delta plate minus screen relationships - pentode

[makepeace]

hi guys. there is a similar discussion that was going on a while back on tdpri, but it did not reach a comprehensive conclusion. i thought i would post it in a more discipline specific environment, so that we can get some enlightening collective thought on the subject.

i've read merlin's article about plate/screen relationship (in small signal application) and it speaks of gain and sensitivity being inversely proportional and proportional (respectively) to the difference in plate and screen voltages.

The Valve Wizard

can anybody offer any explanations (however deep or abstract) as to variations of plate and screen voltage and the relationship to gain and sensitivity and the implications of those factors on output, compression and tone?

cheers

[makepeace]

anyone?

[makepeace]

bump

[Austin]

I'll take a stab: So are you trying to make the pentode the most sensitive and efficient you can? Somebody may correct me but I would try for the highest possible plate and screen voltage allowed. If you want compression, I think a larger than normal value resistor connecting the screen supply to the screen and a faily large value grid stopper in series with the input signal to the control grid would do it but is going to make the sound more mushy or spongey and have much less "spank". The "point of clipping" or whatever is not exactly the same as input sensitivity, they way I understand is you will get more power from the same vpp signal with a stiff regulated screen supply rather than simply a high value resistor because the resistor will allow the screen voltage to vary with the plate current and voltage making it a little bit more like a triode. This would be much less of a big deal with a small signal pentode than a power tube, and it might even improve the sound or make it have a more interactive feel since you are not going for hifi I assume. Theres my two cents.

[J M Fahey]

Too old and dull a subject to spark a lively discussion (you already checked that) , but just to set the boundaries :
1) to start, pull "tone" from the question.
Not a parameter, not measurable, no 2 people agree on it, so better leave it aside unless we want to end into a Bar brawl.
Of course, it must have been the core of the discussion in TDPRI , reaching its inevitable end, which you also found:

it did not reach a comprehensive conclusion.

which must be a very polite way to put it.
2)

it speaks of gain and sensitivity being inversely proportional and proportional (respectively) to the difference in plate and screen voltages.

Had to open the link because I thought you had misquoted it, but yes, it says so.
Well, it's not so.
Reducing screen voltage in fact lowers transconductance (so far so good) or Gm, and since we are not touching the load resistor, lowers gain (still going fine), but we can not say it *increases sensitivity* but quite the contrary.
More gain = more sensitivity.
Less gain = less sensitivity.

To be more precise: *sensitivity" tells me of the signal level needed at the input , to get a single output.
It's used all day long describing power amplifiers: "to achieve 100W of power, the amp needs 1000mV at the input"
If the amp has less gain, needing 1500mV input for same output power, it has less sensitivity (i.e. it's "harder to drive")

3) back to your questions:
a) plate voltage has relatively little effect on pentodes, that's why curves are *almost* horizontal or show a small angle with horizontal.
b) screen voltage, on the contrary, makes all the difference, and it modifies heavily not only plate current, but transconductance, which affects gain directly.
And what about *compression* ?
Well, if increasing signal levels (even going into clipping) increases screen current, and because of series screen resistors , screen voltage lowers, gain will lower too.
Which is a classic definition of signal compression.
A lot? ..... a little .... ?
Well, it depends on how much screen voltage varies with signals.
Both extremes:
Music Man amps, with very stiff screen voltages, compress the least.
Similar case with ultralinears, famous for being *loud*, clean and stiff.
The other extreme?: cheap, old amps, with tube rectifiers, very small filter caps, undersized transformers , low power, so they are *always* being overdriven: all those "Pawnshop Specials" which "breathe" and are killer for raw Blues.

[Alan0354]

hi guys. there is a similar discussion that was going on a while back on tdpri, but it did not reach a comprehensive conclusion. i thought i would post it in a more discipline specific environment, so that we can get some enlightening collective thought on the subject.

i've read merlin's article about plate/screen relationship (in small signal application) and it speaks of gain and sensitivity being inversely proportional and proportional (respectively) to the difference in plate and screen voltages.

The Valve Wizard

can anybody offer any explanations (however deep or abstract) as to variations of plate and screen voltage and the relationship to gain and sensitivity and the implications of those factors on output, compression and tone?

cheers

I have a program posted that give the plate characteristics with different screen voltage based on RDH4 doing the conversion. The curve look very similar.

BTW, what is the meaning of "sensitivity"? Sounds like it's just simply gain!!!

About the top two graph where one without screen resistor and one with. Looks like the difference between beam power tube vs pentodes. The pentodes have a softer knee. does that means pentodes has more compressed sound more suitable for blues. This means I should not use 6L6GC as it is beam power and stiffer by nature? Looking at 6V6GTA, the knee is slightly softer.

[Steve Conner]

I have no idea what Merlin is talking about either. In a small-signal context, sensitivity can only mean the same as gain.

In a large-signal context, it might mean the opposite of headroom before clipping.

[Steve Conner]

I have no idea what Merlin is talking about either. In a small-signal context, sensitivity can only mean the same as gain.

In a large-signal context, it might mean the opposite of headroom before clipping.

[Austin]

BTW, what is the meaning of "sensitivity"? Sounds like it's just simply gain!!!

I am pretty sure it is meant in that where you barely touch the guitar and strings and it feels really loud and lively but when you dig in, doesn't get much louder but rather feels like it gets spongey or compressed, changing the tone. It gets you the swelling effect because as a note decays, it gets louder on its own adding to the sustain. From the term "touch sensitive".

[rshipp]

If you do a search on "trainwreck express touch sensitivity", you'll find a lot of links claiming that it has lots of it. Larry Carlton has a video on his website at Mr 335 TV , where he demonstrates what he considers "touch sensitivity" on an old Tweed Deluxe. I've never played those amps, and I don't know enough about either circuit to guess what gives them that reputation. I have a '65 Bassman that can get loud and dirty, but I wouldn't call it "touch sensitive". On the other hand, I also have an early Brownface Vibrolux that can be clean if picked lightly, but has a sweet distortion when I dig in a little (at the control settings). My Danelectro DM25 is that way even more so.

[Alan0354]

If you do a search on "trainwreck express touch sensitivity", you'll find a lot of links claiming that it has lots of it. Larry Carlton has a video on his website at Mr 335 TV , where he demonstrates what he considers "touch sensitivity" on an old Tweed Deluxe. I've never played those amps, and I don't know enough about either circuit to guess what gives them that reputation. I have a '65 Bassman that can get loud and dirty, but I wouldn't call it "touch sensitive". On the other hand, I also have an early Brownface Vibrolux that can be clean if picked lightly, but has a sweet distortion when I dig in a little (at the control settings). My Danelectro DM25 is that way even more so.

So touch sensitive is like clean when soft and break up when hit hard?

[jazbo8]

David Lamkins has a good description => David Lamkins - Guitarist - What is "touch-sensitive"?

Jaz

[J M Fahey]

Basically yes. But remember it's Musician's lingo, not Tech Speak.
And unrelated to what we were discussing earlier.
A goor example of incredible touch sensitivity:

Beware that I attribute most of it to amazing "hand" technique and to this Musician knowing *very* well his tools (guitar and amp).

[Merlinb]

By 'input sensitivity' I mean the maximum input signal before clipping occurs, alternatively called the headroom or grid base (not to be confused with grid bias!). I've stopped using 'sensitivity' now, as it is indeed a bit ambiguous.

So, higher "input sensitivity" means a shorter grid base, less input signal before clipping?

Yep. A 'highly sensitive' valve would be one that can accept only a small input signal for maximum output swing. Bad choice of words on my part.

[Steve Conner]

So, higher "input sensitivity" means a shorter grid base, less input signal before clipping?

The Glen Kuykendall video was rather good. I'm sure I saw him using the volume controls and even stepping on a distortion pedal, so it's not as if he's achieving that whole dynamic range with "touch".