what, no grid stopper?

[Groover]

After reading this thread: http://music-electronics-forum.com/t19503/ I looked at Nyberg's page and at some of the schematics on the AX84 page. It seems that the inputs on every schematic I saw had a 1 meg load resistor in the input circuit, but no grid stopper.

a) Is it just luck a matter of luck if an input stage designed this way will be stable?
b) Does this give a greater voltage swing at the grid, and hence more gain in the stage? (I would have thought so in the past, but with what I have learned recently on this forum and in Merlin's book I am thinking no.)
c) Why else would the input be designed this way?

[JHow]

Am I looking at the wrong file? Doesn't the "Daisycutter" have a 15K grid stopper on the input?

Oh, okay, I see what you mean, like the "P1" has a note that says the input grid resistor was removed from the schematic.

[Groover]

Yes, I was looking at P1 and some others, not at Daisycutter.

I did not see that note. Thank you.

[Wilder Amplification]

What's funny is that a lot of early Marshalls had no grid swampers at all, while some pre-70s 100 watters only install them on the outer pair of tubes.

[tubeswell]

After reading this thread: http://music-electronics-forum.com/t19503/ I looked at Nyberg's page and at some of the schematics on the AX84 page. It seems that the inputs on every schematic I saw had a 1 meg load resistor in the input circuit, but no grid stopper.

a) Is it just luck a matter of luck if an input stage designed this way will be stable?
b) Does this give a greater voltage swing at the grid, and hence more gain in the stage? (I would have thought so in the past, but with what I have learned recently on this forum and in Merlin's book I am thinking no.)
c) Why else would the input be designed this way?

The input stage (v1) typically has no grid current so to speak, so a grid stopper isn't really necessary. The 68k/34k on fender inputs is more to stop HF crap getting into V1 and getting amplified AFAICT.

Also, take another look at those AX84 schemes - most of the other stages in those AX84 pre schematics have voltage dividers of some sort or other in front of the grids, which help to stop excessive grid current anyhow. Also if you make your coupling caps small enough that will also help reduce/stop grid current. 2CW

[Steve Conner]

Grid stoppers aren't for stopping grid current, they're for stopping parasitic oscillations. They do that by damping the resonant circuit formed by the grid wiring.

The limitation they put on grid current is a (usually unwanted) side effect.

12AX7s usually won't oscillate even without a grid stopper, though.

[tubeswell]

Grid stoppers aren't for stopping grid current, .

Well according to Merlin B they are, something along this rationale. The (conventional) current travels from the HT through the plate resistor of the previous stage, charges the coupling cap as the plate voltage swings up, and if the current is excessive (as it is more likely to be with bass frequencies) then excessive current discharges to the grid of the following stage, whence it flows through the next stages cathode and then to ground, and in the process affects the bias of the stage. Grid stoppers prevent the current that is discharging from the coupling cap getting into the next stage's grid (Well that's the message I took away from him anyhow. If I'm wrong I'll eat my hat as usual)

[Steve Conner]

Well, that is one function that they perform in the circuit, though I'd argue that they're not very effective at it. It's better to shorten the RC time constant of the coupling cap and grid leak resistor, or limit the drive voltage.

But it's not the reason for putting them in the circuit, nor is it the reason why they're called grid stoppers. The reason is to stop parasitic oscillations in power tubes, and in the case of a Fender input stage, to filter out AM radio stations with the help of the first tube's Miller capacitance.

[Merlinb]

Well according to Merlin B they are, something along this rationale.

Stopper serve two purposes- current limiting and low-pass filter.
On the input, though, they are really just for filtering out RF. I don't agree with the practice of eliminating them altogether on the input. A 10k is small enough to be tonally benign, while still offering protection from RF. The P1 designer doesn't have a problem with much airborne RF where he lives, but other builders will.

[Wilder Amplification]

On the subject of swamper resistors and grid current, I'm curious as to why certain tubes have a limit as to how high you can go with the leak resistor. Power tubes such as the 6550 spec no higher than a 50K leak resistor when running fixed bias. I've heard that going too high on the leak resistor can cause the tube to become unstable and run away with the bias. Why is this?

[pdf64]

Yes, it's strange how 12AX7 can be biased using grid leakage current (eg with a 10M grid leak resistor to ground which builds up sufficient NEGATIVE grid voltage to bias the tube adequately, dc isolated from previous circuit), whereas with power tubes a max grid leak/feed resistor is specified. I'd assume from that that the grid of power tubes can build up a POSITIVE voltage, if the grid bias feed resistor is above that max value, leading to runaway.
Or does some other effect come into play with power tubes?

[Merlinb]

Or does some other effect come into play with power tubes?

It is assumed that power tubes are operated with a transformer load, while preamp tubes use a resistive load.
A very large grid-leak will develop a positive grid voltage if there is any negative grid current (and grid current tends to increase with gm, and power tubes have lots of gm), which biases the tube hotter.
In a preamp tube any increase in current is counteracted by the voltage drop across the anode resistor, so its impossible to get into dangerous dissipation territory, but with an OT there is hardly any drop across the transformer, so the dissipation just keeps on rising!

In terms of load lines, the DC load line on a preamp tubes almost never goes above the max dissipation curve, whereas it nearly always does with a power tube.

[pdf64]

Thanks, the transformer load thing also explains why a BF Fender reverb driver 12AT7 redplated when the grid leak resistor went bad.

[Wilder Amplification]

Thanks, the transformer load thing also explains why a BF Fender reverb driver 12AT7 redplated when the grid leak resistor went bad.

Same thing would happen on any fixed biased output section (or cathode biased for that matter). Believe it or not, the only difference between cathode and fixed bias is that with the cathodes grounded, the only way to get the grid more negative than the grounded cathode is to reference it to a negative voltage source via the grid leak resistors.

[daz]

This caught my interest because i recently read a post somewhere in which the person stated something to the effect that a smaller or no input grid stopper eliminated a problem he had. that problem being that very high frequency staticy sorta fizz. I took this to equate to the sound i have always gotten with most any amp i've owned, possibly due to my habit of no tone control in my guitars and a treble pass cap on the volume. So when i turn the guitar down to clean up the tone is very chimey and nice except that staticy super high end garbage. It was always something that i could never get rid of w/o harming the tone's sparkly chimney sound to some degree. Long story short, i tried a small resistor in place of the 68k i had there and it really did cut a lot of that out. oddly when i use smaller or no grid stoppers in the gain stages it does the opposite....more high end.