Autobias amps

[G-Spot]

Hi, I'm new in the forum and to amp electronics.

I have a question... a big problem with Valve amps is setting the bias, right?

So I thought (certainly not the first one) why the amps aren't all auto-biasing?

I discovered this
Untitled Document

Could this be applied to guitar and bass amps?

[MWJB]

This refers to a hi-fi amp. The goals are a little different (low distortion, wide frequency response) when compared to a guitar amp.

Yes it could no doubt be applied to guitar & bass amps, but would probably price the amp in question way outside of the competition.

Many fixed bias amps nowadays come with a bias trim pot & bias test points, meaning power tubes can be biased in a few minutes. In an amp with no bias adjust pot, typically a tech can install one for not much more than an hours labour/bench fee.

Cathode biased amps don't often need to be rebiased, unless you are modifying things & then you should know how to bias an amp anyway.

Bias probes are widely available to allow owners to at least check their plate current, if you take your amp & bias probes to a store/tech with a good selection of tubes, you will normally be able to get it biased up without even opening it up.

So, in short, the probem, is not that "big" in the first place & cost effective solutions are available. Realistically, even these cost effective solutions are more that your average guitar player wants to pay (most players don't think about retubing/rebiasing until a fault develops), so if he won't pay for a simple rebias, why would he shell out for expensive "auto-bias" circuitry?

[Steve Conner]

Autobias is difficult to make work in a Class-AB amp, and even harder in an amp that gets overdriven. (the white paper guy says "Steering signals are kept 5 octaves below the audible region", to me that means complete havoc as soon as you overdrive the thing)

So it's really best saved for Class-A hifi work.

Cathode bias is probably the closest you can get. It's not "far from optimal" when applied to guitar amps, even if some random hi-fi guy says so.

[pdf64]

Traynor seem to have an autobias system, see
http://www.traynoramps.com/downloads...an/smycv80.pdf
My buddy has got one of these, when I get the chance I'll have to see how well it works.
Peter.

[G-Spot]

Thanks for the replies!

So, even if money was not a question and the majority of guitarists actually cared about bias , technically it doesn't makes sense. Nice to know.

[teemuk]

John Broskie has also written a plenty about that stuff in his Tube CAD journal and blog.
John Broskie's Guide to Tube Circuit Analysis & Design
Google search should find the pages pretty fast.

Traynor seem to have an autobias system

The second edition of the SS guitar amps book I'm working on briefly discusses that circuit (along with other SS stuff working hand-to-hand with tube amps). Here's a screencap:

[Enzo]

And Gspot, I'd have to say setting the bias is not a "problem." Bias must be properly set, yes, just as the tire pressure on your car must be properly set. Setting tire pressure is hardly a problem, just something that needs be done.

BIAS BIAS BIAS is a mantra chanted by many, but really bias is not a critical thing. It is critical that it be there, but there is not some critical value of it. Your amp may work great at 40ma of plate current at idle, but it will operate well at half that too. And if your target is 40ma, 38ma or 42ma would be a change very hard to notice. SO in other words, a ball park adjustment is all that is required. ANy further tweaking is up to you. There is not enough bias current, and ther is too much bias current, but between those two there ios a fairly wide range of perfectly acceptable bais current levels.

And as soon as you start discussing bias, somebody will chime in that 45ma sounds best, an someone else will counter that 36ma sounds best. If your amp automatically sets the bias for you, then they had to decide what that target current was. What if they liked 36 and you liked 45?

In hifi gear, they don;t want any crossover distortion, but - to me - a guitar amp set for minimum crossover distortion sounds crummy. A very sophisticated circuit could be made to detect and adjust out crossover distortion, but i wouldn;t want that on my guitar amp.

[G-Spot]

And Gspot, I'd have to say setting the bias is not a "problem." Bias must be properly set, yes, just as the tire pressure on your car must be properly set. Setting tire pressure is hardly a problem, just something that needs be done.

BIAS BIAS BIAS is a mantra chanted by many, but really bias is not a critical thing. It is critical that it be there, but there is not some critical value of it. Your amp may work great at 40ma of plate current at idle, but it will operate well at half that too. And if your target is 40ma, 38ma or 42ma would be a change very hard to notice. SO in other words, a ball park adjustment is all that is required. ANy further tweaking is up to you. There is not enough bias current, and ther is too much bias current, but between those two there ios a fairly wide range of perfectly acceptable bais current levels.

And as soon as you start discussing bias, somebody will chime in that 45ma sounds best, an someone else will counter that 36ma sounds best. If your amp automatically sets the bias for you, then they had to decide what that target current was. What if they liked 36 and you liked 45?

In hifi gear, they don;t want any crossover distortion, but - to me - a guitar amp set for minimum crossover distortion sounds crummy. A very sophisticated circuit could be made to detect and adjust out crossover distortion, but i wouldn;t want that on my guitar amp.

Thanks for you answer Enzo (once again).

I have to be more careful when choosing words. Instead of "bias is a big problem" I should have said: "bias is a big topic"

Yes, it's amazing to see the amount of posts in guitar forums that suggest "the best bias setting" (and don't even mention at which plate voltage).

My question was more into the technical/engineering point of view, but if it isn't really needed or useful in guitar amps, then auto-bias is just an academic exercise

[G-Spot]

John Broskie has also written a plenty about that stuff in his Tube CAD journal and blog.
John Broskie's Guide to Tube Circuit Analysis & Design
Google search should find the pages pretty fast.

The second edition of the SS guitar amps book I'm working on briefly discusses that circuit (along with other SS stuff working hand-to-hand with tube amps). Here's a screencap:

Interesting reading

[Steve Conner]

Thanks for posting that circuit Teemuk. Would anyone care to try analyzing what happens when you overdrive hell out of it?

(hint: what happens when you overdrive the power stage in an AC30? Is the AC30 Class-A? What operation class will an amp with this bias circuit be? What is the maximum possible average plate current? Etc.)

[jrfrond]

And Gspot, I'd have to say setting the bias is not a "problem." Bias must be properly set, yes, just as the tire pressure on your car must be properly set. Setting tire pressure is hardly a problem, just something that needs be done.

BIAS BIAS BIAS is a mantra chanted by many, but really bias is not a critical thing. It is critical that it be there, but there is not some critical value of it. Your amp may work great at 40ma of plate current at idle, but it will operate well at half that too. And if your target is 40ma, 38ma or 42ma would be a change very hard to notice. SO in other words, a ball park adjustment is all that is required. ANy further tweaking is up to you. There is not enough bias current, and ther is too much bias current, but between those two there ios a fairly wide range of perfectly acceptable bais current levels.

And as soon as you start discussing bias, somebody will chime in that 45ma sounds best, an someone else will counter that 36ma sounds best. If your amp automatically sets the bias for you, then they had to decide what that target current was. What if they liked 36 and you liked 45?

In hifi gear, they don;t want any crossover distortion, but - to me - a guitar amp set for minimum crossover distortion sounds crummy. A very sophisticated circuit could be made to detect and adjust out crossover distortion, but i wouldn;t want that on my guitar amp.

As Enzo pointed out, bias is NOT a problem. The "problem" with bias is that it has become a buzzword for guitar geek-techno-wannabes. They actually WORRY about their bias settings! Man, in the old days, we slapped unmatched power tubes into fixed-bias amps (there WAS no tube matching) and sent them on their way! You know what? Everything was just fine and sounded great.

"Worrying" about bias is like worrying about your water pressure or the idle speed on your car. Like Enzo said, it's like checking your tires. Even THAT has leeway built into it.

Speaking of bias and idle speed on cars, I've learned that, in the end, my final test FOR BOTH is with my ears, after 30+ years of doing this. Bias voltage and cathode current numbers mean NOTHING to me, because each set of tubes IS slightly different, even if matched identically. In the end, it's all about sound and performance. Also, keep in mind that guitar amps aren't perfect and never were. They just aren't designed that way.

[G-Spot]

OK, OK, I got the point bias is not a "problem"

As I said, I was not good choosing the words.. I wish I had written "bias issue"... anyway... my post definitely got attention

yes... nowadays, everything is measured and controlled... probably too much.

[tedmich]

can CCS biasing be considered "auto bias"?

It gives the same mA no matter what right?



course there is a bias balance pot....

[Enzo]

Then we get into semantics. To me, when I hear "auto-bias" I think of a machine that adjusts itself to changing parameters. That is different from cathode bias or other circuits that generate their own bias - self-bias, in other words.

Having a regulated bias is just that. Whatever current that little circuit is designed to settle at, it will settle at, no matter if one set of tubes would prefer a different current level. SO in my mind there would be nothing automatic about it. It is just fixed bias.

The distinction in my mind would be auto-bias would have to detect something about the tube operation and react to it. ANything else would be some form of fixed bias.

[Steve Conner]

can CCS biasing be considered "auto bias"?
It gives the same mA no matter what right?

Sure, I'd call that auto bias. I'm going to use this example to develop the point I'm trying to make about response to overdrive.

So, we all know that the AC30 is not a class A amp. How do we know? Because the current draw of the output stage rises above its idle value when you really crank it: a sign that it's going into Class-AB operation.

With the above circuit, the current draw of the output stage can't rise, because it's set by the constant current source. You can drive it as hard as you like and it'll still be the same. This means that it'll sound different to plain old cathode bias under overdrive.

Now, this is an extreme example, but ALL auto bias circuits must have this same problem. In a nutshell, when they're used in a Class-AB amp, they can't tell the difference between bias current and signal current. In the Traynor Bassmaster example, it would temporarily bias itself colder when driven hard, almost as if it were cathode biased rather than fixed. In fact that's all the Traynor circuit is: cathode bias with a DC amplifier so the cathode resistor doesn't have to burn up so much power.

The problem isn't unique to tubes, either: you get the exact same issue in Class-AB transistor amps, and that's why we still use the little temperature-sensing rubber diode circuit after all these years. Real automatic bias of a Class-AB amp is still a holy grail after all these years. Sanken Electric made output transistors with the thermal sensing diodes integrated on the chip, which was a great idea, but then they stopped making them for some reason.

I remember the reissue Ampeg SVTs had a bunch of LEDs on the back, and you were supposed to set the bias trimmers so that they all lit green at idle. This is auto bias of a sort, but with a human in the loop. The job of telling bias current from signal current is left to the "meat in the seat", who presumably read the manual and knows that the amp shouldn't be making noise when he's setting the bias.

I think R.G. Keen once suggested that a tiny microprocessor should be put into the amp to handle biasing. It could measure cathode current and set the bias up every time you flipped the standby switch. Basically the Ampeg LED thing with the micro replacing you, and it could distinguish bias from signal by just muting the signal while it was working.