# Trying to make a more accurate bias chart

Discussion in 'Amp Tech Center' started by peteb, Jul 17, 2017.

1. ### RLee77Friend of Leo's

Posts:
4,437
Joined:
May 15, 2016
Location:
Also consider that max power out does not necessarily equal 2 x 14w for class AB. In push pull you can exceed max power rating on a tube briefly, because it will be off when the other tube goes into conduction, lowering the average. However, the real output rating is going to be somewhere between 14 and 28.

peteb likes this.
2. ### Old Tele manFriend of Leo's

Posts:
3,941
Joined:
May 10, 2017
Location:
Once banned always banned
What I'm saying is idle wattage is totally different from output wattage...yes, they're both the same 'kind' of watts...but they're two different aspects of tube operation. For instance, a pair of push-pull Class-A tubes actually operate at roughly ±100% of their rated plate dissipation wattage, but their actual output (AC) wattage is the result of BOTH tubes combining their currents.

Here a paper illustrating the separation of idle wattage from output wattage and the "70% Rule" for Class-AB push-pull: http://thermionic.info/mccaul/McCaul_70-PercentPlateDissipationRule_2008.pdf

Last edited: Jul 18, 2017
peteb likes this.
3. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,378
Joined:
Dec 29, 2012
Location:
United States
"What I did in this thread is I rebiased my amp from a very cold 20 % to a moderate 50 % by changing the way I calculated the bias, and best of all to me is I was able to raise the bias without actually raising the current, which makes an amp heat up."

?

peteb likes this.
4. ### SnfoilhatTele-Afflicted

Age:
40
Posts:
1,396
Joined:
Apr 8, 2016
Location:
Oakland, CA
This sentence seems to me to hold the key to the ongoing misunderstanding. There's an equivocation with the word 'power.' Idle plate dissipation is heat (waste) necessary for the DC bias the (class A or AB) tube requires to amplify a signal when one appears on the control grid. "AA964 Princeton 12 W RMS" or whatever it says near the speaker jacks is an estimate of the useful work the OT/wires/speaker are going to do within the 5% or whatever noise threshold. They don't add together, happening in different places. The amount of energy the 6V6 plates may be shedding from moment to moment might even track in some way with signal output power, but the '12W' of energy the Princeton 'puts out' (into the speaker) are not happening on the plates in some simple additive way. Right?

Another way to look at it: if all the 'output power' were in fact being shed as heat from plate dissipation, what would be left to move the speaker?

Edit: (I'm not an amp tech, and much more knowledgeable folks have already provided great explanations. I have taught science classes for years and can't help but see this as an interesting conceptual/communication problem. If I step wrong on a technical point, I expect someone will correct me.)

Last edited: Jul 18, 2017
peteb likes this.
5. ### Old Tele manFriend of Leo's

Posts:
3,941
Joined:
May 10, 2017
Location:
Once banned always banned
Idle (wattage) is what must be wasted in order to get output (wattage). Why? Because idle voltage sets idle current which "sets/establishes" the tube operating/conduction-angle conditions: A, AB, B, C, etc.:

Last edited: Jul 18, 2017
6. ### clintjFriend of Leo's

Posts:
4,957
Joined:
Apr 4, 2015
Location:
Idaho
Let's try a different approach to communication here. Uncle Doug is a far better teacher than I am, so let's let him have the floor for a few minutes. There are 2 videos that follow up with biasing a push/pull amp, but we need the basics first.

peteb likes this.
7. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
I admit that I was comparing the idle power to the output power, and confusing the fact that they are indeed two seperate things, and I've been confused by that before. But

The one contributes and leads to the other and there is a reason why they say to set the idle wattage to be 50-70 percent of the total tube output. The total tube out put roughly equates to total power out of the amp. So when you set the idle current to 50-70 percent of the total tube output, you are also roughly setting the idle power to 50-70 percent of the total output of the amp. It will be a higher percentage than 50-70 as the amp power is usually most but not all of the total tube output.

So I was asking if it was ok for the idle current power to exceed total power out of the amp, and I was also saying that it sounded like a hit bias and I think it turns out that it is.

This is from my notes:

On a bassman head, sweeping from 48-72 % bias, the idle power is always lower than amp power out.

On the cold biased 6g2 mentioned above. Because the wattage of the amp is so much lower than the total tube output, and with a more normal bias like 53 %the idle power is more than max power out on the amp, and I can say it sounds fine that way. With a cooler bias of about 30, which also sounds fine and loud and full, the idle power is less than max power out. So it can go both ways, but I think it is more normal for idle power to be less than full power of the amp,

Which reinforces my pioint, on the 12 watt amp, it appears that lower bias points than the typical 50-60-70, will get the idle current to match the amp output power which is normal.

8. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
Clint, I appreciate the video. I will probably check it out.

I consider G webers book amp biasing 101

Aiken amp pages are amp biasing 201

I think the majority of the posters on this thread have taken 101 and 201.

The standard bias calculation is just the beginning, it is a rough rule of thumb and people are told to consider the individual amp when using it.

I'm trying to get closer to understanding the tube behavior when the basic bias numbers can't do that.

I am particularly interested in the bias settings of an underpowerd amp like the Princeton.

9. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,378
Joined:
Dec 29, 2012
Location:
United States
Who says that? I've never seen anyone recommend that. Tube output is very different than max plate dissipation.

The primary factors driving "total tube output power" are control grid AC voltage and plate DC voltage but an undersized output transformer will limit total amp output power--that's why the non-reverb Princeton is rated at 12 watts.

I disagree with your hypothesis that a 12 watt Princeton should be biased cooler than a 20 watt Deluxe Reverb. Typically biasing at 50 to 70% for both amps will get you the best tone.

Last edited: Jul 18, 2017
peteb likes this.
10. ### Old Tele manFriend of Leo's

Posts:
3,941
Joined:
May 10, 2017
Location:
Once banned always banned
Plate dissipation wattage is a "limiting" condition of the tubes design, it's how much heat (in watts) the plate can dissipate continuously without incurring damage.

However, it's not exactly a brick wall number, since many Class-A operating output tubes are routinely biased slightly above their maximum plate dissipation for a reason--to take advantage of a quirk of Class-A operation--whereby 'maximum output' plate dissipation drops slightly due to signal-rectification. Thus, during heavy output, the plate dissipation power actually falls slightly bringing the total dissipation down slightly, roughly matching the tubes specification limit.

Last edited: Jul 18, 2017
peteb likes this.
11. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
OTM! I love class A, and I'm not quite analysing it as I am AB right now, but there may be an axiom that would fit in right now. Class A occurs when the _______power exceeds the ___________power. What is that?

Rob, There was a misunderstanding between you and me on that last pass.

When I say total tube output I mean the max plate dissipation, same as you, the total possible output of your tube type and then set idle power to 50, 60, or 70 % of that.

This is my hypothesis:

High power amps, where the max output power of the amp is 75% or more of the max plate dissipation of the tubes, if biased to the standard 50, 60, 70 rule, will have a idle power of less than the output power of the amp.

Lower powered Amps are different. If you set the bias by the standard 50, 60,70 rule, the idle power is going to be more than the output power of the amp. Therefore, a lower bias than the standard 50,60,70 rule will be needed to bring the idle power below the amp output power.

Side notes: it is OK if the idle power exceeds the amp output power, it still works and sounds the same, but I am not sure if it means crossing into another class of operation or not?

Conjecture: since all of the high powered amps have a idle power less than amp output power, then if a low powered amps idle power is equal to or very close to the amp power, then it is biased high enough, and the 50-60-70 rule does not apply.

That is my confusion, I mean my conclusion.

Last edited: Jul 18, 2017
12. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
Uncle Doug has a nice delivery, he's a sincere and likable guy, but he gets everything backwards.

I give him a pass when he gets cold and hot bias mixed up, but there is a good reason that more current is called the Hot bias, there is more power, and heat.

But when he gets the effect of I increasing the cathode resistor wrong, he says it will cause more distortion, it's not so effective.

I don't know Rob, I really don't, but increasing the bias is like opening up the valve and letting the power flow.

Doesn't it seem right that the 6v6 producing 10 watts of power is going to be biased hotter than the 6v6 making 6 watts?

13. ### BendyhaFriend of Leo'sSilver Supporter

Posts:
2,720
Joined:
Mar 26, 2014
Location:
Northern Germany
Okay ...I'll have one more go at this...

For the sake of comprehension and clarity....try to avoid the term "Idle power" ..."Idle current" is relevant to tube operating point & function...the "wattage" that you call "Power" is only relevant to the heat a tube is creating to fulfill this current handeling function. All this heat is wastage, unwanted (but an unavoidable byproduct), and can be potentialy damaging to the tube, that is why we muct monitor it.

"Amp Output Power"...I can live with. This is usually thought of as.....well various things.....(RMS ?) (Before clipping ? )..the amount of energy delivered to the output transformer...or speakers...? The efficiency of the output transformer makes such huge a difference.

For a totally different analogy;

Spark plugs... They need a certain voltage and current to work. To little = weak spark...to much = burn out. The mechanic must adjust to get the motor running smoothly.

A big powerfull car will have bigger spark plugs drawing lots of energy...even when the car is idling, which is when the designing mechanic set the level. But the potential power of the car is huge compared to the spark energy.

A small single cylinder scooter will still need a big hot spark to get the motor turning, not as big as the car above, but the spark plug will get almost as hot on both, so compared to the
energy difference here, the potential power of the scooter is not that big compared to the spark energy.

As related here, the spark plug power, and the engine output power have a relationship of; big to big, small to small, but there is no relationship that can be drawn from sparkplug heat dissipation/wattage/power, to, engine output /wattage/horsepower.....but this is the conclusion you seem to want to draw for amplifiers.

peteb likes this.
14. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,378
Joined:
Dec 29, 2012
Location:
United States
No. Here's the problem--the 12 watt and 20 watt 2x6V6 fixed bias amps' power tubes are producing almost identical "tube output power". They're both running with around 420v on their plates and they're both biased near the same % of max plate dissipation from the factory (they're both flowing about the same bias current at idle). The two output transformers are responsible for most of the amp output power differences.

peteb likes this.
15. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:

Rob, are you sure about that?

I agree that a Princeton non verb and a deluxe reverb could both (theoretically) be biased to anywhere between 50-70 % of the max plate dissipation of a pair of 6v6 tubes, but is this true in practice?

Are not there deluxe reverbs biased near or even over 100 % and not only are they biased super hot, but the bias level Is 'boxed in' with little room for adjustment? Lower the bias and the voltage goes too high, lower the voltage and the current goes too high? How often do you hear of Princeton non verbs stuck in a high bias? We don't hear much about biasing non verb princetons, I assume that this is because they are not biased too hot, and I am predicting that they are usually biased much cooler than deluxe reverbs.

Is not the Princeton non verbs circuit adjusted to match a smaller OT, and isn't the deluxe reverbs circuit adjusted for the larger OT. I find it hard to believe that (basically) the only difference between the two is the OT.

16. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
Uncle Doug is thought provoking. There was another thread and another uncle Doug video, and I had to pick up where he left off, following his presentation technique, and making pictures that relate bias to operation class. And I made this picture:

17. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
The purpose of this thread is to relate bias points, like the 50,60,70 points, to the operation point of the tubE, including the relationship with cutoff and saturation.

Here goes:

class B: conducts for 180 degrees of the cycle, idle is right at cut off, and that would put it right at 0% of max plate dissipation.

Class AB: conducts between 180 and 360 degrees of the cycle but not 180 or 360, idle is from just above cutoff to almost the halfway point between cutoff and saturation, the percentage of max plate dissipation is from just above zero percent to almost but less than 100%***

Class A: conducts 360 degrees of the cycle, idle is centered or halfway between cutoff and saturation, this is at 100% of max plate dissipation***

?????? Is it true that if a class AB amp is running over 100% plate dissipation that it is actually running class A????

*** these values depend on the 100% of plate dissipation being placed halfway between cutoff and saturation, which is convenient, makes for a nice picture, but I'm not sure how realistic it is. ?????

Back to the original question, can % of max plate dissipation values be used to semi accurately predict the operating point of the tube. Now I'm seeing it. The operating point of the class AB amp is where does the idle fall on the class AB spectrum? How close is it to class B? How close is it to class A? Once the operating point is set, the size of the signal, combined with the operating point will determine how much saturation you get, but the size of the signal is not the focus of this discussion, the bias point is and its affect on operating point.

If 100% plate dissipation really is halfway between cutoff and saturation, (I understand that is probably a gross simplification) then the answer to my first question is relatively simple. 50 % bias places the idle at 25% of the way from cutoff to saturation, and 70% bias places the idle at 35% of the way from cutoff to saturation. And 50 % bias is right in the middle of the class AB range and 70 % bias is 70 % of the way from class B to class A in the class AB range.

Question 2, how does this look on a low powered amp like the 12 watt 2x6v6 Princeton?

Last edited: Jul 19, 2017
18. ### Old Tele manFriend of Leo's

Posts:
3,941
Joined:
May 10, 2017
Location:
Once banned always banned
Uh, before you attempt to prove the unprovable, I suggest you take a look at this brief paper explaining how idle current and output power (actually plate current swing) DO have a relationship, but it is NOT the relationship that you're trying to prove:

http://thermionic.info/mccaul/McCaul_PlateConductionAngle_2008.pdf

As, robrob mentioned above, you seem to be leaving out/omitting the role that OT impedance (Z) plays in the whole Class-AB1 push-pull scheme (basic P = R*I-squared):

Po ≈ (Zo/4)*(Vg*gm')^2 ...assuming there's enough Vp-"swing" available.

The drive signal (Vg) from the phase-splitter combines with the OT tube's effective transconductance (gm') causing the tubes to conduct more. That increased plate current creates an increased AC-voltage drop ("swing") across the OT impedance...the greater the Z the greater the voltage drop. However, that process is LIMITED by the plate voltage available, the plate voltage 'swing' can not exceed the available plate voltage (B+)...and cannot go all the way down to zero-volts due to the tube's "diode line" limitation.

Last edited: Jul 19, 2017
19. ### petebFriend of Leo's

Posts:
4,240
Joined:
Apr 25, 2003
Location:
Let's say The last post only is about amps where they are going for full power from the tubes.

Now low powered amps:

We can break it down by considering amps that either lower the power before the power section, after, or in it.

Before,

The signal fed to the power tube has reduced voltage and amplitude.

The bias is in the regular place, but the AC signal placed on it is reduced in size.
In this case it may be preferable to increase the bias, the signal is small, if tube saturation is desired, then increasing the bias Will help. This could be the case for the blackface Princeton, they are known to stay clean, and a hotter bias would bring the smaller signal up towards saturation.

After:

like Rob said, the Princeton has a smaller OT than the deluxe reverb.

In this case, normal rules apply, the reduction happens after the power tube.

During:

Reducing voltage or current, or both across/thru the tube.

The bias is low, but the signal is still full size.
In this case, just like in the 'before' it will have reduced saturation and increasing the bias would help.

This is the only amp of the three which is meant to have a low bias. It would have less saturation. It would get more saturation with a higher bias, and it could safely be done as the tube is low biased.
This could be the brown Princeton, the voltages are much lower, but brownies saturate and don't sound cold. Maybe the signal is bigger?

20. ### Old Tele manFriend of Leo's

Posts:
3,941
Joined:
May 10, 2017
Location:
Once banned always banned
As mentioned earlier, the Princeton has a cathode-follower phase-splitter which does NOT provide signal gain (and in reality there's a slight loss!), hence, it cannot drive the output tubes as hard as the other amps do because their phase-splitters DO produce gain...and that's regardless of B+ and OT.

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.