Class A vs. A/B - Conducting 360 degrees?

Free

Member
Messages
1,304
I'd really appreciate it someone with technical isnight could clarify what 360 degrees and all the varying degrees of of conduction means exactly - why is at looked at in degrees? Also, what is happening exactly during this conduction?

I've learned Class A and Class A/B amps can be distinguished by these factors of degrees. I regret if this seems like a dumb question - I've got just enough technical knowledge to be dangerous. Thanks much in advance...

-Mike
 

B. Bop

Member
Messages
89
A sine wave is divided into degrees just like a circle. 0 degrees is were the signal crosses the center or 0 volt line and starts to go positive. 90 degrees is were the signal is most positive. 180 degrees is were the signal crosses the center or 0 volt line again and starts to go negative. 270 degrees is the most negative point and 360 degrees is were the signal crosses the center line again going positive. Since audio signals are basically sine waves then a tube conducting for 360 degrees (Class A) is conducting for the entire signal excursion. A tube operating in Class B would conduct for half of the wave say 0 degrees to 180 degrees. The other tube in Class B would conduct the other half of the wave from 180 degrees to 360 degrees. In Class AB the conduction overlaps. One tube may conduct from 350 degrees to 190 degrees and the other tube might conduct from 170 degrees to 10 degrees. In this scenario there would be 20 degrees of overlap between the tubes. The actual amount of overlap would depend on how the tubes were biased. Hotter bias = more overlap.
 

Free

Member
Messages
1,304
I forgot to ask about something. In the scenarios cited (350 to 190 degrees and 170 to 10 degrees) it implies a gap and not an overlap. My question is how is tone effected with a gap in the cycle in Class A/B? Also, is a small gap in the cycle acceptable in A/B? I can understand how overlap would cause crossover distortion, but I'm not clear on how a gap effects tone or how a gap is technically classified.

Thanks!

-Mike
 

Free

Member
Messages
1,304
...the "gap" is called "crossover" distortion and is compensated for by idle adjustment to achieve "overlapping" conduction...ie: both tubes conducting for a portion of the input signal (Class-AB).

I see. Thanks. So, what tonal effects are there with varying degrees of overlapping as opposed to a gap then? Can too much overlap create crossover distortion too? My understanding of crossover distortion was that of waves overlapping - like the crossover distortion one can get stacking distortion pedals, etc. I appreciate the clarification.

-Mike
 

mooreamps

Senior Member
Messages
372
I see. Thanks. So, what tonal effects are there with varying degrees of overlapping as opposed to a gap then?
-Mike

Hi all,

Well. just by virtue of running the outputs tubes in push/pull, regardless running class A or being driven hard enough into class A/B, a push/pull stage tends to cancel out more of the even order harmonics leaving more of the odd-order harmonics. This makes the tone less warm sounding with more sparkle and more crispness.

-g
 

IIIBOOMERIII

Silver Supporting Member
Messages
1,753
Great question Free!!!! I have often wondered about that myself. I do not own a scope so I never really worried about it, I never understood it.

Thank you Tele and B. Bop for the to the point answer and helping me to understand what crossover distortion is. I have even seen pics of scopes showing what the "notch" looks like but, I just never able to wrap my brain around it. You guys are the poop!
 

Wakarusa

Member
Messages
1,458
Regarding gaps and overlaps...

Keep in mind that in a push-pull power amp, the output tubes see equal but opposite (inverted) signals at their inputs. I.e. when half the power section sees max positive, the other half sees max negative.

In class A, all of the power tubes are always conducting. In any other class of operation, the power tubes go into cutoff (stop conducting) for at least a portion of their duty cycle (e.g. when the input signal goes far enough negative).

In class B, the power tube is biased to cutoff at zero volts. By definition then, it's also cutoff for any negative input signal. There's (in theory) no gap and no overlap.

In class AB, the power tube is biased to cutoff at some negative voltage. Keep in mind that any time one half of the power amp sees a negative signal the other half must be seeing a positive signal. Therefore, allowing one half to keep conducting for some part of the negative input means there must be some overlap because the other half is surely conducting with its positive input signal.

In class C, the power tube is biased to cutoff at some positive voltage. Using the same analysis from Class AB, it's easy to see this creates a gap in conduction -- a period in the duty cycle when the entire power section is in cutoff. This is why you don't see class C audio amplifiers :) This is also why class C amplifiers work really well for AM broadcast transmitters -- because all of the information is in the wave peaks, so the "gap" that cuts out the middle of the wave isn't really needed.
 

Free

Member
Messages
1,304
Regarding gaps and overlaps...

In class AB, the power tube is biased to cutoff at some negative voltage. Keep in mind that anytime one half of the power amp sees a negative signal the other half must be seeing a positive signal. Therefore, allowing one half to keep conducting for some part of the negative input means there must be some overlap because the other half is surely conducting with its positive input signal.

Thanks much, Wakarusa. So, this implies that there essentially are no gaps ever, since the negative voltage at cutoff neccessitates some positive signal? Then my previous assumption that crossover distortion is created from overlap is actually correct?

This whole overlap and gap difference still isn't clear to me, with these differing explainations. How can we clarify this in cycle degrees and various bias (hotter or colder) effects on the conduction cycles?
 

Wakarusa

Member
Messages
1,458
First, let's get apples to apples and oranges.. well, you know.

Distortion is any deviation between the output waveform and the input waveform except amplitude. I.e., it can be bigger, but has to have exactly the same shape for there to be no distortion.

Next, any push-pull amplifier is creating its output waveform by gluing together two waves to make one (the glue is the output transformer in a tube amp). Looking at the discussion of class B operation above you'd think that the output is distortion free because each side of the push-pull amplifies exactly half of the input waveform. The problem is that what's been described is the theory. In the real world, there are all sorts of nasty things going on that make the two halves not glue perfectly together.

Said another way, there are other sources of distortion in each side of the push-pull and each side may not quite get exactly half of the waveform. So when you glue the two halves back together there's a nasty bit that doesn't quite match up at the glue joint (i.e. where the signal crosses zero). As you've guessed, this is crossover distortion and it can be pretty bad in a real world class B amplifier.

In class AB when you glue the parts together, they overlap a bit at the zero crossing. When, in the real world, you sum these overlapping bits together in the output transformer, they tend to be a closer match to the original signal than if you had used class B. So, there is still crossover distortion (because the gluing isn;t perfect), but in a well behaved class AB design there's actually less crossover distortion than you'd get in a class B amp.

So, to answer your question directly: In the purely theoretical, there are no gaps ever in classes B, AB, and A. In the real world, there may be gaps in class B. But, whether there are gaps or not, the summing of other distortion makes some overlap in the pushed and pulled signals actually result in less crossover distortion.

Degrees and bias...
A colder bias means that at idle the tube is closer to cutoff than a hotter bias. So as the input voltage swings negative, the colder tube will cutoff before the hotter one. Therefore, the hotter bias will conduct through more of the input wave (more degrees) than the colder biased tube.
 

B. Bop

Member
Messages
89
Free, I e-mailed you a drawing showing sine wave degrees and Class AB conduction. I couldn't figure out how to post it myself. If you could post it it may help clear up some of the confusion.
 

Free

Member
Messages
1,304
Here you go, B. Bop. I got your e-mail. Thanks again! I'm sure everyone else will appreciate the illustration too...


Sinewaves[1].gif
 




Trending Topics

Top Bottom