devils advocate. It's good to have any circumstances that are unrelated specifically to cathode bias included. BTW, plate voltage sag wasn't mentioned. Doesn't this also cool the bias and occur in both fixed and cathode biased amps?

[kg]

to add to merlin's excellent breakdown, technically the 4th is the one most classically discussed (ie rdh) and is the component due to even order harmonic distortion, and absolutely cannot be avoided (because no valve has purely odd order distortion). basically the "top half" of the waveform, where Ip is larger, will always have more area under it than the bottom half where Ip is smaller.

luckily, this is a much more subtle effect compared to cutoff or grid conduction.

[kg]

Thank you guys. Though I'm sure you know that I was speaking as to how it relates to the thread and playing devils advocate. It's good to have any circumstances that are unrelated specifically to cathode bias included. BTW, plate voltage sag wasn't mentioned. Doesn't this also cool the bias and occur in both fixed and cathode biased amps?

yes, b+ sag will offset it to some extent. how much depends on how much sag!

it is possible to at least partially counteract many valve distortions by cleverly arranging them to cancel each other out. imo this is why some hifi amps just sound "right."

[tubeswell]

BTW, plate voltage sag wasn't mentioned. Doesn't this also cool the bias and occur in both fixed and cathode biased amps?

The tube is a conducting device that is 'connected' between the B+ and the ground, and as its own internal (plate) resistance. Current x resistance = voltage (which in the tube is in an inverse relationship in terms of the effect of current on the plate voltage. As more current flows through the tube, the plate voltage goes down. In addition, as more current is demanded from the PT, then the ability of the PT to deliver more magnetising current is ultimately affected (beyond a certain point), causing a momentary loss of transformer efficiency, resulting in sag. If more current is flowing in more tubes, then beyond that point of transformer efficiency, the voltage at the B+ will be dragged down accordingly. That's how I reason things anyhow

[MWJB]

Chuck H asked "BTW, plate voltage sag wasn't mentioned. Doesn't this also cool the bias and occur in both fixed and cathode biased amps?"

Typically not in instrument amps (EFK did allude to this in his 1st post WRT cathode bias), though the voltage sags, it doesn't counter the increased signal current under heavy drive. The harder you drive an amp, the hotter it gets.

Your scenario could well be the case in cathode biased amps with high idle currents, power supply limitations, where plate current cannot be increased under signal & voltage must sag dramatically, but I'd have to speculate that this isn't typical of most instrument amps.

[Chuck H]

Ok, the tube IS conducting, yes, but if the actual -V reletivity of the grid stays the same or increases while the plate voltage goes down is that not the same as cooling the bias?

[MWJB]

My 6G# style amp has a fixed/cathode bias switch. At idle it disspates 70%, 49vdc accross the 680ohm cathode resistor, 486vdc (from plate to cathode) @ 36mA from plate to cathode (17.5W).

At full bore (not max W RMS power), controls set as per stage use plus vol full up (not as per W test) corrected plate voltage drops to 404vdc @ 81mA (less screen current), cathode voltage rises to 110v, grid voltage 89vdc & each tube is now dissipating 32.6W...didn't measure screen current, but let's say it maxes out around 10mA, that's still 28W of plate current.

At whatever power is being made at this point, dc dissipation is up. This is just an example & not meant as a precedent...I guess we need more cases to build a robust sample of typical behaviour? ;-)

[Chuck H]

OK... Everyone is talking about where the power is going and how much power is being dissapated, but... Isn't the bias (which is what we're discussing) the point between cutoff and saturation as determined by the grid conditions? So, if the grid -V relationship to the cathode remains the same (as in fixed bias) and the plate voltage drops, does that not incline the point between cutoff and saturation closer to cutoff?

I admit to being tech impaired and I'm trying to wrap my head around how all the different aspects of operation affect that point between cutoff and saturation independant of the grids -V aspect.