Single Ended EL84 tube amp plate voltage thoughts...

[Mike Sulzer]

With this statement it should be observed that the knee frequency for a given cap value will change with the resistance (impedance) value. The higher the circuit impedance, the more LF a given cap value will pass. I don't know the formula for this because I'm not very tech at all. Someone may chime in about that. I use a lot of on line calculators when I design. ampbooks is a good one. Play with the resistance values for the "bypass cap calculator" and you'll see this effect in action.

An EL84 has a transconductance of something like 11.3 mA/V. The inverse of that, 88.5 ohms, is the impedance looking into the cathode. The bypass capacitor sees this in parallel with the bias resistor. With 150 ohms, this is 56 ohms. Changing the resistor does not do as much for the low frequency "knee" as you might think since the input of the tube has a greater effect since it is smaller, but the effect is there. You just have too compute the parallel resistance to see how big it is.

[Chuck H]

Thanks Mike. I wasn't aware that Ck saw the Rk in parallel with the tube itself so this is pretty important. Or not, as you noted. I mostly offered the brief tutorial as it applied to the preamp cathode circuits kevtronic mentioned. I guess in that case a 12ax7's internal resistance (60k+?) can be neglected with something like a 820R or 1.5k cathode resistor but would still be an important consideration for things like partial bypass cathode followers (like some effects loops) or so called "cold clipping" stages with 10k to 39k cathode resistors. Good info.

[Mick Bailey]

If you're looking to maximise your clean sound you may wish to experiment using a local feedback loop as per the RH84 hi-fi amp. I've built three amps using this technique (for Hi-Fi) and the extra clean headroom is considerable, as well as an improvement in linearity. To some degree it will also tame your excessive drive level to your power tube.

[eschertron]

If you're looking to maximise your clean sound you may wish to experiment using a local feedback loop as per the RH84 hi-fi amp. I've built three amps using this technique (for Hi-Fi) and the extra clean headroom is considerable, as well as an improvement in linearity. To some degree it will also tame your excessive drive level to your power tube.

Local feedback gets a few 'pluses' in an application like this. Extra headroom as mentioned, and better high- and low- end frequency response. That for about the same component count to add dividers between the stages.

[pdf64]

Thanks Mike. I wasn't aware that Ck saw the Rk in parallel with the tube itself so this is pretty important. Or not, as you noted. I mostly offered the brief tutorial as it applied to the preamp cathode circuits kevtronic mentioned. I guess in that case a 12ax7's internal resistance (60k+?) can be neglected with something like a 820R or 1.5k cathode resistor but would still be an important consideration for things like partial bypass cathode followers (like some effects loops) or so called "cold clipping" stages with 10k to 39k cathode resistors. Good info.

12AX7 internal cathode impedance seems to be a bit under 2k, see Designing Common-Cathode Triode Amplifiers especially appendix B (give it a few goes, try not to get too put off by the math).
It does seem to line up with my measurements of the output impedance at the cathode.

[pdf64]

...the two main reasons I originally started the post was whether I should stick with 250 volts on the plates (specs) or push beyond like others I have read about. Once that is done and set I can move on to the rest type thing. It does seem acceptable to many to push the plates 300-350 on an SE El84, so I think I will go that route to see what happens...

Bear in mind that pretty much any realistic SE operating point is class A, so any point that sounds acceptable and doesn't overly stress the tube may be seen to be ok.
And that class A is defined by plate current cut off under signal, not directly related to plate dissipation at idle.

It may be that idling at 95% with 250V HT is hotter than centre bias, and if so, the under biasing would be reducing the clean output power to some degree.

I find nick b's interactive valve data sheets page very helpful Interactive Valve Data Sheets

[Chuck H]

12AX7 internal cathode impedance seems to be a bit under 2k, see Designing Common-Cathode Triode Amplifiers especially appendix B (give it a few goes, try not to get too put off by the math).
It does seem to line up with my measurements of the output impedance at the cathode.

Thanks Pete. I thought it was the more typically referenced tubes internal resistance being considered since that's the commonly referenced spec. Here's a paste from the link for the sake of the post:

With a plate resistor of 100K, and internal plate resistance of 62.5K and a mu of 100, the 12AX7 will have a cathode impedance of (62.5K+100K)/(100+1) = 1609 ohms. This impedance is then in parallel with the actual cathode resistor, so if you used an 820 ohm cathode resistor, the actual cathode resistance would be 1609||820 = 543 ohms.

[kevtronic]

The cathode bias resistor will also alter how much signal the gain stage can accurately amplify before clipping AND how much it amplifies it AND (this is something you'll want to know) how that stage clips when it's overdriven. ... You ask good questions. ... These are strictly an experienced layman's observations and understanding. Maybe some of the more technically inclined will step in and touch up my response.

Loads of info in there Chuck H, I will have to re-read it a couple times for sure. I kind of got side tracked working on another amp beefing up the tremolo a bit. But, back to busting gain stages. Thanks for taking the time to respond like you did. Oh, yes, 100K not 1K on the plate load resistor (3K/100K).

*** edit starts ***
Ok, I've read down to here. Great stuff. I can see I am a green-horn in the land of pre-amp stages
I also think it's going to take more than a couple days to get a good grasp at my next level of understanding on this subject...
You've guys have provided a wealth of bunny-trails to take and buckets of trail-mix to chew on - lol...

I'll be spending more time here:
Amp Books: https://www.ampbooks.com/
Thanks Chuck H

And I'll be reading Designing Common-Cathode Triode Amplifiers:
http://www.aikenamps.com/index.php/d...ode-amplifiers
Thanks pdf64

By the way, just in case someone isn't familiar with Aiken Amps website it has a wealth of information on it:
http://www.aikenamps.com

MUSIC ELECTRONICS FORUM - ask and you will receive, and much, much more too
Had to put a forum coodles in there, man...

[Mike Sulzer]

Thanks Pete. I thought it was the more typically referenced tubes internal resistance being considered since that's the commonly referenced spec. Here's a paste from the link for the sake of the post:

With a plate resistor of 100K, and internal plate resistance of 62.5K and a mu of 100, the 12AX7 will have a cathode impedance of (62.5K+100K)/(100+1) = 1609 ohms. This impedance is then in parallel with the actual cathode resistor, so if you used an 820 ohm cathode resistor, the actual cathode resistance would be 1609||820 = 543 ohms.

(62.5K+100K)/(100+1) <== (Rp + Rl)/(Mu + 1) = (Rp + RL)/(gm*Rp + 1).

Notice how the limiting cases work out. If we consider a pentode, Rp is much larger than Rl, and so we ignore Rl. gm*Rp is much larger than one, so we ignore one. Then we have 1/gm for the pentode.

Now consider a triode used as a cathode follower. Rl is zero, and if we ignore one relative to 100 we also have 1/gm, or about 625 ohms.

[Chuck H]

Thanks Mike. And thanks again Pete. Another question though? If the cathode impedance is acting on the level of NFB as well, if it does? As I would think is case with an unbypassed Rk. Then how is Ck accountable to the parallel Rk and cathode impedance? Wouldn't it be more like two series impedances with only one bypassed making Ck accountable only to Rk WRT it's knee frequency?

[pdf64]

Yes, I've pondered that one too.
As the cathode resistor and the internal cathode impedance do seem to be in parallel, the cathode seems to magically know where its 0V reference is.

[olddawg]

Honestly if you want a good functioning single ended EL84 amp with only a volume control, look at the Epiphone Valve Jr schematic. (Or buy one for $85 off of CL and put fresh tubes in it). As my mom always said, "You can't make a silk purse out of a sow's ear!" There is only so much clean overhead there no matter what you do. I used one for years. Using a compressor really helps the clean volume. The amp overdrives nicely with a TS type pedal. An efficient 12" speaker will do wonders for the low end. I preferred an open back cab. You can play games with circuit machinations all day long. Imho, at the end of the day you are just reinventing the wheel. Maybe one that is no longer round. If it's too gainy, try a different 12A_7 tube. Harp players do that a lot.

[Mike Sulzer]

Thanks Mike. And thanks again Pete. Another question though? If the cathode impedance is acting on the level of NFB as well, if it does? As I would think is case with an unbypassed Rk. Then how is Ck accountable to the parallel Rk and cathode impedance? Wouldn't it be more like two series impedances with only one bypassed making Ck accountable only to Rk WRT it's knee frequency?

Suppose we turn it on, let it warm up, but apply no ac input voltage. Measure the voltage V to ground at the parallel combination of the cathode resistor, the bypass capacitor and the tube cathode. Suppose that someone sneaks in and somehow quickly deposits enough charge on the capacitor to raise the voltage by Vr. What determines the time constant for the voltage to return to V? Current flows both into the tube and the resistor, and so it is the parallel combination that determines the "RC" time constant. The "RC" time constant and the frequency response are related by the usual equation.

[Chuck H]

That helps. I'm going to run that through my knob awhile because it adds/modifies my perception of tube function. Thank you.

[Malcolm Irving]

... Current flows both into the tube and the resistor ....

To expand on that a bit (to clarify – I hope ):

Initially, there is a DC conventional current flowing down through the tube and down through the cathode resistor to ground.

After some (positive) charge has been hypothetically added to the top (cathode end) of the bypass cap, the cathode voltage becomes more positive by Vr.
This increase in voltage at the cathode increases current through the cathode resistor to ground, and at the same time reduces the current coming down through the tube (since Vgk has become more negative).

This is the same thing as imposing additional currents, ‘down’ through the resistor and ‘up’ through the tube, added to the quiescent current.