Tone Stack Straight Into EL84...

In my preamp, after a voltage gain stage and a cathode follower stage, I have a normal tonestack a la Fender, I'm coming off the Treble wiper, 220K stopper resistor (is that what it's called?) and then through a 1K5 into the EL84 grid.
I have a feeling that this should be wrong as there are usually stages after the stack (PI or another gain stage) to buffer the output. The thing is it sounds pretty darned good up to a certain point, after which it gets a little nasty. I am guessing that the point after which it starts complaining is where grid current begins to flow for a portion of the wave cycle. I have scoped around it and the waveform looks okay going in at the top of the tone stack but pretty wonky coming out.
Any thoughts would be appreciated, and I'll try and scribble a schem for you to see what I'm doing...
Oh and I have a variable B+ on the EL84 for variable power o/p, and I realise that as I ramp the B+ down, at some point grid current will flow, but I just want to tame it a little...



p.s. yes there is power to pin 7 of the EL84, I just neglected to draw it. sorry!

Chef

The circuit looks basically OK, although I would increase the grid leak resistor from 220K to 470K. 220K is a bit too much of a load on the tone stack. 1M would be even better from the perspective of loading on the tone stack, but (IIRC) the grid resistor limit for a self-biased EL84 is 470K.

The 150-Ohm cathode resistor is a bit low unless you're running an extraordinary low B+. What are your plate, screen and cathode voltages?

Increasing the 1.5K grid stopper might alleviate any nasty "blocking" under overdrive conditions.

The response of the tone stack is anything but flat, so don't be too concerned about the fact that the waveshape looks quite a bit different at the output of the treble control. The closest you'll come to a flat response is with treble and bass at minimum and the midrange cranked.

> Tone Stack Straight Into EL84...

You've guessed the basic problem with that. Power tubes are optimized for power, not gain. They have as little gain as they can get away with. Basically this means that a small tube eating nearly the same supply voltage should be just able to drive the power tube to maximum output. Pentodes need a little less drive than triodes, and the EL84 is one of the most sensitive pentodes around. Still, if you put a big lossy attenuator between the EL84 and the tube before it, the preceding tube will clip before the EL84 is maxed-out. And remember that the passive tone-stack does NOT boost; it cuts everything but more or less depending where you set the knobs. The usual settings on the Fender stack are very un-flat, but assume it has 20dB or 10:1 average loss. Assume it takes 15V peak grid drive to max-out. That requires 150V peak at the tone-stack input. Assuming a driver triode needs a B+ about 5 times higher than the peak audio, you need a 750V B+ at the driver. You probably have 300V. The driver is giving up before the power tube does. Using these sample numbers, the EL84 is making about 16% of maximum power when the driver distortion gets bad, and maybe 1/4 full power when the driver is whacking its rails. If you crank the Fender bass and treble, it may not be that bad, but still the driver is not giving the EL84 the full smack it needs. And if the EL84 is running at "5 Watt" conditions, there are cheaper ways to get a 1 Watt amp.

Since you can make quite a racket with a Watt in a good speaker, I have no doubt it sounds darned good, until the drummer arrives.

Modifying the Fender stack for less average attenuation reduces the amount of "boost" possible at "10", which is not a good tradeoff. There are simpler less-lossy tone schemes, but they lack the proven Fender sound-range. Building a huge driver gets insane and costly real fast, and if you are insane enough you could be smoking pot (CentralLab, not Jamaican). Leo put a driver between the fancy tone-stack and the output grid for a reason, not just to get excess 12AX7s out of the way.

The point between driver and output is also not the best place for a gain control, but if you go for two gain controls this is where the second one goes, for the same reason it is a bad place for the tone control. Reducing gain here lets you clip the driver before the output reaches full power, allowing a fat tone without huge volumes.

> a cathode follower stage

Cathode follower????? There is no place for a cathode follower in a small guitar amp. It has no gain, it has no distortion, and any impedance issues can be patched-up some simpler way. It is just dead meat. A CF might be useful for a direct out or FX loop in a big amp, but not a small amp. Lose it or give it real work to do, and see if you feel better.

[quote author="NewYorkDave"]The circuit looks basically OK, although I would increase the grid leak resistor from 220K to 470K. 220K is a bit too much of a load on the tone stack. 1M would be even better from the perspective of loading on the tone stack, but (IIRC) the grid resistor limit for a self-biased EL84 is 470K.[/quote]

Yep I've been messing around with that grid leak resistor and it is surprising how much it affects the sound...Or perhaps it isn't that suprising!

[quote author="NewYorkDave"]The 150-Ohm cathode resistor is a bit low unless you're running an extraordinary low B+. What are your plate, screen and cathode voltages?[/quote]

Va=243V
Vg2=255V (?) eh?
Vk=7.8V

Not quite sure how the screen voltage is up there but I'll do some investigation.

[quote author="NewYorkDave"]Increasing the 1.5K grid stopper might alleviate any nasty "blocking" under overdrive conditions.[/quote]

Will investigate this too!

[quote author="NewYorkDave"]The response of the tone stack is anything but flat, so don't be too concerned about the fact that the waveshape looks quite a bit different at the output of the treble control. The closest you'll come to a flat response is with treble and bass at minimum and the midrange cranked. [/quote]

Gotcha!

I'll report back later...

chef

[quote author="PRR"]> Tone Stack Straight Into EL84...

You've guessed the basic problem with that. Power tubes are optimized for power, not gain. They have as little gain as they can get away with. Basically this means that a small tube eating nearly the same supply voltage should be just able to drive the power tube to maximum output. Pentodes need a little less drive than triodes, and the EL84 is one of the most sensitive pentodes around. Still, if you put a big lossy attenuator between the EL84 and the tube before it, the preceding tube will clip before the EL84 is maxed-out. [/quote]

Yep I guessed it. I kind of knew but wasn't sure why, so I'm thankful that i can come here and hear from NYD and yourself!

[quote author="PRR"]... it has 20dB or 10:1 average loss. Assume it takes 15V peak grid drive to max-out. That requires 150V peak at the tone-stack input. Assuming a driver triode needs a B+ about 5 times higher than the peak audio, you need a 750V B+ at the driver. You probably have 300V. The driver is giving up before the power tube does. [/quote]

D'oh! My head resists accepting what my intuition wants me to know.
I feared this but thought "Hey it sounds good up to this point so what's the problem?"
If I can get more loud out of it then I can challenge my drummer to a duel knowing I can beat him with my 100dB efficiency 1 x 12.

[quote author="PRR"]Using these sample numbers, the EL84 is making about 16% of maximum power when the driver distortion gets bad... there are cheaper ways to get a 1 Watt amp. [/quote]

Granted. As I said if I'm losing out on volume...goodbye c-follower.


[quote author="PRR"] Leo put a driver between the fancy tone-stack and the output grid for a reason, not just to get excess 12AX7s out of the way.[/quote]

Ha ha. Yep, head-in-sand, so I'll lash it up and see how I go.

[quote author="PRR"]The point between driver and output is also not the best place for a gain control, but if you go for two gain controls this is where the second one goes, for the same reason it is a bad place for the tone control. Reducing gain here lets you clip the driver before the output reaches full power, allowing a fat tone without huge volumes. [/quote]

Yep. My schematic perusal has drawn that conclusion, and I may yet go for a master volume, but that is a little way off yet.

[quote author="PRR"]Cathode follower????? There is no place for a cathode follower in a small guitar amp. It has no gain, it has no distortion, and any impedance issues can be patched-up some simpler way. It is just dead meat. A CF might be useful for a direct out or FX loop in a big amp, but not a small amp. Lose it or give it real work to do, and see if you feel better. [/quote]

the reason I went for that was because one of my favourite amps has a
c-f before the stack and I really liked the sound and 'Touch' of it so thought I'd suck-it-and-see. I'll dump the CF and see if the 'feel' is still as good.

I'll be back later with a report.

chef

Sorry for the hiatus!
Well, I lost the CF and now have an input stage which is a paralleled 12ax7 with a bypassed 1K5 on the cathode and a 220k plate resistor through a cap into the tonestack, through a volume pot and into the final generic 1k5 byp + 100k plate res voltage gain stage.
To begin with I had just the volume pot after the i/p stage which then fed another voltage amp+cf, and then into the tonestack, and then the EL84. I have to say that the amp does sound smoother in distortion now although it might have lost a tiny bit of 'touch', whatever that means...
My issue now is what to do with that other 1/2 12ax7? Should I run it as the dreaded CF (I can't handle any more volt gain!!) to seperate the stack and the volume pot?
I know you're going to wonder 'why the paralleled input stage?', well I have tried it with just the single and I think the double sounds a bit meatier. As I understand it, there will be more current drive (about twice that of a single side?) and I assume the o/p impedence drops by a factor of 2. Am I on the right track? IF not then why does it sound different?

Cheers

chef

Here's a good write up on grid resistors.

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

There's quite a bit of good stuff on that site.

The cost-effective design would have just the one 12AX7: input stage, and a buffer/booster after the tone circuit. Save a pound of weight and many dollars cost (including heater transformer capacity). You do have to be careful pulling a gain of over 2,000 (50*50) in one socket.

The doubled-up input stage makes musical sense to me. I might even go a lot lower on the cathode resistor: 1K5 sounds a bit high for 2 sections 12AX7. Of course flavor is more important than theory. Whatever works.

Then logically, if you are stuck with a spare triode, double it up with the buffer/booster after the tone stack. If nothing else, bad socket contacts won't stop the show, since at least one of the paralleled triodes probably has all 3 pins working. You get a little more drive to the output tube, which may be good. You get twice the input capacitance, which may load the high-Z tone stack and volume pot. You may need a slightly larger treble cap and brite cap, like 330pFd instead of 220pFd.

[quote author="PRR"]The cost-effective design would have just the one 12AX7: input stage, and a buffer/booster after the tone circuit. Save a pound of weight and many dollars cost (including heater transformer capacity). You do have to be careful pulling a gain of over 2,000 (50*50) in one socket. [/quote]

Yeah, the more I delve into this stuff, the more I realise how far Leo went to discover just where he needed to be on the curve of diminishing returns. I know he has his detractors who say it was all just app. notes frome valve manufacturers, but he clearly got down to fighting the devil in detail to a very fine level, and surely anyone has to admire that!

[quote author="PRR"]The doubled-up input stage makes musical sense to me. I might even go a lot lower on the cathode resistor: 1K5 sounds a bit high for 2 sections 12AX7. Of course flavor is more important than theory. Whatever works. [/quote]

For the duration of this project I have made sure that where possible individual components could be compared by switching to see what kind of effect, if any, swapping different values/ types of components has. I rigged one half of that input bottle with a switch between it's cathode and resistor/ cap to that I could effectively 'turn off' one side, and while I'm sure this is not the same as disconnecting everything, it was good enough to establish that I preffered it with both sides in. Likewise I had a play with the cathode resistor values and by ear, pretty much ended up back at 1k5. None of the other values sounded bad but with the 1k5 it sounded a little 'chunkier'.

[quote author="PRR"]Then logically, if you are stuck with a spare triode, double it up with the buffer/booster after the tone stack. If nothing else, bad socket contacts won't stop the show, since at least one of the paralleled triodes probably has all 3 pins working. You get a little more drive to the output tube, which may be good. You get twice the input capacitance, which may load the high-Z tone stack and volume pot. You may need a slightly larger treble cap and brite cap, like 330pFd instead of 220pFd. [/quote]

Yep again, I have ended up paralleling the buffer stage and again, the sound got a little meatier, and has the added bonus, as you said, of stopping socket crud turning into a show stopper!
Leo would probably turn in his grave at the recklessness of my valve usage, but I can afford the crampons to hike a little bit further up that curve. :thumb:

chef

[quote author="Swedish Chef"]I have a feeling that this should be wrong as there are usually stages after the stack (PI or another gain stage) to buffer the output.

Chef[/quote]
I thought to remember that the old H&K Crunch Master box had it's tonestack also right in front of the EL84 but I just checked it and it wasn't so.

So FWIW: http://www.amptone.com/images/schem_crunch.gif

Without having read this thread completely:

Have a look at the P1 and HiOctane at ax84.com, these amps run the tonestack directly into the EL84, and the EL84 definitely gets hit hard enough (i built both). The Marshall tonestack has 10..15db loss driven from a ECC83 plate, so there is still plenty drive for a EL84 to clip.

I also dare to disagree with PRR, the dc coupled CF in the std marshall pre does clip. It can't follow the previous stage's anode to full B+...

;Matthias

EZ81 and CR:
thank sfor both of those comments.
All the food for thought that I can get is greatly appreciated!
I'll 'Save' the schems and study them when I get a chance.

Cheers

chef

Since you guys are talking tone stacks, I need some opinions! I built hybrid guitar preamp 2 years ago. It's got an IC input which drives a 2 stage tube section for flavor. I modded it and added two sweepable eq's (low/mid 150 to 2.4kHz and mid/high 800 to 10kHz). The eq's are peak type +/- 15dB cut/boost and work very well. But...they're more like a console eq than an amp tone control. I can easily add the tone stack between tube stages with minimal changes (I need to change the load on the stack)

Would this be a good idea to add? I find the existing eq is "specific" (and useful ) where the tone stacks are broad but I'm not sure if adding this is a waste of time or not.

Opinions welcome!

Thanks,
Jeff

Electric guitar needs strong EQ to sound normal. Naked, it is all midrange; you gotta sag the mids so the low and high will come out. There are at least three main schools: Fender, Marshall, and Gibson. Fender is strong, Marshall lets more mid-bass through, and Gibson used a twin-T notch applied here and there in various ways.

The logical thing is to use the Fender stack: it has proved itself. For big-balls crunch, a minor change (and the cathode follower I forgot) makes it Marshall. Keep the console-like EQ for further flavor, but try to use it very little. With a nice Fender or Marshall stack and good speakers, you would almost never want/need more than a few dB more/less than the guitar stack gives you. (And adding big boost on top of the bass-treble boost of a guitar stack is begging for trouble.)

I am totally in awe of all these explanations, but at 200 watts flat out after the drummer arrives, with even a tiny amount of distortion, I just wonder if in the end it actually sounds any different.
Stephen

I think I need to "doo the math" on the Fender tone stack as the PRR comment rings so true in my practical experience. On the little 7watter that I'm fooling around with at the mo' I started with just a "High" roll off and it sounded very honky, and not very appealing at all, but as soon as I began to elaborate the tone circuitry the sound started to become much more pleasing. After messing about with quite a number of different combinations I ended up with the Leo Classic. It makes sense really, as he had a few units/ years to get it right. It's just so flexible, and yes, perhaps there are other stacks that are better for specific tasks but as a jack-of-all-trades it does VERY well for my tastes.

Stephen: My 7W single EL84 through a 1x12 Celestion Blue (100dB sens) stands up pretty well to my drummer and yes, the tone does make a diffference ;-)

sc

Well, sounds like the tone stack is in on my preamp! Darn easy to do so why not.

For those of you considering building a low wattage simple amp....don't hesitate. I have played through a number of amps but can't seem to beat the sound of my 1955 Fender Deluxe. There is no substitue for the classic tone of these amps. Since you can get these tones at low volume, they're great for recording and I agree, they stand up nicely in a small room or venue even with drums and bass!

Regards,
Jeff

Roger:

I'm not really at liberty to post a full schem as my boss wants to sell these, but if you read the whole thread you should be able to piece together the whole thing.
If you want to query any values pm me but it's all pretty much in the thread :green:
Don't tell my boss...

At the moment I'm running Westlabs 6 watt tx set but have a few other options lined up which are a little closer to home...

sc

Roger:

I'm not really at liberty to post a full schem as my boss wants to sell these, but if you read the whole thread you should be able to piece together the whole thing.
If you want to query any values pm me but it's all pretty much in the thread :green:
Don't tell my boss...

At the moment I'm running Westlabs 6 watt tx set but have a few other options lined up which are a little closer to home...

they're great for recording and I agree, they stand up nicely in a small room or venue even with drums and bass!

Click to expand...

Soooo true, Jeff :thumb:

sc

[quote author="RogerFoote"]

I bought stuff to clone the Champ, but I do like the EL84 sound. It would be cool to have a blend between the Champ and Vox amps.[/quote]

A session player friend who has used a Champ for years just made the first pre-order for the little amp I've come up with. Side by side with the Champ which is an early 70s model, the Champ sounds like it's being played under a cardboard box, while the new one sounds crisp and clear.l I'm not trying to 'big myself up' rather am encouraging you to persue your goal; it's really satisfying...as you already know! :green:

sc