BIAS: is it worth testing it and why?

[tenser75]

i'm a newbie... so be patient!

given: "Transistor Biasing is the process of setting a transistors DC operating voltage or current conditions to the correct level"
I never understood what is this "correctness". how do i know when is correct?? is there a V value I need to know?
is it 4.5v? I mean, do I have to check the voltage at the first transistor is 4.5 (given 9v the half point...is that the point?)

Every pedal i built, most of them fuzzes based on given new & vintage schematics, they seem to work well...even when i replace resistor in the 10% range when i dont have the right ones... so how accurately does the transistor need to be biased? are there huge differences in the final tone?

does the bias affect the sustain?
does the bias affect the harmonics?

[GibsonGM]

Hi "newbie" (ha ha),

Your instincts are pretty right on.  The degree that the bias point will affect a circuit's operation varies, and is very specific to that design.   Many fuzzes ARE sensitive to where the bias is set, and with things like FETs, it can be critical to dial it in (that is why many circuits call for trim pots).  In some ckts, the bias setting DOES affect sustain and the generation of harmonics.

And then, there are others that are not sensitive, and you just go with what has been called for on the schematic. 

Bias is quite often 4.5V, or 1/2 supply voltage - but not always!  Sometimes it is set otherwise for a purpose, such as to make a transistor clip positive or negative peaks, and so on.   

If you are designing a circuit, you would set up what parameters you want the transistor, etc, to operate on over its characteristic curve...where you place the bias point determines the starting point, before you apply signal (quiescence).   

[tenser75]

very helpful thanks!

also: is the bias basically the resistance that goes from 9V to the base... or the collector... or both? how can I bias somethingif i have to test both resistors, i mean it sounds like a pain in the butt..
or is it common practice to test a resistor first then the second one etc..


thansk!!

[PRR]

Audio swings both ways.

Transistors can only swing one way.

The hack is to "bias" the transistor so the no-signal condition is "about half" of the transistor's total possible swing.

Yes, in 9V stuff this is "about 4.5" in many cases.

> based on given new & vintage schematics

Some Designer has *already* thought about the bias. And often thought about what can go wrong if the factory has to substitute different resistors and transistors. "Many" circuits are very robust and will do a right thing even with different (but similar) parts. However a few of our beloved Classics (and a few experimental ideas) can get very off-bias and need some changes.

Mostly, if the plan is good and your copy is good, and no trims are called for, it "should work".

Since we here are not always about "clean" audio, we sometimes find pleasure in "mis-biased" amplifiers. If half-way is 4.5V, and you mis-bias to 2v or 1V, the circuit distorts early and with a distinct color. Many-many more tricks. You can get lost in an infinity of "mis-biases". Go there only if you feel you must.

[bool]

Power amp designers take a lot precaution to get the bias right - because in their world wrong bias usually leads to failure or malfunction (breakdown etc).

In fuzz and other distortion generating circuits, the bias must be "wrong just the right way". If you stick with it (and listen), you will get it eventually -  and become able to measure and set it up with a DMM. If you stick with it some more, you will recognize the pattern of that "just right mis-bias" and become able (at least) to f.e. "design it" with a help of a simulator etc.

[teemuk]

The hack is to "bias" the transistor so the no-signal condition is "about half" of the transistor's total possible swing.

This applies to "Class-A" operation only.

Different class, different quiescent point.

...in 9V stuff this is "about 4.5" in many cases.

"About" is the right term. It assumes voltage swings from 9V to 0V, which it doesn't in practice. There is voltage drop across the device so it never swings to full 9V, and usually some DC voltage potential at emitter, source or cathode, so it never swings to 0V either. So in practice the voltage swing range is different and very much defined by the overall circuitry.


Additionally, class-A bias with "symmetric" voltage swing range of both negative and positive halfwaves is ideal only for "HiFi" applications that hunt for maximum clean headroom and low distortion. Clipping starts late and is largely symmetric, creating odd order distortion and very little (if any) even order.

Many tube guitar preamp stages actually are not biased in such way.

Guitar amps and effects - which really are more EFFECTS than ideal linear amplifiers - do not neccessarily even try to achieve goals such as linearity and symmetric clipping distortion. They may intentionally aim to "offset" bias: Clipping distortion starts later on one half wave, sooner on the other. Distortion is -asymmetric- and harmonic products -both- even and odd. AC coupling will interactively slide bias points of gain stages and create effect of shifting harmonics, which sounds more interesting.

So "biasing" also affects the degree of distortion's symmetry and is often used artistically to create different flavours of distorted tones. "Cold cathode biased" gain stage in the preamp of many high-gain amps (e.g. SLO100) is a very good example of such design. In practice, with usual input signal levels, it acts more like a class-AB stage than an ideal class-A one. And that's the whole point when distortion and harmonics are generated intentionally.

[PRR]

> This applies to "Class-A" operation only.

Quite true.

But you are climbing near the top of the mountain. tenser75's words suggest he has got off the bus and is climbing the field to the first camp site. There's much more for him to discover. But step by step.

We could also note that "about half voltage" is very much a resistance-loaded concept. Is there any other kind of amplifier? Lots. A Champ's transformer-loaded 6V6 will "bias" with plate at full (95%) supply voltage. At that point it "can" carry 90mA, no more (in negative grid). Since we need a both-ways swing (and some other concerns), we bias to 40mA so it can almost swing zero to 80mA. It is "half", but the at-rest condition appears as half current. (Same in R-C coupled but voltage is easier to probe.)

I would say that classes B and C are quite rare in "stomp boxes". Oh, the outputs of chip opamps, but we buy that pre-made, don't directly design it. Some of the nastier fuzzes approach class B, low-low idle current. Of course such things violate a base assumption of "class A B C" teaching: good output with low distortion. Distorters which "gate", cut-off small signals, are often worked in class C. And the level detectors in many limiters work class C (off until output reaches medium-loud). Some of the total-destruction distorters might be analyzed as class D, though usually violating all design concepts of "good class D". (Class D power amps could be applied to battery pedal power amps, but DIY designs are rare now we can buy the chip for a buck.)

Many of these applications are SO far outside 1930s A-B-C teaching that it may be better to ignore "class" and look at what it actually does in the specific circuit.

[bool]

... Every pedal i built, most of them fuzzes ...

99,77% of them in class-A

[antonis]

is the bias basically the resistance that goes from 9V to the base... or the collector... or both? how can I bias somethingif i have to test both resistors, i mean it sounds like a pain in the butt..

The REAL pain in the butt will come after your OPEN circuit "perfect" biasing.. 

(it will appear as a nightmare with many titles, like "Loading", "Impedance matching", "Treble loss", etc..)