Neg Bias Quest...

[trem]

I skimmed the bias section of Merlin's power supply book, and did not see it...maybe I missed it.
I did a search and could not find the answer...but I am sure this has been addressed, and am very happy to read a link.!

With that said .....is it a bad a idea to run the neg bias straight to the power tube grids, and not across the power tube grid resistors.?
Specifically...in the schem below...is it wrong to run the neg bias from the 330k resistor, and then right to the power tube grid.?
That is to say...can I by-pass the 47k grid resistors with the Neg Bias Voltage, and run it from the 330k resistor and then right to the power tube grid.?
Thank You

http://www.drtube.com/schematics/vox/ac100jp.gif

[Enzo]

it isn't a matter of right or wrong. Those 47k are called grid stoppers. Not all amps have them, many do. You can leave them out, but then you no longer have what they do.

Look up "grid stoppers", I am sure we have discussed it more than once.

[Chuck H]

If you locate the bias supply direct to the grids the grid stopper becomes part of a voltage divider relative to the impedance of the bias supply. The effect will likely be small, but it might not. 47k as it relates to 330k forms a voltage divider worth considering. I'm not sure what affect this has on the function of the "grid stopper" circuit. Since grid stoppers are best located directly on the tube pin I would think it inconvenient to apply the bias there as well. Locating the grid stopper on the board (or anywhere that isn't directly on the tube pin) takes away a lot of it's effectiveness for the purpose of squelching interference and HF oscillation.

[Enzo]

I THINK he was asking if he could run the 330k right to the grids, eliminating the 47ks. I don't think he wanted to connect the bias supply itself directly to the grids.


I don't know where these 47k are located within the amp, but it seems to me the only way they form a voltage divider is through the grid, which usually is not drawing any current, and so no voltage drop.

Or are you talking leaving the 47k and moving the 330k to their other ends. That could form a voltage divider for the signal, but not the bias.

Or am I missing the point.

[Chuck H]

Or are you talking leaving the 47k and moving the 330k to their other ends.

That was my interpretation. We'll see I guess.?.

[Alan0354]

I read bypass as jumper across the 47K!!!

Is grid stopper mainly to form a RC low pass using the capacitance of the grid to filter out RF and high frequency? Fender and Marshall use much lower value. 47K is kind of high and I would imagine loosing some treble.

[trem]

Yeah...sorry...written communication is an art I am not versed in.
I believe Enzo explained my thought.
I do not want to physically eliminate any components.. I was wondering about the wisdom of moving the 330k resistors right to the tube girds.
So the AC Signal would would hit the 47k...the bias voltage would hit the 330k, and the 2 would meet at the grid. (Sorry...I REALLY need to learn a schematic program).
The grid stoppers are right on the tube...Pin 5 and 6 I believe.? Pin 6 is the unused node, and is used as a tie point.
Thanks Again....I hope that is more clear.

[R.G.]

You can do it either way, but it partially defeats and muddies the operation of the grid stoppers.

Grid stoppers are there to suppress parasitic RF oscillation. The output tubes have internal capacitances between every possible pair of electrodes and also capacitances between every electrode and the external wires. If you have the bad luck to get things just barely wrong, these cause feedback paths that can make the tube self oscillate. Grid stoppers help prevent this by forming a low-pass filter between everything outside the grid terminal and the effective capacitance of the grid to ground.

For this to be effective, and not just another antenna that picks up external junk, the grid stopper resistor body needs to be as close to the grid terminal as you can get it. Literally, the body of the resistor should be nearly touching the tube socket lug. The closer the better, given that you don't break the resistor lead loose bending it in the socket lug.

Moving the grid leak resistor from the "outside" of the grid stopper to the inside of the grid stopper makes the grid leak resistor another "grid stopper", but a clumsy one from a separate part of the circuit, and means that it, too has to be right on the grid terminal to avoid antenna-ing in extraneous signals. It gets clumsy. If the grid leak (330K in this thread) resistor is outside the 47K grid stopper, then the wiring length and capacitances of the grid leak resistor are less critical and don't increase the chances of interference and feedback oscillation.

I keep saying "probably" and "chances" because exactly what happens depends on the arrangement of the leads, components, and chassis shielding. Probably (there's that word again) moving the grid leak to inside the grid stopper would work OK most times, maybe nearly always. But there will come a time when it doesn't. Unless you enjoy debugging odd problems caused by aging tube sockets, new tubes and/or continuity of the chassis shielding, the simple and safe thing to do is to put the grid leak resistor outside the grid stopper.

[trem]

Thanks Everybody -
OK...now I understand why...
It was just a matter of ease for me to relocate the 330k...but I will wire it per the schem.
I looked at Randall Aikens pages...and checked Merlin's book, but did not see my question addressed.
IS there a link to a written explanation (like the excellent synopsis by RG above) that talks about location of Power Tube Grid Stop Vs. Grid Leak resistors.?
Thank You