Fixed V Cathode Bias

[TIMBO]

Hi guys, This may have already been covered but doing a search couldn't fine anything.

Tubenit asked why I did a fixed bias rather than a cathode bias on a recent build.

I couldn't really answer that question other than I believed I could get a better sound/performance by having the bias at a HOTTER setting.

I also had a fixed/cathode switch on a Jubilee and I felt that the cathode bias was "holding back" the sound/performance a little.

I was under the impression that biasing power tubes HOT I could get some nice P/Tube breakup.

Fixed V Cathode Bias,what do ya reckon.

[HotBluePlates]

... I believed I could get a better sound/performance by having the bias at a HOTTER setting.
...
Fixed V Cathode Bias,what do ya reckon.

Simple Answer:
Use cathode bias everywhere you can, but use fixed-bias when you can't.

Economic Answer:
What's cheaper - a resistor (& maybe a bypass cap); or a bigger resistor, diode, bigger cap, pot, bias winding/tap, maybe an extra resistor/cap? Old amp designers would use self-bias whenever possible as it was much cheaper to use.

Safety Answer:
The coupling cap to your expensive output tubes shorts, putting the plate voltage of the phase inverter on your output tube grids. With fixed-bias, the tube cheerfully passes a current limited only by the PT and whether a fuse pops before the tube can melt itself and burn up your output and/or power transformer.

An amp with cathode bias in the same situation would tend to have a rising bias with rising tube current, turning off the tube. If the failure is catastrophic and the cathode resistor wasn't over-rated in a silly way, it will burn up and open the circuit which shuts off that tube's current. Only a hard short from plate/screen to ground which is external to the tube will cause damage that can't be corrected by the cathode resistor (this same fault will also kill your fixed-bias amp).

Esoteric Design Answer:
Cathode bias will do the same job as fixed bias until you raise the plate voltage very high, while idling very cool, to operate in deep class AB for very high power output.

  - In class A, all output devices pass current all the time. In theoretical class B, all output devices are biased exactly at cutoff and conduct exactly half the time.

  - Since the cathode resistor creates a bias voltage by having a voltage drop created by tube current flowing through the resistor, it is impossible to cathode-bias a true class B amp because there's zero current so 0v across the cathode resistor, which tries to turn the tube on fully, etc... So a cathode resistor is unable to develop the needed bias voltage when the tube idles.

  - Hot class AB amps that briefly stray into class AB by having one side cut off can use cathode bias. During the class A portion of the swing, the turn-off of one side is offset by the turn-on of the other side, maintaining current at an average current roughly equal to the idle current. The total current averages about double the idle current of one side at all times.

  - When one side shuts off, the remaining side increases its current at a faster rate than before, and makes up for the other side being off briefly. For brief turn-offs, the average current doesn't rise too much above double-idle. This scenario is the same as saying the on-side has smaller peak current.

  - When one side shuts off for a time more-approaching half-off, or using a small load impedance, the peak current rises to very much more than double-idle. If there was a cathode resistor for that side, it would be developing a very large voltage across the the cathode resistance, which would tend to shut the tube off. So a cathode resistor would limit peak current in this situation.

You should see from this that cathode bias is unsuitable for deep class AB operation or low load impedances (which is also a feature of class AB). Fixed bias can be used to bias an output stage to class A, AB or B, but its cost and lack of inherent safe operation traditionally limited its use to cases (deep class AB, class B) where there was no other feasible way to bias the output tubes.

[HotBluePlates]

... I believed I could get a better sound/performance by having the bias at a HOTTER setting.
...
I was under the impression that biasing power tubes HOT I could get some nice P/Tube breakup.

What does fixed-bias have to do with "hotter setting"?

I assume "hotter" means more idle current. You can't get a tube any hotter than grounding the grid reference resistors (as in cathode bias), and sticking 1Ω resistors from cathode to ground. They'll probably melt. So more resistance, a la cathode bias, is actually running the tubes cooler.

As I showed in the "esoteric design answer" above, cathode bias only limits how high the peak tube current can be under driven conditions. But you were probably just referring to the idle condition.

I also had a fixed/cathode switch on a Jubilee and I felt that the cathode bias was "holding back" the sound/performance a little.

How so?

LooseChange has built at least one amp with Fixed/Cathode bias switching. The amp was designed to operate equally-well in either mode. I have the impression that when he set the fixed bias to apply the same bias as the cathode-bias setting, there was no real sonic difference between the settings. In other words, using both the same way without hobbling one, they sound the same because the tubes are operating the same (they don't know if bias is developed by a resistor or a d.c. supply).

But most modern amp companies provide a fixed/cathode bias switch as a "feature" or a to have a "low power mode". If both switch setting sounded the same, it wouldn't be a feature.

So some will take advantage of the fact that if you try to cathode-bias an amp running deep class AB, it will tend to limit the peak current as I explained earlier, which will then limit the peak output power. You may also hear this as "power compression" and "lower output" which reinforces the marketing hype above cathode bias that it causes compression/low power. In reality, only a misuse/misapplication of cathode bias will cause these.

Peeping a 2550 Silver Jubilee schematic, I see a Hi/Lo switch, but not a fixed/cathode bias switch. The Hi/Lo is Pentode/Triode mode; triode mode will seem to have less treble (because of less odd harmonic distortion) and somewhat lower power output (because of the shape of the triode characteristics). If you're used to pentode edge & power, the Lo setting might have seemed less gutsy.

[TIMBO]

HBP, Fantastic explanation and some of it I under stand The jubilee had a fixed/cathode switch not a pentode/triode switch. I may not understood the difference when flicking between the two thinking that if I had a cathode resistor that held an el34 @ 22w when calculated is that 70% biased and using a fixed I can get the el34 to the 70% bias or push it to 100% if I was an idiot.

Maybe another way to look at it, can I cathode bias power tubes to be 100% bias (if 70% is the recommended setting)

 

[HotBluePlates]

Maybe another way to look at it, can I cathode bias power tubes to be 100% bias

You can cathode bias tubes to any percentage of max dissipation, except 0%. That implies no current, so you can't develop any bias voltage across the cathode resistor.

... thinking that if I had a cathode resistor that held an el34 @ 22w when calculated is that 70% biased and using a fixed I can get the el34 to the 70% bias or push it to 100% ...

At this point, you have to point to a specific amp.

For vintage amps, every amp that was class A used cathode bias and idled at/slightly above 100% dissipation (above only because you may not have subtracted out screen current when calculating).

Most 50's class AB amps (and some 60's) used cathode bias. The exceptions were bigger amps with bigger output tubes where the maker was trying to squeeze every last watt out. At the same time that you see fixed-bias appear, you see higher supply voltage appear, cooler idle dissipation (which you can see from Fender schematics in larger bias voltages shown)... There would also be lower load impedance; all of that designed to create an amp that had higher peak output power.

The tradeoff of all the changes to obtain higher peak power is that if you biased hot at idle like you do with class A, the tubes would redplate with a large applied signal. The idle bias had to be made cooler to keep the tubes within their dissipation limits when driven.

... can I ... bias power tubes to be 100% ... (if 70% is the recommended setting)

Can You: Yes, you can do whatever you want to do.

Should You: Depends on the amp. If 70% was the "recommended setting" then it was likely recommended because 100% at idle means melt-down with an applied signal.

In a way, this gets to the heart of why Fender moved away from a bias adjust (which most of the time the end-user wouldn't be fiddling; and in those days, new tubes were dropped in without a re-bias) to a bias balance control based on a fixed bias voltage setting. That is, a proper value was found for the operation of the output stage given the supply voltages and expected drive from the phase inverter for maximum power output.

But that aside, do you have a particular amp in mind, or is this all hypothetical?

If you're trying to get 12-15w from a pair of 6V6's, or 24-30w from a pair of EL34's, cathode bias is probably your best friend. If your supply voltage is 300-375vdc (maybe a bit more, depending on tube type), cathode bias is your friend. If you're trying to squeeze 60w from a pair of EL34's running 500v+ on the plates, you probably want fixed bias to get all the power output possible.

... The jubilee had a fixed/cathode switch not a pentode/triode switch. ...

I dunno, as I only saw one of those in person once and never did more than glance in its direction. The schematic shows a pentode/triode switch (but I might have picked the wrong amp).

[TIMBO]

Sorry HBP , I meant that the first build that I did of the jubilee I had a fixed/cathode sw. and there was a difference when flicked between them thinking that the fixed had a better sound 

[John]

I'll chime in with a thank-you for that explanation as well. So far I've used cathode bias because it's simpler, so less chance of me mucking it up. 

[loogie]

From Randall Aiken:

When do you use fixed biasing instead of cathode biasing?

    Since cathode biasing eliminates the need for a special negative DC bias supply, why don't all amplifiers use cathode biasing?  Well, cathode biasing is not without its faults.  It turns out that in order to keep the DC bias supply voltage at the cathode constant while the input signal is changing, the cathode resistor must be bypassed with a large capacitor.  This capacitor effectively "shorts" the AC signal component to ground, while allowing the DC voltage to remain relatively constant.  If the capacitor is removed, the cathode DC voltage will have a signal voltage superimposed on it, which will subtract from the grid-to-cathode signal voltage, and reduce the gain of the stage.

    The problem comes in when there are large signal level changes, and the average DC level of the cathode voltage changes.  This causes a bias shift, usually in the direction of a colder bias point. This bias shift can be audible, but is sometimes desirable for guitar amp use, as it adds varying harmonic overtones to the sound.   If the bias shift is severe, the tube will go into cutoff, and large amounts of "crossover" distortion will occur.  In addition, the current flow through the cathode resistor generates a necessarily large bias voltage on the cathode for proper tube operation (typically 30-50 volts for most higher power output tubes).  This voltage subtracts from the total plate voltage, which decreases the available output power.  Between this voltage decrease and the bias shift, the output power in cathode biased operation is reduced when compared to fixed bias operation.  Therefore, fixed bias is usually used for higher power amplifiers (50W and higher), and cathode bias is usually used for lower power amplifiers.

Preamp tubes are almost universally cathode biased, because they are used for signal amplification, not power amplification, and the side effects of cathode biasing are not as important.  Also, cathode biasing makes the circuit less dependent on tube parameters, and more forgiving with respect to bias point.  This allows replacement of tubes without having to rebias the amplifier stage.   Output tubes that are cathode-biased should always be checked when replacing tubes, however, because they vary widely in terms of idle current for a given cathode resistor value, and it may be necessary to change the cathode resistor value to return the output stage to it's proper bias current.

http://www.aikenamps.com/WhatIsBiasing.htm

Richard Kuehnel also writes about the differences in his book: Guitar Amplifier Power Amps.  Very good reading

[HotBluePlates]

    The problem comes in when there are large signal level changes, and the average DC level of the cathode voltage changes.  This causes a bias shift, usually in the direction of a colder bias point.

Thanks for pointing this out, as it was a complication I didn't want to dig into much.

Even apart from the use of cathode bias, bias shift (termed "rectification" in the Power Amp section of RDH4) can move the apparent operating point when a signal is applied in a colder or hotter direction, and all output stages suffer from it to some extent. But yes, having a resistor and cap to create and store a d.c. voltage as a result of normal rectification can be a problem in an output stage.

Sorry HBP , I meant that the first build that I did of the jubilee I had a fixed/cathode sw. and there was a difference when flicked between them thinking that the fixed had a better sound 

I see. Well, no telling if you gave each a level playing field. That said, nothing wrong in a preference either.

[jjasilli]

JP Morgan once quipped that a man has two reasons for the things he does, a good reason and the real reason.  Aiken points out a good reason.  But I suspect that for mass manufacturers the "real" reason is marketing: power output in watts.  Why should my 6L6 amp be only 30W when my buddy's is 50W or more!?!  

Fixed bias allows for a higher B+ supply voltage, while yielding a correct idle bias.  W = V²/R, so a higher B+ supply voltage yields a larger nominal output in Watts.  By contrast, cathode bias requires a correct B+ supply voltage per the tube chart, in order for a cathode resistor to do its job. If the supply voltage is high, then correct idle bias cannot be achieved with a cathode resistor. Fixed bias enables manufacturers to fulfill a perceived need of the buying public that "bigger is better" in terms of nominal power output.

[Ed_Chambley]

HBP Commented
Fixed bias can be used to bias an output stage to class A

What amp can you point me to which uses a Class A Push Pull output?  I have heard the old vox ac-15 and 30 are, but I do not think so.  Least not from what I understand to be Class A.

I understand how a SE amp can be class a, just cannot get my head around Push Pull Class A, especially when the comment was made it could be also Fixed Bias.

[jjasilli]

HBP Commented
Fixed bias can be used to bias an output stage to class A
What amp can you point me to which uses a Class A Push Pull output?  I have heard the old vox ac-15 and 30 are, but I do not think so.  Least not from what I understand to be Class A.
I understand how a SE amp can be class a, just cannot get my head around Push Pull Class A, especially when the comment was made it could be also Fixed Bias.

Ed you asked a trick question! How to bias PP tubes for Class A operation is in the tube charts.  See. e.g.: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6L6GC.pdf  But you asked what amp uses Class A Push Pull output.  Statistically speaking it's a safe bet that no production amp does this.  Per my earlier post it just doesn't satisfy the marketing aspect of bigger is better. 

Speaking roughly, a Class A, PP 6L6 amp can put out about 20W.  A Class AB PP 6L6 amp can put out 50W or more.  However, so long as the Class AB PP 6L6 amp is putting out only 20W or less, it will be operating in Class A.  The true Class A amp is always operating at full power and has nothing left to give.  The Class AB amp can have at least another 30W of power at its disposal, but this requires that it crosses over into Class AB territory. 

For a mass manufacturer, as long as big pentodes are being used anyway, they might as well raise the B+ supply voltage and claim big output to satisfy the mass market.  But boutique or DIY builders can design to taste.

[Ed_Chambley]

I see, I just have never stopped long enough as I am usually not looking a voltages that low.

I have no good reason for asking really except I have heard so much that the Vox tone comes from the power section being class A.  The assumption being class A makes the tone.  To complicate this further Wampler pedals is making a new pedal that is supposed to sound like a Class A amp, problem with this is I have a couple of Hifi, class A monoblocks that I am sure would not sound good with guitar.

So is the talk about class A tone coming from an amp is just that, talk?

I also understand the marketing comments, I guess most people think 100 watts is 2 times as loud as 50 watts when it is really closer to 10 watts, but who wants a 10 watt amp, right?  Well I would and I am working on a design that is 10 or 20 watts. 2 or 4 6v6's.

Sorry for the hijack!  Just trying to understand what is folklore and what is truth.

[jjasilli]

2 points: voltage & performance.  Any system tends to perform at its best when at or just approaching its limit.  Picture a pot of boiling water.  Just simmering is not its best.  A rolling boil is peak performance.  More heat and it becomes volatile, sputtering and splashing incoherently.  So a 50W amp simmering at 20W is not its best performance. 

Then there's voltage & resulting waveform.  At higher plate voltages the amplitude of the signal increases.  A sine wave signal input, rounded like the top half of a ball, gets stretched upwards, being made taller.  Its walls approach 90 degrees; its rounded shoulders become sharper, and the top narrower.  It morphs to the shape of a square or triangle wave.  The sine wave generates warm second order harmonics; but the morphed waveform produces harsher third order harmonics.  If it clips, the sine wave generates a soft clip; the morphed wave a harsh clip.  So overdrive sounds different.

[HotBluePlates]

... as long as big pentodes [and push-pull] are being used anyway, they might as well raise the B+ supply voltage ...

JJ hit it on the head.

If you have a big V8 engine that gets 10 miles per gallon, would you be happy if it only made 85 horsepower? If you wanted only 85 HP, you could use a much thriftier 4-banger engine (or maybe even smaller); if you're gonna pay for V8 and lots of gas, it better have some grunt.

Same with guitar amps. There was a post recently about 4x EL84 SE. That's unusual, because low powers are more economical with SE running small tubes, and it's generally cheaper to switch up to push-pull and run class AB if you're gonna pay for more than 1 bottle, 1 socket, etc.

... I have a couple of Hifi, class A monoblocks that I am sure would not sound good with guitar. ...

I can plug my guitar into a console strip and get a decent funk sound. You can probably do the same with your monoblocks. But they're also probably designed for minimal distortion, and you might like some beef on your guitar pickup's sound.

... So is the talk about class A tone coming from an amp ...

What's "class A tone"?

All preamp stages in common guitar amps are class A because few are push-pull stages; their tone could be all over the map, right? You might say phase inverters are "push-pull" but they develop the signal, and occasionally push-pull drivers are used between the PI and output tubes. However, few of those in guitar amps are designed for each side to alternately cut-off, which is the definition of class AB.

[HotBluePlates]

... power section being class A.  ...

What is Class A?

Class A is defined as a device conducting for 360 degrees of the input signal cycle; i.e., it never cuts off.

If the tube never cuts off, then it swings from idle downward a certain amount approaching zero current. You get the most clean power output if the tube swings upward an equal amount towards more current (double-idle), and if the tube idles at 100% dissipation. If the tube swings an equal amount above and below idle, the average tube current is equal to the idle current.
   * You can idle less than 100% dissipation, and swing an equal amount above and below idle, but you'll get less power output.
   * You can swing beyond double-idle current, but average tube current increases along with distortion. If you idle at 100%, the tube redplates as you raise average tube current.

The above never exceeds a power output that is 1/2 of plate dissipation rating, and generally only approaches that.

If you want bigger output power, you need bigger current swings. But we established this will raise average tube current and cause redplating (under driven conditions). You have to let the tube turn off during part of the signal cycle to bring average current and plate dissipation down. Enter Class AB.

Any turn-off of plate current is class AB, but if the tube is off exactly half the time you're operating class B. So that's a lot of potential variation. Most manufacturers eventually pushed for greater outputs by running further towards class B.

If you want more tube current, you need bigger supply voltages. Recalling Ohm's Law, a smaller load impedance will also give bigger current for a given voltage. Amp makers attacked it both ways by lowering load impedance and boosting supply voltage.

Push-Pull Example
Let's say you're running a 6L6 with 400v on the screen/plate. The tube can only pull its plate down to ~40vdc leaving 360v of swing. The data sheet shows that at 0v on the grid the 6L6 will pass 310mA. The voltage change divided by current change is the load impedance. 360v / 310mA = 1161Ω, but this is what one side sees when the other is shut off. The plate to plate load is 4 times this, or ~4600Ω (you'd use a 4kΩ OT).
   * This push-pull setup cannot operate class A because 310mA / 2 = 155mA idle per tube, and 155mA * 400v = 62w per tube.
   * Peak output power is peak voltage swing times peak current, 360v * 0.31A = 111.6w
   * RMS Power is 1/2 peak power, or 55.8w

If you idled this output stage at 70%, that would still be 52.5mA per tube. However, that is much, much less than 1/2 the peak current (as you'd have with a maximum output power class A condition).

The tubes are happy under full output conditions in spite of the big peak current and voltage swings because the total power input to the plate is reduced by the amount of power transferred though the load to the speaker.

Bottom line, idling cooler plus power transfer to the load keeps the tubes from overheating. The cooler idling forced tube current to cut off some portion of the input cycle.

[printer2]

Class A with 6L6G's, a big 15W output.



Class A amp I built



Was in Class A with 70% of max plate dissipation. Increased the dissipation to 100% by lowering the cathode resistor. Basically had the tubes dissipate more but got out pretty much the same power out. Did not notice any sweeter sound. Just for kicks put in a switch to turn off the signal to the upper output tube. Being biased Class A had a sort of SE amp with half the power.

The marketing on the Vox's about being Class A is not quite right as they are run Class AB but a hot AB. Many amp makers now market their amps as Class A even though they are not but rather they are cathode biased. Why do the Vox`s sound like they do? Partly due to the choice of output tube but also because of the choice of phase inverter resistors. Look at them and the Vox and 18 Watt offshoots and compare them to Fender values.

[HotBluePlates]

... Why do the Vox`s sound like they do? Partly due to the choice of output tube but also because ...

No negative feedback around the output stage.

Celestion speaker.

EF86 in some amps, unusual tone circuit in others.

Clean, Vox amps don't sound really any different than any other amp.

[TIMBO]

Thanks for all that info, I looked for my voltage recordings of the jubilee and I only took the readings of the fixed bias with 420v on the plates and 41mA on the cathode so I am unable to tell you what the cathode bias was doing.

Anyway. The FiSonic amp is a cathode bias and as I found this amp was built to a price and this may have been why it is cathode bias, not sure.
Could cathode bias save on parts......
This amp is supposed to be 30w but by my calculations it is spitting out about 40w.
6L6s with plate voltage of 400v, 370 plate to cathode and 32v across 1 ohm resistor and cathode resistor of 270 ohms.
Now here is the tricky bit the OT is only rated at 25w and doesn't get any hotter than warm.

The PT was dead when I bought it and to find any specs where long gone. Thankfully another amp was able to give me the specs I needed and checking the output power I was concerned the I could fry the PT so far all is good 

This may not be what we have been talking about or is it 

[Voxbox]

... Why do the Vox`s sound like they do? Partly due to the choice of output tube but also because ...

No negative feedback around the output stage.

Celestion speaker.

EF86 in some amps, unusual tone circuit in others.

Clean, Vox amps don't sound really any different than any other amp.

Agree with HBP. I would like to add that if you dont have the blue Celestions, it just doesnt sound like a Vox AC30 at all. My friend and I both have 1963 AC30s. He has original blue Celestions and so do I except mine were reconed with non-original cones which was done a very long time ago before Celestion started manufacturing them again.
My friend's amp sounds like an AC30. Mine doesn't! if I play my amp through his speakers then bingo!
Cheers, VB