Fixed vs Cathode Biasing

[plexi50]

I know this is an old topic but i need a refresher course.
Is there a difference in tone between say a cathode biased Deluxe reverb and a fixed biased DR?

Or Plexi jtm45 for that matter? Like the November amp layout that has the option to switch between fixed or cathode biasing

[tubenit]

Yes, there is a significant difference in tone between cathode and fixed biased.  Fixed biased tends to be louder & it has a crisper and cleaner tone than the cathode biased which I tend to think of as a softer and perhaps more compressed tone.

Think of the difference between a Deluxe Reverb (or black face Deluxe) and a 5E3 Deluxe (tweed).

With respect, Tubenit

[HotBluePlates]

Fixed biasing makes it easier to run the tubes with much higher plate voltage, while idling at much less current. That is, deeper into class AB. Cathode biasing the same tube into deep AB doesn't work too well, because it's harder to turn the tube more-off with a resistor after you reach a certain point. You might use a technically too-small bypass cap along with cathode bias in an amp trying to push deep into class AB; that will cause a difference in distortion and compression, but it's almost more about a misuse of cathode bias.

As a result, I think a lot of the difference is a comparison of two amps with the different biasing methods which have too-different operating characteristics to make a fair comparison. Think about it: no one ever complained their preamp was squishy and compressed, yet almost every guitar preamp is 100% cathode biased.

Ask LooseChange how different the sound is between fixed/cathode biased in his Standard. I think I recall him saying that the sonic difference was minimal; I'm guessing that's because there wasn't a big shift in the operating point when going back and forth between the biasing modes.

I'd suggest you shouldn't pick biasing method based on tone differences. The old-school rule of thumb was use cathode biasing eveywhere you could, because it helps prevent catastrophic failure of the tube. If the tube draws too much current, bias voltage increases and turns the tube more-off. If there's a short, the cathode resistor will burn up before something expensive.

Fixed-bias will cheerfully let your tube burn up, along with taking out an output or power transformer.

I suppose you could use servo-biasing to build safety into fixed-bias, so that it could sense and counteract an overcurrent situation. But you'd have to have a very compelling reason to go through all that trouble to use a bunch of extra circuitry instead of a single resistor.

[jjasilli]

I installed a Fixed/Cathode bias SW in my el84 Traynor Bassmate, used for regular guitar, anticipating great tonal changes.  Honestly, I can't hear a difference.  As Hotblue says, the tonal difference may be plate voltage, especially in the bigger pentodes which can be operated over a wider voltage range.  Usually fixed bias is used together with hi plate voltage; cathode bias with lower voltages.  The different operating points of the tubes affects tone.  This makes it hard to tease-out whether tonal differences derive from the operating point, or from the different bias methods.

[plexi50]

Excellent reading and explanations. I have a cathode biased 2x 6V6 Mark Huss plexi 1987 build that sounds exactly as you said tubenit. (softer and perhaps more compressed tone) Thats what has been bugging me about it. It just is too soft and doesnt have a loud fullness about the amp. Sounds like it is being held back and trying to get off a leash but cant quit make it. To compressed and bland. Great explanation HBP. I'll get around to it this week and make it fixed bias. I took the quick way out to cathode bias it and now i know i cant do that with just any amp. My ear will report back this week

[tubenit]

I have a cathode biased 2x 6V6 Mark Huss plexi 1987 build that sounds exactly as you said tubenit. (softer and perhaps more compressed tone)

IF you are using a tube rectifier, try a 5AR4/GZ34  .............. or a solid state plug in rectifier like Hoffman sells.

Also, you can increase the cathode cap value to 220uf which tightens it up and simulates more like fixed bias.

It's interesting that you say your Mark Huss 6V6 plexi is cathode biased?  The schematic on his website shows his 6V6 Plexi  to be fixed biased.

With respect, Tubenit

[LooseChange]

Not much difference in tone. I have a "Standard" with with a switch for bias. The two modes operate the  tubes at about the same level. The tone is about the same. I get sense that the amp is a bit more sensitive in the Cathode mode. The Standard uses a 470ohm resistor (any power tube) with no bypass cap. KOC says the cap makes Cathode bias sound more like Fixed so it's not included.

[plexi50]

I have a cathode biased 2x 6V6 Mark Huss plexi 1987 build that sounds exactly as you said tubenit. (softer and perhaps more compressed tone)

IF you are using a tube rectifier, try a 5AR4/GZ34  .............. or a solid state plug in rectifier like Hoffman sells.

Also, you can increase the cathode cap value to 220uf which tightens it up and simulates more like fixed bias.

It's interesting that you say your Mark Huss 6V6 plexi is cathode biased?  The schematic on his website shows his 6V6 Plexi  to be fixed biased.

With respect, Tubenit

I am using a 5AR4 and also have the SS recto plug. Ive been bouncing back and forth with them. I used the Mark Huss schematic but deviated from the bias circuit and used cathode r and cap.

LC do you have the Standard schematic or where i can find it?

[seedlings]

I just built a 6L6 PP amp cathode bias.  It's biased at about 35V at idle, but when the strings are strummed the increased current moves the voltage to 45V.  So, the load line shifts.  (For fun I tried a 1000uF cap across the 330R || 33uF to see if it could be pseudo-fixed-bias, and it didn't change the voltage drop at all.)  I recently played a 1963 Fender Bandmaster AB763, and the compression is obvious in my build compared to his, and lends a Vox-ish character.

CHAD

*edit: Just hit me that my build also doesn't have a NFB, so this may in-and-of-itself account for tonal differences.

[HotBluePlates]

... Great explanation HBP. ...

I actually dropped the ball, twice. Called LooseChange by the wrong name (before edit), and also forget to prove why deep class AB is hard to do with a cathode resistor.

So I think we understand class A is easy to cathode bias, because the average tube current doesn't change much. Let's go to the other extreme: textbook-definition class B.

According to the definition, in class B the tube passes current for exactly 50% of the input signal cycle (class A is 100%, class AB is >50 but <100%). To do this exactly in accordance with the definition, you bias you tube right at the point plate current cuts off. That way, the whole positive signal cycle causes plate current to flow, while the tube is cut off for the negative half-cycle. Obviously, the other side of the push-pull output picks up the slack.

But Voltage = Resistance * Current. If there is no tube current, there is no bias voltage across the cathode resistor.

"But wait!" you say. "I could use a really huge resistor to bias the tube on just a little and be close enough, right?"

Sure, and that's how tubes are generally biased to class B, but not with a resistor. The problem is the bias-shift Seedlings mentioned. This isn't a big deal in class A, where a theoretical perfect device has the exact same idle and average current. Real devices have increased screen current under driven conditions, as well as an apparent increase of d.c. due to tube distortion. As you progress from class A to class B, the ratio of average signal current to idle current keeps increasing; theoretical class B would go from 0mA idle to few-hundred mA peak.

You'd need a big cap to fake fixed-bias with your BIG cathode resistor to make our class B stage. What happens when a BIG transient hits? The BIG cap is need to source the big current for that transient (to effectively bypass the big resistor). Once the current is drained, other high-level signals cause the resistor to have a bigger voltage drop, turning the tube more-off.

If low-level signals follow, the tube stays cut off. That's because the big cap value and big resistor lead to a long time constant, so it takes longer for the cap to settle back to the proper bias voltage. At an extreme, you could have each side of the push-pull cut off except for the very peaks of the input signal, so the majority of the input is chopped out. This would be "crossover distortion" to a pretty extreme level; the output would be biased into class C (works great for RF transmitters, not good for audio, which is probably why most readers won't know what it is).

So that's an extreme limit case, but it serves to explain why cathode bias works great for class A, can work good for stages straying lightly into class AB, but isn't used for deep class AB (or B) stages trying to get big output power from the tubes.

[plexi50]

I finished changing to fixed bias last night and it adjusted right up on the bias trimmer pot. I just played this morning and there is a big difference in clarity to me. The muddness is gone and the amp is more tonally responsive to the ear. Who's ear?
An old fog horn leg horn ear. Big lesson learned. I cant say that i have run into this exact problem before using a cathode biasing on other builds or noticed it as much as with this amp build. As far as plexi circuits go though i wouldnt use a cathode biasing again. I think it makes the amps tone too bland. In my case.