Have you tried buffering power valve grids?

[Mostro]

Hi guys.

I´m currently debugging my build, a modded AC30. It has voltage regulation, but I have a problem, when the voltage is low it has lots of crossover distortion. Horrible sound. So I found out that a way of solving that is buffering the grids of the EL84´s. I started to investigate the subject but there´s little info on that, the only thing I found with an actual schem, besides some mentions of it in diverse places, is this:

http://www.geofex.com/Article_Folders/m ... tfolly.htm

Has anyone tried this? If you did, does it work OK?

Any info on this subject: schems, experiences, etc etc will be welcome.

Thanks!

EDIT: Forgot to mention that it´s a cathode biased amp.

[pdf64]

I remember Gingertube posting some great info on this topic, either on TAG or MEF or both.
I'm considering such an arrangement for a KT120 bass amp.
However, have you considered a lower tech solution, eg a PPIMV to tame the output from the LTP? Excessive blocking distortion may be pushing the grids negative.
Pete

[Colossal]

Vox AC30s use large value output coupling caps; 0.15uF. The dreaded fizz of crossover distortion is a well characterized phenomenon (google Paul Ruby fizz fix) that some builders experience with Marshall 18W 1974 circuit builds. Not all builders do however. The solution as first documented by Paul Ruby was to install zeners post grid leaks and some also elevate the power tube cathode with a zener as well. The AC30 has a slower bias excursion recovery time with the large coupling cap. It doesn't take much to drive EL84s either, only about 10-12V, IIRC. Cathode/source followers seems like an interesting option and might add some nice smoothing. Try it out!

[Mostro]

Thanks PDf and Colossal for the answers.

I have installed a pre PI master, I have to try a post PI one. Yes, the problem is the grids going into conduction I believe, not helped by the 100n coupling caps...

I´m curious about the cathode R zener. How would that be? Between the R and ground?

The post grid zeners I have read about, but have yet to try. The problem is, since I have a Variable Voltage Regulator, how would varying the voltage affect that solution?

[dave g]

The zeners may not do their job when you turn down the VVR since their reverse breakdown voltage is fixed, yet at lower VVR settings you will need less swing to drive the output tubes into distortion. I haven't tried it though and it's entirely possible that they would still work. I think a better solution in general than the zeners is to lower the coupling caps between your PI and EL84 grids to something like 0.01uF, and increase the value of the EL84 grid stoppers to something like 47k or 100k.

DC coupled cathode followers driving EL84s into heavy grid conduction scare me a little. They'll get rid of the crossover distortion, sure, but they're capable of supplying more than enough current to fry the grids of your EL84s. If you go that route, I would definitely stick some 100k resistors between the CF drivers and the EL84 grids.