stephenl wrote:
Hi Ian,
Does the "too large grid leak issue" apply only to fixed bias amps or cathode bias as well? On my cathode biased KT66 PP PA, I have 10k on the stoppers and 220K leaks. Should I lower the leaks?
Respectfully,
Steve
Steve,
A bit iof a "rave" but here goes:
For power tubes it is NEGATIVE Grid Current we are concerned about.
This grid current is mostly due to residual gas in the tube. As electrons accelerate "up" the tube, from cathode toward anode, some of them collide with residual gas atoms. The collision is energetic enough to strip an outer orbit electron which makes the atom into a positively charged ion. That accelecrated back "down" the tube toward the cathode AND the grid (which will be the most -ve element in the tube structure and so offer the most attraction to the +ve ions). Electrons must flow into the grid to neutaralize those +ve ions which gather at the grid, which is the same as saying that current must flow out of the grid.
When grid current flows it develops a voltage across Rg1 which opposes the bias voltage (by making the grid more +ve).
That reduced bias results in increased tube current, more collisions, more ions, more grid current, less bias etc. etc. round and round until the tube melts from over current.
The maximum value of Rg1 is specified such that this bias shift is limited to values which will not cause a "run away".
With auto bias (cathode bias) there is a mechanism to oppose the increased current from teh bias shift, that is, the increased current will genertate more bias voltage.
With fixed bias there is no opposing mechanism and that is why we see 2 values for max Rg1, one value for fixed bias and a typically X4 higher value for cathode bias.
The max Rg1 values are specified at max tube dissipation. In VERY many fixed bias guitar amps the Rg1 value exceeds the max Rg1 data sheet value and that is why they are biased at 70% of maximum dissipation. By By dropping the tube dissipation to 70% you can use a Rg1 value X2 higher (inverse square relationship). If they were to be rebiased at 100% dissipation then they would very likely blow up, or at least as soon as tubes started getting a little old and gassy they would certainly blow up.
This last paragraph needs a caveat. It should read "if Average Power Dissipation is reduced by X2 then the max Rg1 can be increased by X2). At idle you reduce dissipation by X2 by reducing current by root 2 and that is where teh 70% comes from (1 over root 2 = 0.707)
At high volume (high power) the Average Power dissipated goes up, particularly when you start overdriving and the waveshape becomes square'ish (power is proportional to area under the curve). The assumption that you can use an Rg1 of X2 max if you bais at 70% then becomes seriously dodgy and the amp becomes unreliable.
Ever wondered why HiFi tube amps seem a lot more reliable that tube guitar amps, well it is not just how hard they are expected to work, they typically are better designed, particularly when it comes to Rg1 values on the output tubes.
For KT66 see here:
http://tdsl.duncanamps.com/pdf/kt66.pdf
See Note 2 and Note 5 on last page.
Cheers,
Ian
