[Ericus Rex]

I see it on many of the amps on the market today.  Seems so convenient.  Makes me wonder why all amps aren't self-biasing.  There must be some sort of trade off.  What is it?

[doug s.]

i am not technically savvy, but here's an explanation from someone who is - the mfr of mcalister amps; from his website:

http://www.mcalisteraudio.com/
"...Cathode/ automatic/ self biasing amplifiers use a large resistor impedance between the cathodes of the output tubes to ground. This raises the cathodes above the ground potential in effect creating the required negative grid number one voltage on the output tubes that determines the current flow and class of operation. Amplifiers with this type of bias do not require any adjustments when output tubes are replaced and are less expensive to the manufacture as no additional negative power supply or controls are required- just a cathode resistor and by-pass capacitor. With cathode biased amplifiers the speed and rise times are much slower due to the large value cathode resistor that is in series with the audio output. Further a large value electrolytic cap in parallel with this cathode resistor has to be used for a shunt AC path ground return and impedance reduction. Electrolytic capacitors in the signal path further degrade the sound quality. Fixed negative bias supplies connect the output tube cathode to ground and have an additional negative voltage power supply. The negative voltage is adjustable and fed to the control grid of the output tube that determines the current flow and class of operation. All of my power amplifier designs use a fixed bias voltage that eliminates the cathode resistor and electrolytic by-pass cap producing an amplifier with fast rise times, superior dynamics and transparency. Bias adjustments are required whenever output tubes are replaced- I personally prefer to keep my bias controls inside the amplifier requiring the removal of the bottom cover to adjust the bias. As an option on customer demand I will make the bias adjustments accessible on the top chassis..."
=============

doug s.

[GMuffley]

Operating the output stage of a tube amp with fixed bias produces greater power, and usually less distortion, than with cathode bias. 

[JoshK]

There are trade offs.  Fixed bias means you have to reset the bias every once in a while as tubes age.   But it is usually reported to sound better.  You don't have a big ass cathode bypass cap (usually an electrolytic in commecially built amps) in the AC current path (signal path) with fixed bias, you do inevitably with cathode bias.  Some tubes aren't very stable with fixed bias (usually high gm, low rp tubes like 6C33) and can "runaway" causing destruction of the output circuit.

Cathode bias (aka auto bias, aka self bias) is tube circuit 101, its the generic, plain vanilla circuit.  Fixed bias is usually a step or two up in complexity, usually because of stability more than compexity of circuit.

[Roger A. Modjeski]

Good on Ya all those above on this topic. I agree and appreciate your efforts in getting the right information, a rarity on the internet.

One point to add. Honest (Emission Labs for one) modern tube makers of 45s , 300Bs, and 2A3s freely admit that they cannot hold the bias tolerances nearly as well as RCA, GE, Sylvania, WE and others could in their day. There are lots of reasons for this and I see and agree with what they are up against. Therefore, modern tubes in a cathode bias amp can have quite a range and in some cases run way too cold or way too hot (dissipation wise). Amps with adjustable bias and a way to measure current make these variations unimportant. We use adjustable bias in all our amps except for the EM7-2.5. For those amps we select tubes that fall into a range we like. We make this amp for those who don't want to fool with bias and we take the responsibility of providing selected tubes that hit the right bias point.

[Ericus Rex]

Sounds like "self biasing" is a misleading name.  Makes it sound like the tubes are miraculously biased to the full potential on their own.  Maybe "hodge-podge biasing" or "lump-sum biasing" is more accurate? 

[Roger A. Modjeski]

Sounds like "self biasing" is a misleading name.  Makes it sound like the tubes are miraculously biased to the full potential on their own.  Maybe "hodge-podge biasing" or "lump-sum biasing" is more accurate? 

You are correct, self bias can be +/- 20% and you can't do anything about it other than changing the cathode resistor. Below are the generally accepted terms in all their seeming contradiction. These terms were widely accepted by the early 1920's.

1. "self bias" that the tube biases itself via a cathode resistor and a grid at 0 volts. Considerable power supply watts are wasted across this resistor and it gets HOT. Europeans like to call this "auto bias" which suggests some additional circuit monitoring cathode current but they are referring to the cathode resistor.

2. Fixed bias is actually variable via a bias pot. The cathode is usually returned to ground via a 1 ohm resistor where the current can be measured. In this example a 1 ohm cathode resistor produces 1 millivolt for every milliamp. If the reading is 60 mV it would be better to say "my cathode current is 60 mA". In early amplifiers there was generally no bias pot because the tubes were very consistent and had low transconductance (Gm). The 2A3 is an early high transconductance tube and the first to have mention about circuit design changes (though not till there was some trouble in the field) that should be considered when using this tube. The RCA manual has notes about fixed bias to educate circuit designers about these tubes. Low Gm tubes like the 71 and 45 were not that fussy.

It is important to note that the large cathode resistor is usually bypassed by a capacitor. This combination degenerates the DC gain (bias) but preserves the AC (signal) gain. It also limits power output and pre-maturely clips on sustained loud passages.

With fixed bias the AC and DC gains are the same and equal to the MU of the tube.

[Ericus Rex]

Thank you all for the clear answers.

Roger,
In your experience, what are the sonic drawbacks of cathode biasing?  And why did you choose it for your 2.5 watt amp?

[Roger A. Modjeski]

Ericus,

That's a good question. I see you are sincere in your desire to learn to design amplifiers.

Cathode bias removes the need for the user to do any measuring or adjusting. Many users prefer that and if 2.5 watts is all they need then they should buy the 2.5 watt. It sounds just as good for the first watt and will deliver good peaks beyond 2.5 watts. Depending on the program material it may choke up a bit on loud sustained passages where a fixed bias amp would not.

By adding a bias pot and cathode current measuring jack we get 5 (actually 6.5 watts) out of the same tube. A higher B+ and lower cathode current gives us this power with no additional load on the tube itself. The dissipation is actually lower in the 5 watt version and the tube runs a bit cooler.

In addition to what I have said previously (please read my expanded answer #2) here is a note from my study of early amplifier design as it applied to fixed vs cathode biasing. Although RCA knew what they were doing, they could not anticipate some of the problems that might occur in the field. In the Applications section of the 2A3 data sheet there is a major revision of the paragraph on biasing the 2A3 in Technical Series RC-12 (1934) only a year after it's release in RC-11. Revisions like this do not ocurr often and you generally find the same tube descriptions from year to year and even from GE to RCA. Never again was there such consistancy from maker to maker. BTW, tube-roller-aholics might consider this information when they wax poetic about the differences from maker to maker. As I have often stated, the sound of any given tube in any given circuit is due principally to how that tube will bias-up in that particular circuit.

In addition to operating point variations there are stray effects between individual elements and and tubes with two sections in one bottle. I am doing some research on this currently and will publish my findings when I completely understand this phenomen and how to deal with it. I do believe I am onto an answer of why some tubes sound absoluetly horrible in some circuits. My first experience of this goes back to the circuit in the NY AudioLabs preamp using a 12AX7 in cascode (not a good idea). While that circuit worked with a Tungsram 12AX7 it produced horrible distortion with a Yugo (EI) tube. That was in 1982 and I have been thinking about it ever since. I am interested in hearing from any of you who have these preamps and your experience with the 12AX7 in this circuit. Sorry I can't remember the model number but it was not the NB2.

[Carl V]

Thank you al for the info.
I have cj amps....they have user adjustable bias
pots, with a red LED light.  Is this a viable alternative?

Do some tube types/circuits tolerate changes easier?
i.e., Pentode EL-34, 6550?  PP pr Parrelled SE, SE--variants?

I'm not looking for an amp by amp comparison or bashing.

Just a general thought.

[Roger A. Modjeski]

What are the instructions concerning the LED indication? Every maker is different. Some have the light out for too high and too low and only lit for correct bias +/-  some amount chosen by the circuit designer. I find those hard to deal with.

On the RM-9 I simply put an LED across the cathode resistor with a couple of tricks to make it both accurate and reliable. When it begins to light you are on the cool side. Brighter is higher and it flickers with the music.

[Carl V]

After 30 minutes of warm up
no music or signal present..turn
the pots till the RED just turns on or off a flicker...viola done.
It will pulse with music.  You shouldn't need to adjust very much
or very often.  YMMV.

[Roger A. Modjeski]

Which is it, on or off? There needs to be a clear distinction to know which side of the proper setting you are on. How do you know if you are too high or too low?

[Carl V]

Which is it, on or off? There needs to be a clear distinction to know which side of the proper setting you are on. How do you know if you are too high or too low?

I guess you don't or I don't.

I went back to the manual.
"At idle with no signal adjust it till the light just goes out."

If I recall it was my dealer at the time...many moons ago...
who suggested following the cj instrucitons i.e., till it just goes out,
then his suggestion was to apply the just a slight bit of pressure back
in the reverese direction...not so much as to turn on the red light.

[Roger A. Modjeski]

What happens if you turn the bias all the way down, does the light stay on?

What happens if you go past the light out point on the way up?

Your dealer's term "pressure" does not seem accurate. You can put lots of pressure on a car before it moves at all. This backing off would better be expressed as degrees of rotation.

[Carl V]

What happens if you turn the bias all the way down, does the light stay on?

What happens if you go past the light out point on the way up?

If you turn it all the way down Couter clockwise (past the 'ideal' setting)
the light extinguishes & remains off.  Turning it clockwise past the 'correct'
setting the light remains on.  I've never listened to the amps in either of
these extreme settings.

Your dealer's term "pressure" does not seem accurate. You can put lots of pressure on a car before it moves at all. This backing off would better be expressed as degrees of rotation.

[Roger A. Modjeski]

That sounds like the same system I use. Does the light come on abruptly or slowly being dim at first and getting brighter as you go further?

You would not want to be at either end of the adjustment.

Why not let the light be on as instructed?