biasing preamp tubes?

[daz]

i built a 18 watt and the 4 stage preamp is cascaded with 2 volumes and a master and has a good bit of gain. I have a bit of a issue with the stages in V1 tho. It's a bit bright and buzzy when the first volume is cranked. as it's turned down it goes away at about 12:00, but then of course so does the gain. I don't mind losing some gain, but not quite that much. So I thought maybe the cathode and plate resistors might need to be tweaked so that the tube is biased correctly, but i know nothing about doing this on a 12ax7. The plate is 100k and the cathode is 1k with a 1uf bypass. How low can i go on the plate resistor? i tried larger cathode resistors but they tend to cause it to sound a tad dead by the time i try one large enough to kill the buzz.

so i'd really like to know how to tell when a preamp tube's bias is withing ok specs, plus whats the difference in effect of changing cathode resistors to reduce gain as opposed to changing plate resistors to do the same?

[tbryanh]

Please post a schematic or a link to one. Thanks

[cbarrow7625]

Sounds like you may need more voltage swing in the cathode and/or plate circuits. Distortion in generated in a few ways in a tube preamp, the ugliest sounding of which is plain old clipping of the DC rails. Good, smooth harmonic distortion created by compression & soft clipping of the signal is the most musical, desired sound. "Fizzy" distortion is often a sign that you are clipping the DC rails and/or just driving one of the stages further down the signal path too hard (and causing it to clip when you don't want it to). The highest gain distortions require the clipping but it is done in such a way that the tube is driven into full on square wave clipping and tons of harmonics are generated (including odd order). When you are just getting into clipping you get increasing amounts of odd order harmonics that can be out of balance with the sound you are going for.

Counterintuitively, you may actually need to increase the values of the plate resistor & the cathode resistor to increase the voltage swing before clipping. THis will also result in more gain in that particular stage. Depending on what you have downstream, this may cause you other problems.

What happens when you turn down the second volume control (which I assume is between stage 2 & 3)while the first one is cranked? Does the bright & buzzy sound go away? If not, I suppose that you have localized the problem to the 2nd stage. That's good! Tweak away on the second stage and see what you get.

As far as biasing is concerned, there is not much you can do to hurt the tube. For a 12AX7 plate resistors as small as 27k are used sometimes to reduce gain (Marshall SLX preamps) and plate resistors as high as 470k get used regularly in many circuits. The most common values lie between 100k-220k (all along the range). Try differnet values and see which ones work in your circuit.

For the cathode circuit, a lower value resitor is going to give the sound more of a "hard feel" and "attack". Higher values will make the sound mushier. For the second stage in a hi gain preamp, I think 1k is probably too low a value (typically) and to get the sound you want you may have to comprmise a little bit on the "feel". I would recommend trying 1.5-1.8k to start. Those values will feel a little "softer" than the 1k but not so much that it just feels like mush. Anything much higher than about 3.3k will usually begin to feel ovely soft but those higher values can be used for "effect" to actually create more non-linear distortion in that tube stage. The 10k Marshall cathode resistor & the world famous 39k cathode resistor found in the SLO100 (and all of it's wannabe knock offs) are used for just that. The high cathode resistor values greatly reduce the gain of the tube stage but the increase dramatically the distortion created withing the tube itself (not a bad trick really). The trade-of is the reduced "attack" and "feel" of these kind of circuits.

Hope that's enough to keep you tweaking. A schematic would be helpful if you want to dive into more specifics.

Chris

[tbryanh]

The highest gain distortions require the clipping but it is done in such a way that the tube is driven into full on square wave clipping and tons of harmonics are generated (including odd order).

Transistors can provide full on square wave clipping too. Why do you think tubes are still used for this?

[daz]

Great stuff, thank you !!!! Really, thats some of the best detail i've found so far. I tried your suggestion of a higher (3.3k) cathode R on the second stage and it did indeed seem to soften up the tone, which was another issue i had that was really bothering me. However, now theres too much gain and i'm wondering what you would recommend to lower ? The plate is already 100k. I guess the smart thing to do would be lower the stage 1 gain because i'm sure it's stage 2 thats being hit too hard. reaosn being is that when i crank the first volume i hear blocking distortion starting to happen. But with stage 1 down a bit, even with the seond volume cranked the blocking isn't there anymore. So the 3.3k worked great on stage 2 to soften the tone, (even thinking to go higher because it's no where near mushy) but i now need to figure out how to lessen the gain in the first stage. But with the plate at 100 and cathode at 1k all i can do is go up from there, not down? correct? And so how do i reduce gain? Should i use a fairly large grid R into the second stage?

EDIT: I also just noticed that the softness i gained seems only to affect the high gain tone. But when i turn the guitar volume down to clean up, the bridge p/u sounds very muddy and dull and hard still. Yet the single coils are totally opposite when i turn down. On other amps usually they are both affected the same way.

Sounds like you may need more voltage swing in the cathode and/or plate circuits. Distortion in generated in a few ways in a tube preamp, the ugliest sounding of which is plain old clipping of the DC rails. Good, smooth harmonic distortion created by compression & soft clipping of the signal is the most musical, desired sound. "Fizzy" distortion is often a sign that you are clipping the DC rails and/or just driving one of the stages further down the signal path too hard (and causing it to clip when you don't want it to). The highest gain distortions require the clipping but it is done in such a way that the tube is driven into full on square wave clipping and tons of harmonics are generated (including odd order). When you are just getting into clipping you get increasing amounts of odd order harmonics that can be out of balance with the sound you are going for.

Counterintuitively, you may actually need to increase the values of the plate resistor & the cathode resistor to increase the voltage swing before clipping. THis will also result in more gain in that particular stage. Depending on what you have downstream, this may cause you other problems.

What happens when you turn down the second volume control (which I assume is between stage 2 & 3)while the first one is cranked? Does the bright & buzzy sound go away? If not, I suppose that you have localized the problem to the 2nd stage. That's good! Tweak away on the second stage and see what you get.

As far as biasing is concerned, there is not much you can do to hurt the tube. For a 12AX7 plate resistors as small as 27k are used sometimes to reduce gain (Marshall SLX preamps) and plate resistors as high as 470k get used regularly in many circuits. The most common values lie between 100k-220k (all along the range). Try differnet values and see which ones work in your circuit.

For the cathode circuit, a lower value resitor is going to give the sound more of a "hard feel" and "attack". Higher values will make the sound mushier. For the second stage in a hi gain preamp, I think 1k is probably too low a value (typically) and to get the sound you want you may have to comprmise a little bit on the "feel". I would recommend trying 1.5-1.8k to start. Those values will feel a little "softer" than the 1k but not so much that it just feels like mush. Anything much higher than about 3.3k will usually begin to feel ovely soft but those higher values can be used for "effect" to actually create more non-linear distortion in that tube stage. The 10k Marshall cathode resistor & the world famous 39k cathode resistor found in the SLO100 (and all of it's wannabe knock offs) are used for just that. The high cathode resistor values greatly reduce the gain of the tube stage but the increase dramatically the distortion created withing the tube itself (not a bad trick really). The trade-of is the reduced "attack" and "feel" of these kind of circuits.

Hope that's enough to keep you tweaking. A schematic would be helpful if you want to dive into more specifics.

Chris

[cbarrow7625]

Yeah, they can but its still a different transfer function. Transistors tend to accentuate odd, higher order harmonics. Tubes still clip softer even when hitting the rails so they tend to accentuate lower order harmonics and the even harmonics are ususally higher in amplitude than the odd harmonics.

Its also a matter of voltages too. If you could run transistors on 250-300V rails, I'm sure the clipping would be much "softer". 100 VAC peak signal hitting 80 volts of headroom clips a lot softer than 100V peak hitting 15 volts of headroom in a solid state circuit.

[cbarrow7625]

If you have blocking distortion going on, chances are that your coupling cap going into the second stage is too large and/or the grid resistor is too large. What matters with blocking distortion is just the time constant formed by the cap/resistor combination.

A neat trick for this is to put a lower value resistor (like 100-250k) from the wiper to ground on the volume pot. That way, when the volume is turned all the way up you reduce the time constant because the impedance of the pot seen by the coupling cap changes from 1M to 100-250k (depends on the resistor you use, of course). This will also reduce the bass repsonse of the signal when fully cranked (something you typically want anyway to avoid farty-ness). When the volume pot is turned down, the cap sees a larger resistance and more bass is passed through the cap/resistor combo.

Now, if that doesn't work you will need to add a voltage divider before the stage (not just a series resistor). A series resistor into a tube grid does not change signal level. you need to put something like a 470k series resistor + a 470k resistor to ground following it into the grid of the stage. That will reduce the level going in.

Iwould recommend keeping the first stage plate resistor at 100k but increasing the cathode resistor to 1.5k-1.8k (maybe even up to 2.7k if you want a Marshally kind of grittier sound.)

Increasing the cathode resistor up to about 3.3k will usually end up giving you maximum gain for the stage. Any higher than that and the gain starts to go down again. You can use that to your advantage.

If you are not fond of the sound of the resistor on the wiper of the volume pot then just reduce the value of the coupling cap between stage 1&2. That will have a similar effect but it may get too thin sounding as you turn down the volume.

Keep us updated.

Chris

[daz]

Ok, i tried the voltage divider and wasn't happy with the resulting loss of gain. It dropped way too much. however, i may try a 1 M for the grid>ground R and see how that does. The pot resistor i already tried. The coupling cap would cut too mych high end because i just finally got it to where it's got just enough highs as is. So the quest goes on. But your first advice has already helped me a lot so your help is appriciated in any case. thanks. I'll keep tweaking and let you know if i find the magic bullet.

EDIT: i tried a 470k at stage 2's grid, tho w/o the ground resistor. i tried this before with a 100k and for some reason i didn't like it. this time it seemed to work quite well tho. maybe placebo effect? Who knows, but it seems good now. I'll just have to play with it a while to be sure. but don't get me wrong..the amp never sounded bad. It's as Nigel says, "nitpicking really". But seriously, it's been a killer tone since i cascaded the stages. Just trying to perfect it is all.

[cbarrow7625]

Cool. Sounds good. Yeah, I suppose the 470k series resistor reduced the current enough to the tube to stop the blocking distortion. I forgot about that effect. Glad to hear it worked.

I know the feeling though...it's all about the tweak. The final tweak & the next "final" tweak, and the next one....... I have a 5 channel amp with 12AU7's as output tubes that I have been "tweaking" on fo about 5 years now. I just need to wring out that last little fraction of perfection and I'll be done...no, really.......

[daz]

I know the feeling though...it's all about the tweak. The final tweak & the next "final" tweak, and the next one....... I have a 5 channel amp with 12AU7's as output tubes that I have been "tweaking" on fo about 5 years now. I just need to wring out that last little fraction of perfection and I'll be done...no, really......

LOL!!! So true and funny. I tell myself i want to get it nearly perfect and be done with it, but i fear that may never happen. On a serious note. being my first amp i am quickly learning that tweaks that accomplish what you wanted may after all not be quite so great as you 1st thought. After a day i'm now wondering whether i like that resistor there ! It got rid of that blocking distortion, which by the way was just beginning to happen at 10 but would go away by turning the 1st gain down a bit. But it also left me with a less elastic feel to the string....a bit less sag i think, which was the first *symptom* i would get as i turned the 1st volume up even before any of the nastiness began. So maybe it was GOOD after all and i just needed to use a smaller pot to limit the output. Or maybe that 470 just needs to be smaller, or maybe or maybe or maybe.... LOL. crazy stuff man, but fun !

On a serious note, at this point there is really only one thing i desperatly need to put the amp squarely in the "best i've owned" category. Thats would be the midrange. The cathode resistors you suggested really helped. But it's still there to a degree, and theres something else that i think is probably caused by the same thing thats causing the hard mids. that would be the tome you get when using a strat with high gain and playing single notes up around the 10th fret. That fat hollow sound that sounds sorta flute like. Thats just not there at all, and it's a sound i use and love a lot. I don't know if it could have something to do with the output section, but i have tweaked the pre ad nausium so i wonder if it does. I've had other EL 84 amps that do it so it's not the tube type. I'm certain tho that if i could get that tone the hardness would be gone and visa versa.

[cbarrow7625]

I hear ya on the strat thing. It's difficult. "The Tweak" is why I have 5 channels. It's really hard to make asingle channel do everything you want.

With 4 gain stages you end up with a fairly symmetrical clipping. Each stage basically clip hard on half of the wave. The negative swinging part of the wave can only swing down to the cathode voltage. The positive going half of the wave can swing up until it hits the DC anode voltage. The second stage only provides the beginnings of a soft clip. The 3rd & 4th provide a hard, symmetrical clipping of the full wave.

A 3-stage preamp will have a tendancy to give a more assymetrical cliiping. That is where that real nice strat tone often lies. It's the assymetry of the clip that can make a strat sound like its full & singing. Not having a 4th stage to completely clobber the signal and turn it into full on high harmonic squished hi gain glory helps too.

It's hard to get that great strat sound with that kind of setup.

For your other issues, have you tried just putting a 100k-470k resistor in series between the coupling cap & top of the volume control? It sounds like you may just need to knock down the signal a little bit to get rid of the nastiness. Remember to try combinations of all of this stuff too. Quite often subtle changes in many differnet locations (rather than one big change in one place) can make all of the difference ingetting the "glory tone" and missing it altogether. Tweaking is all about stringing together a bunch of subtle changes to get that perfection squeezed out.

[daz]

Well maybe i should try eliminating a stage, tho then i won't have the amount of drive i'd like to have. that would be fine for gigging, as i don't use near as much as the amp has and i always get more if i do need it with a clean boost.(not to mention with the output stage driving at stage level the gain would have to go WAY down anyways) But since i don't play out anymore with rare exception, it's a home amp and the gain is more important to me there because the low volumes i need at home, the only decent sound i get are gain tones.

So i suppose i could try it, tho i'm not quite sure i want to unless i could be sure the tone i'm looking for lies there. But i'll take it into consideration. I'll try the resistor between the pot input and cap too. Thanks.

[daz]

I just realized it was a simple move of a couple wires to remove a stage, and i did so with no sucess as to that sound. Sounds the same but less gain. So i'll give that resistor on the pot a try.

[daz]

I just realized it was a simple move of a couple wires to remove a stage, and i did so with no success as to that sound. Sounds the same but less gain. I just that resistor on the pot a try to and no go. But i'll try anything else you can think of, so if you have any more suggestions, the iron's still hot.

[Jag]

After a day i'm now wondering whether i like that resistor there ! It got rid of that blocking distortion, which by the way was just beginning to happen at 10 but would go away by turning the 1st gain down a bit. But it also left me with a less elastic feel to the string....a bit less sag i think, which was the first *symptom* i would get as i turned the 1st volume up even before any of the nastiness began. So maybe it was GOOD after all and i just needed to use a smaller pot to limit the output. Or maybe that 470 just needs to be smaller, or maybe or maybe or maybe....

For your other issues, have you tried just putting a 100k-470k resistor in series between the coupling cap & top of the volume control?

+1 on what cbarrow7625 said. I've got an SE amp I built that was too nasty above about 8. I put a 220K in series with the coupling cap and the top of the 1M volume control and it sounds great. I think, in your case, 470k is just a bit too big. Just out of curiosity, what is the value of your volume control?

I hear you on the "day after the mod blues". No matter how good a change sounds to me, I always have this little devil on my shoulder saying "are you sure you like that" . It's the same feeling I get the day after I buy a new guitar. The little devil says, "you've got three Strats, a couple Tele's, a PRS, and a bunch of acoustic instruments, do you really need that 335?".

It's an addiction!