Using cross-over distortion to set bias

[Enzo]

And what is not to like about a socket adaptor for monitoring tube current?

if that is a sin, learn the shunt method.


This is a guitaramp, as was said above, there is no "accurate" setting, any more than ther is an accurate amount of salt to shake on your steak.

If the thing doesn;t sound like crap on the cold end, and the tubes are not red plating on the hot end, then the bias is adequately set.

[J M Fahey]

I use the scope because itīs the quickest method which gives me info on *this* particular set of tubes Iīm using.
Going to any "kitchen recipe" value, by definition, has been arrived at using *other*, earlier tubes.
The ones I just installed may (will) be different, unless I buy pre-tested/matched/graded ones.
What if these are not ?
Or just from a drawer with no further indication?
I set bias cold enough so as to show a definite kink, and then raise current until it dissappears.
Go back and forth a couple times (5 seconds total time) until I like what I see.
If Iīm suffering a *real* OCD attack that day, I lift NFB which otherwise tends to hide it.
Only after that , I measure idle current (DC dissipation) so as to ensure I'm not silly overheating my tubes.
All adjustments at no more than 50% of full power out.
Adjusting for no kink when overdriving is meaningless because grid rectification wreaks havoc with bias voltages, itīs an entirely different problem.
"Crossover distortion" is a *low signal level* problem, which kills sustain, works as a farty noise gate, you name it.

When the amp is overdriven itīs a very different game.
This is personal opinion: waveform kinking when overdriving (because of grid rectification) is part of the tube sound, I for one donīt want to kill it.
Peavey Transtubes even generate it on purpose! To better have a "Tube flavor".
Agree.

jm2c

[Gtr_tech]

Bias FAQ

[Jazz P Bass]

Bump.
Notch.
Same thing.
What the crossover artifact is not, and this is what matters, is a linear transition.

[Chuck H]

Adjusting for no kink when overdriving is meaningless because grid rectification wreaks havoc with bias voltages, itīs an entirely different problem.When the amp is overdriven itīs a very different game.

I've never been able to adjust for NO kink. Even minimizing the kink, you need to pay attention not to have the tubes overdissapating at idle. I have a hard time with "meaningless". Yes the grids are current starved and shifting during clipping, but if this is the conditions under which the amp will be used "I" feel the need to maximize the tubes behavior under those conditions. The crossover kink does continue to increase as the waveform continues to flatten. And increasing current does minimize it to some degree. Saying that adjusting bias for these connditions is "meaningless" implies that there is no affect. And that's not the truth.

This is personal opinion: waveform kinking when overdriving (because of grid rectification) is part of the tube sound, I for one donīt want to kill it.

I dig that. I've tried some experiments to create desirable "swirl" without much luck so far. When it sounds bad, it sounds bad. OT, specific tubes, speaker,etc... May all play a roll in why some amps sound good swirling and others sound like a busted speaker. And when it sounds bad I want to kill it. I'm not inclined to leave it as part of the tone for any nostalgic reasons. No more than I would try to heal myself with leaches because it was traditional.

[hasserl]

I've never been able to adjust for NO kink. Even minimizing the kink, you need to pay attention not to have the tubes overdissapating at idle.

^^ This

Ahh, The Tone Lounge.
Bar none, the best tech site on the net.

Tone Lizard For The Win!

A few years back there was some quite heated discussions about the Tone Lizard around this place (the old Ampage forum). Ken Gilbert and the Tone Lizard had some flaming arguments.

Technical Articles on Tube Amps | ken-gilbert.com

The Big Ass Guitar Amplifier | ken-gilbert.com

[J M Fahey]

Thatīs what I meant by "meaningless".
Dunno, maybe somebody can suggest a better word?
I mean itīs not realistic trying to kill that "kink".
After all, itīs "built-in" , inherent to tubes.
What we call "crossover" in fact is a result of the rather important non-linearity in the transfer curves.
Itīs in the characteristic curves published.
Itīs part of that Tube personality, like it or not, and has always been, since it was created what ... 50/60 years ago?
Maybe *we* are obsessing now, about something that was there all the time, and nobody worried about before.
Specially Leo or Jim

[JoeM]

If the amp is used cplipping the power tubes hard and you want to minimise crossover distortion then you would need to have the amp clipping hard when you adjust the bias for minimum crossover distortion. But of course you then need to check the idle current since you don't want the tubes overdissapating (or nearly) at idle and also working as hard as they can when conducting. In fact, on an AB1 amp with moderate to high plate voltage, if you attempt to bias out all crossover distortion under hard clipping you will likely end up overdissapating the tubes at idle. You need to find a safe and acceptible compromise. There are still other facets to it that are also relative to playing style. Like trying to get that classic metal/hard rock sound. Biasing for more current can make the amp feel softer and less dynamic which is bad for this style of playing. So you bias for a little less current and end up with a little more crossover distortion but better dynamics. My point was just that WRT guitar amps I think the tubes should be biased for the players style and not to any arbitrary %age or for maximum clean output (unless that IS the players style). These aren't reference amplifiers or home theatre systems and they aren't used like those amplifiers are. Guitar amplifiers are used as instruments (or at least signal processors) and I think they require a different criteria. Sometimes that means shorter tube life. Not always. Guys who work with tool steel go through more files than guys who work with mild steel. And guitar amps often go through more tubes than home stereo amps. Smoke 'em if you got 'em.

OK, I found the answer on the Aiken site which helps. Here's what Randall Aiken says about trying to 'bias out' the crossover notch:

"When driven into the positive grid region at clipping, the output tube grid acts as a forward biased diode and clamps the positive peaks of the grid waveform to a point slightly above the cathode voltage. As the input signal level is increased, the clamping action forces the average value of the grid waveform downward, effectively increasing the average negative grid bias. This results in more crossover distortion, even if the amp is biased higher into class AB. Because of this clamping effect, the amount of crossover distortion that you are trying to "bias out" will change depending upon how far into clipping you set the grid drive. If you keep trying to eliminate the "notch", you will bias the amplifier too hot, and your tubes will be destroyed."

[Joe L]

Randy is IMO slightly wrong in his description. What is actually happening is the bias point is shifting downward due to the positive charging of the PI coupling capacitor when grid current is occurring. And that is important because if you can drain the capacitor on each down stroke of the input wave, it won't shift the bias downward. This is just what the "Paul Ruby Mod" does by using zener diodes to drain off the positive charge when the outputs are cutoff.

I have implemented this in an 18watt Marshall which is cathode biased and it made a huge tonal difference. As soon as I can clear my "todo list", I am going to research how well it works on a fixed bias Marshall I need to perform 20 year maintenance on.

As far as crossover distortion improving the sound of an amp, I don't agree. It isn't harmonically related at all to the fundamental frequency and is just noise.

[Chuck H]

Randy is IMO slightly wrong in his description. What is actually happening is the bias point is shifting downward due to the positive charging of the PI coupling capacitor when grid current is occurring.

That's actually what he said. He just didn't say it was due to charging the coupling caps. And if you want to get technical, which Randall is prone to do, the specific cause is the time constant. The caps can't discharge fast enough in any of the typical guitar amp PI circuits to allow for small journies in and out of grid current.

And that is important because if you can drain the capacitor on each down stroke of the input wave, it won't shift the bias downward. This is just what the "Paul Ruby Mod" does by using zener diodes to drain off the positive charge when the outputs are cutoff.

Great mod. The only thing I don't like about it is that I didn't think of it first. Paul should have patented that one.

I have implemented this in an 18watt Marshall which is cathode biased and it made a huge tonal difference. As soon as I can clear my "todo list", I am going to research how well it works on a fixed bias Marshall I need to perform 20 year maintenance on.

As far as crossover distortion improving the sound of an amp, I don't agree. It isn't harmonically related at all to the fundamental frequency and is just noise.

That is just an opinion (which you did concede). In some circumstances crossover distortion is part of the formula for "swirl". Which can sound very dynamic. Example: If you listen to the Komet clips of Mississippi Queen and compare the Komet to the actual TWreck you'll hear that the TWreck has more phase shift to the note bloom. The Komet sounds great, but kind of flat and bland by comparison. I think Ken's ability to "voice" the amps has to do with him learning some specifics about his circuit that allow him some control over this elusive property. He's mentioned things like "on the edge of unstable" and "moving wires" for tone shaping. If you listen to any TWreck clip from any Ken voiced unit you'll hear this property. And you'll also hear a good amount of crossover distortion. But for some reason it sounds good with these amps. The same can be said of MANY vintage Marshall clips. My point is... Don't throw the baby out with the bath water. I do my best to eliminate crossover distortion because I don't know how to make it work for me instead of against me. But I haven't given up trying.

[Joe L]

That is just an opinion (which you did concede). In some circumstances crossover distortion is part of the formula for "swirl". Which can sound very dynamic. Example: If you listen to the Komet clips of Mississippi Queen and compare the Komet to the actual TWreck you'll hear that the TWreck has more phase shift to the note bloom. The Komet sounds great, but kind of flat and bland by comparison. I think Ken's ability to "voice" the amps has to do with him learning some specifics about his circuit that allow him some control over this elusive property. He's mentioned things like "on the edge of unstable" and "moving wires" for tone shaping. If you listen to any TWreck clip from any Ken voiced unit you'll hear this property. And you'll also hear a good amount of crossover distortion. But for some reason it sounds good with these amps. The same can be said of MANY vintage Marshall clips. My point is... Don't throw the baby out with the bath water. I do my best to eliminate crossover distortion because I don't know how to make it work for me instead of against me. But I haven't given up trying.

Very good points Chuck. I only have one amp - the 18 watt Marshall, that I have had to deal with too much crossover distortion. And as I mentioned, I used zeners on both the grids and cathodes and could disable them by switches. I let the guy that I was building the amp for chose which position to leave the switches in and he chose the zeners. But EL84s are pretty complex sounding and without feedback, even set clean the amp sounds rich.

And another problem I am fighting is hearing loss. I played an Ampeg V4 with full stack Marshall cabinets in the early 70's with the channels jumpered and dimed and I'm paying for it now. 10K is the upper limit in my left ear and 13k in my right. I don't trust what I hear like I used to.

[Chuck H]

I"ve only used the PR mod with el84's. I use it in combination with a zener across the cathode resistor that "fixes" the bias as the tubes start to clip. My amps also use a little NFB to tighten the bottom end. I also use a big chunk of iron for the OT. I also use a very small shunt filter on the OT primary to control some of the spikey-ness that happens when el84's are pushed hard. The combination sounds very big for a pair of little bottles.

I played through 100W heads all the time when I was in bands and such. But being 43 I was more in the half stack era rather than the full stack. Not having the speakers right at ear level has helped preserve my hearing.

And to be frank my name surely isn't Doug

[g1]

And to be frank my name surely isn't Doug

Um... ok Frank
I was wondering about that, I've seen J. Frondelli calling Enzo Doug and was starting to wonder "how many guys here are really named Doug?"

[Enzo]

I let Frondelli get away with it because them Italian guys from New York scare me...

[kg]

my ears are burning. :lol:

i've never used an oscope for adjusting fixed bias. lots of other, also important stuff--but not that.