ampage
Tube Amps / Music Electronics
For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

Search for:  Mode:  

Grid stoppers at power tubes


 :
9/21/2000 7:34 PM
Sven Wickström
Grid stoppers at power tubes
I have done some testing with high value grid stopper resistors in a cathode biased 2x6L6 amp.  
The purpose is to tame grid blocking when the power tubes clips.  
I have tried up to 100 kohm resistors.  
There might be a loss in high end, but its hard to hear.  
Looking at a scope connected before the grid stoppers, the curve goes from flat top to a round top when the value is raised.  
 
Anyone has opinions on this?  
 
Sven
 
Today
ampage
archive
And now, a word from our sponsors:


 
9/21/2000 8:58 PM
ken gilbert
it's a good way to cut out some of the flappy, farting, blown out sound. just be cautious that you do not inadvertantly exceed the max grid circuit impedance for the particular power tube in question, lest they run away on you.  
 
kg
 
9/22/2000 8:45 AM
Sven Wickström

Max grid circuit resistance for a 6L6GC is 500 kohms in cathode bias, so with 220 k grid load resitors, 100 k grid stoppers should be ok.  
 
Sven
 
9/22/2000 3:31 AM
Bruce

quote:
"Looking at a scope connected before the grid stoppers, the curve goes from flat top to a round top when the value is raised."
 
 
Did you increase the drive level to the actual grids of those power tubes to see what happens or was this with the drive level at the same for all tests?  
Did you need to reset the bias voltage on the power tubes when using 100K?  
Did you happen to also look at the output of the power tubes during any of those setups?  
I've never heard anyone say "the grid stopper is used to tame grid blocking"... is that the conclusion you came to, or ???  
Have you tried reducing the grid LOAD instead of increasing the grid stopper resistance, as KG hinted at?  
Anytime you increase the grid stopper you are also increasing the grid resistance and the "Zed" of the grid bias network.  
Way too high and the tube will try and bias itself through the grid leak method.  
In regards to the O'scope pattern, I'm curious if you couldn't have seen the same rounded tops, just by reducing the overall output from the PI/driver.  
 
Bruce
 
9/22/2000 11:11 AM
Sven Wickström

Let me first say that this is a DIY project, and that i dont have the theoretical knowledge of this.  
 
With no or small gridstoppers, the input of the power tubes clips at the same point at the output, with the same "flat" curve.  
I want to change this point so the inputs clips at a later point.  
I got the idea for this test from Randalls Aikens article about http://www.aikenamps.com/BlockingDistortion.html" TARGET="_blank">Blocking distorsion  
I did the test with a 1 kHz sine wave input, output into a 4 ohm resistor load. Input signal was set so that output on the dummy load was a few volts over clipping point.  
quote:
"Did you increase the drive level to the actual grids of those power tubes to see what happens or was this with the drive level at the same for all tests?"
 
No, i used the same input signal.  
The measured output voltage at the dummy load was the same with 100k gridstoppers vs. none.  
quote:
"Did you need to reset the bias voltage on the power tubes when using 100K?"
 
Bias voltage, measured over a common cathode resistor, is the same at no signal. At the signal above, the bias voltage raises about 3-4% with 100 k gridstoppers.  
quote:
"Did you happen to also look at the output of the power tubes during any of those setups?"
 
It was the same, and the curve did not change.  
quote:
"I've never heard anyone say "the grid stopper is used to tame grid blocking"... is that the conclusion you came to, or ???"
 
Got it from Randalls Aikens article.  
quote:
"Have you tried reducing the grid LOAD instead of increasing the grid stopper resistance, as KG hinted at?"
 
I tried that, from 220 k to 100 k.  
I had to raise the input signal to get the same output, but could see no change in waveform.  
quote:
"Anytime you increase the grid stopper you are also increasing the grid resistance and the "Zed" of the grid bias network.  
Way too high and the tube will try and bias itself through the grid leak method."
 
Maximum grid resistance is 500 k at cathode bias, so that should be ok, from what i know.  
Can the bias voltage raise come from grid leak, or is it because of a change in operating point?  
 
Regards, Sven
 
9/22/2000 3:05 PM
Bruce

Well, yes "bias voltage" can change with grid leak biasing but it takes a large resitance on the grid to start changing the voltage and I don't think 500K would be enough to change much at all.  
Testing or experimenting with my power amps, I've gone up as high as 47K grid stoppers on a couple of my amps, mostly with no ill effects.  
They seemed fine with 1k5 to 6k8 though so that's where I stop.  
Some old tube amps use 5M with a blocking cap in front of the grid load to do this.  
Go check out the schem of something like an old 5B3 Fender Deluxe.  
I was taught that was real grid leak biasing... of course that was MANY years ago!  
I don't know how a low mu tube would react to that though.  
Yes, I respect Randall greatly but I never heard anyone state that the grid stopper resistor is used to tame grid blocking, ao, I'll have to pop over there and read the whole thing again.  
It's been a while since I did.  
I wonder if there is a terminolgy thing going on since the grid load and the grid stopper appear to be attached to the tubes at the same place but IMO have two to three different functions.  
Good stuff Sven.  
I haven't spent that much bench time on experimenting with this one aspect like that.  
Bruce
 
9/22/2000 1:51 PM
MBSetzer

Confirmed Sven's findings a couple years ago.  
 
I think with high enough ohms it would be more difficult for the blocking cap to charge, and this at least changes the sound of the resulting distortion when overdriving.  
 
Another one of those key components that aren't even present in some legendary amps, but now I consider a worthwhile tweak to include values more extreme than traditionally used, especially since the entire signal passes through the component. Naturally, the voicing changes in relation to the values of nearby components, with secondary holistic effects depending on what signals are available from upstream and what you will be doing with them downstream.  
 
Experiment, you can change the grid stoppers a fairly large degree, like doubling them, or doubling them again before responding with more minor tweaks to other parts. Conversely, if tweaking other components moves you in a desirable direction but you think there is increasing unpleasantness underlying, adjusting the value of the grid stoppers (usually upward from traditional values) might be an additional cure to have available along with the more common remedies. I have used up to 100K but still consider zero ohms ideal for some amps, so somewhere in that ballpark ;) there's a lot of room to find a sweet spot. You might just have to go from zero to 100K in about half-dozen steps with accompanying changes to nearby components, keeping everything else constant and playing the same licks on both Les Paul & Strat in all pickup positions and various amp volume levels. The waveforms do change, but once you hear it you'll know what to do with it.
 

  Page 1 of 5 Next> Last Page>>