Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Jonas on March 05, 2023, 12:56:49 pm
-
I'm doing some Sunday snooping around tube data spec sheets and pulled up the EF86. I was surprised to see the amplification factor was only 38. I thought the EF86 was supposed to have higher gain compared to the 12AX7 (amplification factor = 100). What am I missing?
-
> amplification factor was only 38
As a Triode, right?
Pentode 'amplification factor' is so high, and so not real world, that they don't cite it anymore, not since the very first ones.
-
Might have been good to say what datasheet you snooped.
Philips 1970: https://frank.pocnet.net/sheets/010/e/EF86.pdf
Mu(g2G1) means "with G2 in control", which is triode connection.
-
Yes, I see that now - I was indeed looking at the Philips data, specifically the quick reference data at the top.
Thanks!!!
-
Ciao PRR
.... they don't cite it anymore, not since the very first ones
Have you a link to one of those very old datascheet ?
Thanks
Franco
-
> Have you a link to one
-
Cool PRR
amplification factor from 360 till 1050 :grin:
Many Thanks
Franco
-
I'm doing some Sunday snooping around tube data spec sheets and pulled up the EF86. I was surprised to see the amplification factor was only 38. I thought the EF86 was supposed to have higher gain compared to the 12AX7 (amplification factor = 100). What am I missing?
According to wiki it is 185. If that is the truth, is hard to know?
https://en.m.wikipedia.org/wiki/EF86 (https://en.m.wikipedia.org/wiki/EF86)
/Leevi
-
... the EF86. ... the amplification factor was only 38. ...
According to wiki it is 185. ...
https://en.m.wikipedia.org/wiki/EF86 (https://en.m.wikipedia.org/wiki/EF86)
This is a perfect example of why Wikipedia is not allowed to be used as a reference in academic work. After all, you could go change that right now to " Amplification Factor 1,004,578,985" if you wanted.
The "gain of 185" for the cited condition comes from the 4th column of the top section on Page 3 of this EF86 data sheet (https://frank.pocnet.net/sheets/010/e/EF86.pdf). And it is "gain" because it is "volts out / volts in."
This same data sheet provides (triode connected) Amplification Factor as 38, because that value is fixed by the relative spacing of the cathode, grid and screen.
Notice that all "Voltage gain" figures on the 3 triode-connected sections on Page 3 show values well below 38. That's because in-circuit gain is always smaller than "Amplification Factor" or "Mu". The plate load resistor (RL) and the tube's internal plate resistance (ra) form a voltage divider such that we only exhibit a gain close to Mu when the external plate load is near-∞Ω:
Av = (Mu x RL) / (RL + ra)
The reason I point this out is triode curves are a diagonal line moving upward, while pentode curves are a near-horizontal line. For large changes of plate voltage the EF86's plate current barely changes which indicates a high internal plate resistance. Page 7 of this EF86 data sheet (https://frank.pocnet.net/sheets/030/e/EF86.pdf) shows a 2.5v gridline where plate current is 2mA at 100v plate, and increases to only a bit over 2.1mA at 400v:
Resistance = Volts / Current = (400v - 100v) / (2.1mA - 2mA) = 300v / 0.0001A = 3MΩ
Now, try plugging in the Wikipedia Condition's "Amplification Factor" of 185, RL of 220kΩ and the example ra of 3MΩ:
Gain = (mu x RL) / (RL + ra)
Gain = (185 x 220kΩ) / (220kΩ + 3MΩ) = 40.7M / 3.22M = 12.6 ---> less than a 12AU7
The math doesn't work anymore, because the internal plate resistance is so high. So if we go backwards and estimate "Mu" from the observed gain of "185" what do we get?
Gain = (mu x RL) / (RL + ra)
185 = (Mu x 220kΩ) / (220kΩ + 3MΩ)
185 x (220kΩ + 3MΩ) = Mu x 220kΩ
Mu = [185 x (220kΩ + 3MΩ)] / 220kΩ
Mu = 2707 (!)
But where Mu is nearly constant in triodes, the implied Mu changes with every condition in pentodes, which is why it is never cited, and why we use a different formula to estimate pentode gain (Gm x RL).
-
> amplification factor was only 38
As a Triode, right?
Pentode 'amplification factor' is so high, and so not real world, that they don't cite it anymore, not since the very first ones.
Mu is 33 in triode mode, so unlikely Mu is 38 wired as grounded cathode amp strapped as a triode.
--Pete
-
Mu is 33 in triode mode, so unlikely Mu is 38 wired as ...
It is curious that the GEC and Mazda data sheets (https://frank.pocnet.net/sheets/020/e/EF86.pdf) mention "33" and have both figures (38 and 33). Meanwhile, no Mullard, Philips or Telefunken data sheets have that "33" figure.