Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: shooter on February 10, 2017, 07:16:59 pm
-
From reading here and others, when you lower G2 volts it has the result of increasing gain by allowing a smaller G1 signal to get the tube to “max” with less effort.
So if Vout/Vin = Vgain, lowering Vin increases Vgain. So does Vout stay the same? In other words, G2 doesn’t alter the “max swing” of the plate signal?
Thanks as always
-
For one I do believe you will be changing the capacitance of G1 by changing the electrostatic force on G2. One result, I believe, will be a decrease in compression, slight, change frequency response. I haven't read everything here but probably will later. It looks like an interesting site if you go to the home page.
http://oestex.com/tubes/screens.htm (http://oestex.com/tubes/screens.htm)
silverfox.
-
See: https://www.ampbooks.com/mobile/amp-technology/pentode-gain/ (https://www.ampbooks.com/mobile/amp-technology/pentode-gain/)
It appears that pentode voltage swing is determined by the plate voltage -- it swings between that and ~0. Changes to screen voltage affects plate current. But does a higher plate current cause a plate voltage drop? I think I'm now asking your same question.
Meanwhile, yes when you lower G2 volts it has the result of increasing gain by allowing a smaller G1 signal to get the tube to “max” with less effort. But at the cost of commensurately lower input headroom (voltage swing) at G1.
-
Interesting thread :thumbsup:, I like this argument
Some time ago I was surprised seeing that a reseller (Tube Town) give this indication (for a VVR they sell and call VoCoM)
Variable voltage control for voltages up to 600 V
Possible applications:
- G2 Control
- Power adjustment
https://www.tube-town.net/ttstore/Kits/Power-Supply/Kit-TT-VoCoM-Variable-Voltage-Regulator::5651.html (https://www.tube-town.net/ttstore/Kits/Power-Supply/Kit-TT-VoCoM-Variable-Voltage-Regulator::5651.html)
https://www.tube-town.net/cms/?DIY/Amp-Tools/VoCo_Variable_Voltage_Control (https://www.tube-town.net/cms/?DIY/Amp-Tools/VoCo_Variable_Voltage_Control)
(http://i.imgur.com/ghPisnv.jpg)
You can read something about here
http://el34world.com/Forum/index.php?topic=14308.msg135099#msg135099 (http://el34world.com/Forum/index.php?topic=14308.msg135099#msg135099)
Ciao
Franco
-
But does a higher plate current cause a plate voltage drop? I think I'm now asking your same question.
exactly, I get the sense I'm in a circular firing squad :laugh:
I will follow the links later, time is way compressed today :BangHead:
-
I now remember some threads on this Forum. Lowering screen voltage is not simple, at least at idle, because at idle the screen draws near -0- current. So simple use of a screen dropping resistor is problematic. The Bogen CHA-20 does this using a large value 10K 2W resistor to the 6V6 screens, along with a small 5uF filter cap. (Though it slams the poor 6X5 rectifier with a 16uF cap for the plate supply). But what does this do under signal conditions? For a mere entry-level PA amp in 1960, primarily used for voice announcements, no one may have cared.
So, how to drop screen voltage @ idle? Other alternatives: use zeners to drop screen voltage supply; separate PS iron or PS taps; or separate OT taps (UL). Unless one's prepared do do these things the question is academic.
-
Lower G2 is also lower current.
So is this a small-signal stage? Or a Power stage?
-
Power stage?
I have my breadboard set up ~330 plate, (Tube drag dependent), cathode biased, unloaded preamp produces ~ 45Vacrms
tried kt88, 6550, 6V6, 6K6 and they all yield ~ 4W ac rms @ DL
The only thing I change is Rk for idle bias, and amount of drive to get max clean @DL.
before I start building voltage dividers n such for G2 I wanted to "prove/disprove to myself" that I could/couldn't get more max clean by doing that. Lowering G2
-
Finally got around to playing with G2, results totally unexpected.
Original settings;
Vp 336, Vk 21.37, Rk 550, Vg2 338.2 (UL tap), Ip .0388A, Pdis 12.23
My 1st power rail tap is a stacked cap (see schematic), so I used that as a “test point” for new G2 voltage tap
After G2 volts change;
Vp 359, Vk 7.5, Rk 550, Vg2 180, Ip .0136 Pdis 4.79
So, should I now re-adjust Rk n such to get back to where I was?
NOTE: schematic is now tweaked and set up for 6V6 so values are in this text, not on the schematic
-
"Of course" dropping G2 drops current.
Yes, re-bias back near rated dissipation.
However you do NOT want to "increase gain" by mucking with the Power stage. Power tube makes power. Little tubes make gain. You do need a power tube sensitive enough that a little tube can smack it seriously. All the now-common power tubes are easy to drive. (There are 2A3 on another thread which need giant smack.)
If you have 300+V supply and a 5K load, the power tube should idle near 300/5K or 60mA, 18 Watts. If it is a 6V6, that is HOT. If it is a 6550 (I am lost) that is a waste of a 42W tube, you wanted a 2.5K load and 120mA (but B+ sag will probably fight you).
-
If it is a 6550 (I am lost)
makes 2 of us. I'm using a 6V6 for now.
Before I lowered G2 I had the 6V6 at ~ Max plate dis (12.2W)
I had enough drive to distort the 6V6 but adjusted for max clean out.
essentially the PA was "tuned"
Lowering G2 sent me back to square 1
If I re-bias, get PA back to 12W I don't see it getting me more power at speaker, 12W is 12W weather G2 is 180vdc or 339vdc.
I do see? an application though, can I switch G2 volts for like a 1/2power, something like make before break switch?
-
Lowering Vg2 (all other things being equal) decreases gm* and increases ra. Since gain=gm(ra||Ra), then lowering Vg2 also lowers gain.
lowering Vg2 squeezes the grid curves together more, which decreases the range of available current swing at the plate for any given amount of signal voltage swing at the control grid
-
then lowering Vg2 also lowers gain.
Ya, being thick as a brick, I finally tested the math :think1:
I am gonna play with it some and see how switching Vg2, effects sound, distortion, etc. My brain is cooking at 101/2 degrees so thought is somewhat skewed. I think I'm gonna ask for my money back on the flu shot :BangHead:
-
> lowering Vg2 also lowers gain.
No. This has been discussed.
Internal plate impedance is always overwhelmed (in pentodes) by load impedance.
On a self-biased power pentode, lowering Vg2 lowers plate current and maximum output. Gain goes up some. It breaks-up sooner and won't go as loud. May be a useful "choke". A 100K 2 Watt pot bypassed with >=0.5uFd will usefully adjust Vg2 on 300V amplifiers.
-
> lowering Vg2 also lowers gain.
No. This has been discussed.
Internal plate impedance is always overwhelmed (in pentodes) by load impedance. ...
making voltage gain approximately equal to gm x Ra,
in which case, how can reducing gm also therefore not reduce A?
What am I missing?
-
A 100K 2 Watt pot bypassed with >=0.5uFd will usefully adjust Vg2 on 300V amplifiers
This way :w2:
(http://i.imgur.com/RzPpYSv.jpg)
May be I can wrongly understand the presence of the pot, but I don't understand at all the bypass cap :dontknow:
Franco
-
It's just a filter cap that provides an AC ground to the wiper of the pot.
-
An AC ground .... via the PS (as I've draw) or directly wiper > cap > ground ?
Franco
-
The G2 bypass goes to ground.
If you do not bypass, the 0-25K resistance of the pot wiper against the ~~20K impedance of G2 seriously reduces gain at mid-voltage. 0.5u is about a minimum, and inside a NFB loop may provoke a bass bump.
-
> What am I missing?
Reduce G1 bias to get back near original current.
It is a 4-way problem and there are many possible tradeoffs.
-
(http://i.imgur.com/ospeKix.jpg)
---
0.5u is about a minimum, and inside a NFB loop may provoke a bass bump.
:rolleyes: :rolleyes: :rolleyes:
:huh:
increasingly difficult :dontknow:
:w2: :w2: :w2:
Franco
p.s.: Now I remembered this (mine) thread
* in case where the screen ..........
http://el34world.com/Forum/index.php?topic=16111.msg157302#msg157302 (http://el34world.com/Forum/index.php?topic=16111.msg157302#msg157302)
(http://el34world.com/Forum/index.php?action=dlattach;topic=16111.0;attach=39106;image)