KT66 grid leak resistors

[Pierre]

Hi everyone,
I wanted to find out the general consensus on grid leaks on KT66 tubes. I just built an amp with 2 x EL34L tubes push-pull, fixed bias, with a Larmar 250K//2.2M PPIMV. I changed to a set of Valve Art KT66s biased at 35ma with ~450V on the plates. I only looked at the data sheet after installing the new tubes and saw 100K listed as max for grid leaks in fixed bias. The bias circuit adds about another 20-30K of resistance. It seems that data sheet values are exceeded on a regular basis in guitar amps. Should I change it or leave it?
Also, I have 5.6k grid stoppers and the data sheet recommends 10-50K. Probably just a matter of personal taste, but I'm interested in other's opinions.
Thanks in advance for your input.

[martin manning]

As the Lar-Mar is turned down the grid leak resistance is reduced, so that's in your favor. The grid stopper adds to it, so don't go too far. The bottom line is, is the bias stable? Monitor it over time and while playing. If it's steady, I'd say you are fine.

[Roe]

220k tends to work, even on the 4xkt66 amps

[Pierre]

Martin & Roe,
Thanks for the replies. Everything sounds great and seems to be fine. I wasn't sure if once pushed, things would start to melt down. That' s where experience is important.
Thanks again

[pdf64]

Bear in mind that the 100k g1 fixed bias resistance limit seems to apply to all the beam tetrode type tubes commonly used in guitar amps.
G1 current tends to increase with tube operating temperature; I think gingertube has suggested that a benefit of the 'bias to idle at 70%' guideline is that it keeps temperature reasonable, hence grid current reasonable, and so the 100k limit can be breached.

[nworbetan]

I don't remember where I read it, but I've read that as tubes age (by being used, not shelf time) they become more susceptible to runaway grid current, i.e. the grid current goes up which causes an increased voltage on the grid leak resistors, which causes grid current to go up, which spirals out of control and destroys the tube. I can't remember the exact cause though, something about gas escaping from the metal elements maybe?

At any rate, the thing I took away from what I read is that higher grid leak values are fine as long as you're okay with the tubes life span being somewhat shorter than with a lower grid leak value. This kind of trade off happens pretty often in guitar amps with voltages that meet or exceed the manufacturers' recommendations.

[trobbins]

Outgassing is the term used to cover the generation of parasitic gas coming out of the metal structures (mainly the anode plate structure) and glass - and depends on temperature, operating time at temperature, pre-conditioning before manufacture, and operating age of the valve. Gettering material normally reacts with that parasitic gas when the valve is operating, and so a reaction rate balance occurs that normally keeps the parasitic gas level suitably low. That unwanted gas causes leakage currents between electrodes, and grid leakage current needs to be kept suitably low or bias voltage will walk-away over time and result in red-plating.

Some valves don't start with the same level of 'vacuum', or use the same process to outgass the plate structure prior to valve assembly, or the gettering is not as sufficient as needed, or a valve pin seal is very slowing leaking, and so can exhibit higher parasitic grid conduction levels as part of normal manufacture tolerances or within normal shelf and service life time expectations. Some valves get accidentally red-plated during life, and that can cause relatively gross outgassing of the plate which can't be recovered from.

The datasheet grid leak resistance is then a 'catch all' value that should ensure acceptable valve performance when looking at a large population of valves placed in equipment.

So even sticking to the recommended grid leak value does not 100% guarantee no related grid leak issues, nor does using a much higher grid leak value mean death is imminent.

[Andy Le Blanc]

I ran into an issue with EL/KT types and para-phase inverter with the inverting side directly coupled to a resistance divider.
The G1 current can throw off the inverter bias. Have to use a blocking cap. 6l6/6v6 don't suffer the issue so you can simplify.
A little G1 current is normal under typical conditions for some tube types. Follow the data sheet recommendations for bias type and application.
I set bias as PS condition and don't correlate it to the device unless as a design limit...IE, design center values for older tube type.

Eyeball the edge of the getter if your worried about gas contamination. you'll see a slowly increasing dark band around the silver in side the envolope, as the getter is used up.