• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Grid stopper resistor required (KT66)?

I've built Mikael Abdellah's KT88 SE amp, and was wondering if I need to add a grid stopper resistor if I want to use a KT66? The data sheet suggests a value of between 10k & 50k. Doesn't this resistor control gain of the stage (along with the grid leak resistor?) I know this resistor has to be soldered right at the tube socket. Can I leave this resistor in the circuit if I want to go back to the KT88's & EL34's?
The only thing I don't like about this, is that it's another component in the signal path.

Thanks in advance.
porkchop61 said:
I've built Mikael Abdellah's KT88 SE amp, and was wondering if I need to add a grid stopper resistor if I want to use a KT66? The data sheet suggests a value of between 10k & 50k.

Hi Glenn

>10 K seems a little high to me. For me, 2,2 K maximum 4,7 K is about right!

Doesn't this resistor control gain of the stage (along with the grid leak resistor?)

No, the gain of the stage don't change.
Can I leave this resistor in the circuit if I want to go back to the KT88's & EL34's?

Yes, you can.


diyAudio Moderator Emeritus
2004-01-27 8:55 am
With any OP tube that is capacitor-coupled from the previous stage, it's a very good idea to use a grid stopper of minimum 1k.

With a pentode/beam tetrode, the grid stopper value can sometimes be a lot higher, to minimize the effect of blocking distortion on overdrive due to transients. A screen stopper, from 47 to 470 ohms, is also a very good idea, even if the tube is triode-strapped.

There is no effect on gain, because a grid stopper does not form part of a potential divider, but there is an upper limit dictated by the tube's maximum grid-ground resistance and, with a triode, the tube's Miller capacitance.

Without stoppers, you run the risk of parasitic oscillations, which can be at ultrasonic frequency. To be effective, the lead from the tube socket pin to the body of the stoppper resistor must be very short and the resistor must not be inductive. Carbon comp. resistors are best if you can get them. Normal wirewound are the worst type and may actually make things worse.
I have a 100R screen stopper resistor in place already per Tubelab's suggestion. This is for Triode strapping the output tube (normally I run in UL, but it's nice to have the ability to switch to Triode & Pentode connections).

I will play around with a grid stopper resistor. What is the best type to use for this application? After all, it is now a component in the signal path. I thought I also read somewhere that this resistor changes the high frequency roll off of the circuit. I guess this is what we're after anyway to eliminated the oscillation. I guess the trick is to keep the roll off above the audio range.
DISCLAIMER: I am a beginer, I read a lot on the net, but thats all... :angel:

I removed the output tubes grid stoppers on my EL34 PP amplifier, following the advice of a friend that is experienced, and it was one of the most effective mods I did to my amplifier... :bigeyes:

At last, all arshness is gone, it sounds natural, very nice... :sing:
Physical topology has an impact, that is to say the layout of the amp, lead inductance etc. Tube type has plays a large role, and some of the types popular in SE use might be more susceptible to it. The OPT can also play a role in some oscillations.

Grid stoppers are highly recommended in cathode followers, and in tubes designed for VHF use.


2006-12-30 11:55 pm
I think you should try yourself to get it to oscillate to confim whether it has enough (properly implemented) resistance.

Switching on/off, plugging/unplugging sources, touching the inputs, volume control up/down/centred, square waves, switch on the fridge etc...
That would be my call.  I have not used grid-stop resistors on my last several projects, and experienced no oscillation whatever.  This includes an EL34 direct-in power amp with 6dB global negative feedback, and in that (and every other) instance, I observed absolutely no resonant behavior (output peaking above 0dB, or at all) up into the low megahertz range.  The place to look, if you have a generator and meter, is for a response that (after rolling off starting between 10K and 30K) rises again and peaks somewhere. In a power amp, the usual suspect is the output transformer primary resonance, which will almost always fall between 50KHz and 120KHz.  Doesn't hurt to pan on up to a meg or so, just to make sure.

In every experiment with grid stoppers I have performed, they were noticeably audible, to the detriment of the transparency of the piece.

Try without, and test.  If you absolutely have to use them, do.  I think that those instances will be rare to none.


Thanks to everyone for the responses.
I don't have a generator for audio (only RF as I work on radios mostly).
No one has mentioned using any grid stoppers on this amp design yet, and a few people have built this amp on this forum. I only really started asking because I want to try a set of KT66's on this amp. so far I've run KT88's, EL34's, and 6L6GC's in this amp just to see how they change the sound. I actually like the 6L6GC's, so I thought I'd try the KT66's as I believe they have a higher rating for plate dissipation.
If no one here has seen this amp design, I have a schematic on my web page, along with some voltage readings for different power tubes I've tested.


My KT88 SE build
Poinz's post #10 explains the situation quite well. But I am going to be a little blase here.

I think the G1 stopper on a power stage is overrated. Sure, it is only a resistor, so why not - I use them as well. But often being a rebel, I once tried a test with and without, and had to increase to 27K before any noticable effect - not hearing-wise, just on the scope (frequency plot).Then, again I suspended such a resistor on up to about 10cm of wire - no trace of difference on the scope.(And the amp was good up to 160 kHz.) One does not say go crazy with wire lengths, but no need to cut off the last every mm.

In tube amplifiers I would say that generally the output transformer has a by far greater role to play in the >20 kHz region than grid stoppers.

So as pointed out, if you can test with a scope/signal generator up to 1 MHz, try it. Square waves are revealing as an initial indication of possible trouble. But again, one piously tend to put it in.

Further (perhaps unnecessary) detail. Hope you are not simply plugging KT66, KT88, EL34 or 6L6GC into the same circuit. There are bias differences, apart from variations of Gm from 6600 umhos to over 11 000. The playground will not be level by simply "rolling" different tubes.

To conclude, Glenn, please forget about this business of "another component in the signal route" (particularly referring to resistors and capacitors). This is another overrated pious urban legend, simply because it has a ring of logic to it. Unless one is using heroically inferior components, they cannot nearly have an effect in audio simply because of their presence - on the contrary.

PS: Poinz, your immediately above post: True, but one must just specify 6L6GC, not earlier GA, GB etc models of lower Pa - small detail.

Edit: Typos
Poindexter said:
Pa max for the 6L6 is 30w. Max for the KT66 is 25w, same as EL34/6CA7.

My bad, I thought I saw 30 somewhere. No big deal, as I changed my bias anyway to run EL34's at a more conservative dissipation. They were running right at the 25w limit, so I changed the cathode resistor to 560R from the original 470R. Now I'm around 22W.

Thanks for your comments. I haven't actually tried adding a grid stop resistor yet, so I may try it just to see if I can actually hear any difference. As for just plugging in different tube types, I'm aware that the bias is not optimal for every one of these tubes. I've actually sat down and looked at where the tubes are operating with the voltage/current measurements I have recorded for this circuit. I was actually thinking of changing the biasing circuit to make it a fixed bias instead of the cathode bias as it is now. I'd have to look for some circuits that are like this. I was actually thinking of something where I could access a trim pot and test point on top of the chassis to set the bias easily for each tube type, like a Dynaco ST70.
Aren't there also circuits out there that "Auto-bias" power tubes?
I was just looking at the schematic for the "Simple SE" board from Tubelab, and I noticed that R16 &R26 are 100ohm 1/4 W metal film resistors. I was under the impression from things I've read that grid stop resistors should be carbon composition or at the least carbon film. Is this true? I've been putting off installing grid stop resistors because all I have are metal films.