The so called "grid stoppers" make a low pass filter with the resistor value itself and the inout capacitances of the triode (Grid to cathode, to earth and to plate) plus the miller capacitance, which depends on the stage gain. So, the higher the resistance value, the lower the amp frequency response and worse are the transients. I never use them unless I am fully convinced of the necessary of them. Sometimes, small inductors (some few microhenry) will also do the job, with a double pole in the frequency roll off.
1k (even 1.3k) is not much of a low-pass. And the grid is driven from typically higher impedance than that.
The problem is: you think you are building an audio amplifier. But the tube doesn't know that, and has gain out to MHz. It has internal capacitances, and external wires which act as inductances. Capacitance and inductance is a tuned circuit. Tuned circuit in both plate and grid is a potential oscillator.
1k and 30pFd is 1.3MHz. So it is likely that a tuned circuit with these values will be low Q due to R becoming greater than Z(c). The low-Q tuned circuit is much less likely to support oscillation.
The problem is: you think you are building an audio amplifier. But the tube doesn't know that, and has gain out to MHz. It has internal capacitances, and external wires which act as inductances. Capacitance and inductance is a tuned circuit. Tuned circuit in both plate and grid is a potential oscillator.
1k and 30pFd is 1.3MHz. So it is likely that a tuned circuit with these values will be low Q due to R becoming greater than Z(c). The low-Q tuned circuit is much less likely to support oscillation.
Grid stoppers don't have to be carbon composition. In some cases a little inductance will acually help the stopper do its job; sometimes a lossy inductor will do a better job.
Grid stoppers don't have to be expensive even if you believe in expensive resistors elsewhere in the amplifier. If you can 'hear' the grid stopper than there is something seriously wrong with your amp design.
Grid stoppers don't have to be expensive even if you believe in expensive resistors elsewhere in the amplifier. If you can 'hear' the grid stopper than there is something seriously wrong with your amp design.
I bought a DIY amp that was basically a rat's nest wiring - one of the channels seemed to be lower in level than the other. With a scope I found out that the bloody thing is oscillating at ca 600 kHz and removing feedback did not help (I thought there was wrong polarity). I installed 1k grid stoppers on all 4 tubes and voila, the amp works perfectly. So the moral of the story is - I will never have any amp without grid stoppers.
I am with DF96 - the small resistors behave like linear resistance way higher than audio frequencies and since it is linear, you cannot hear any difference - at least that my idea of it.
I am with DF96 - the small resistors behave like linear resistance way higher than audio frequencies and since it is linear, you cannot hear any difference - at least that my idea of it.
It is probably that when no signal current goes through the component, the component contributes nothing to the signal. Grid stoppers in HiFi amps have no current through them since the grid is never positive. ClassA2 and AB2 does draw current and some people claim to hear different resistor types.
Musical instrument amplifiers are often over driven (mostly guitar) and ni those cases carbon comp is usually preferred, not just for their great pulse- and overload-tolerance, and non inductive properties, but also for sound.
Substituting grid stoppers with inductance is chancing it. The inductor may have high impedance at the frequency where the tube likes to oscillate, but what if you use an inductor with self resonance below the frequency where the tube can oscillate? It may very well get worse. The whole point of 'place grid stopper right at the pin' is to introduce as little stray inductance as is physically possible. Carbon comps are the best for this purpose. Since no current is passed, their so called excessive noise is not an issue, they give no more noise than the best metal film.
The resistor is 1k or so.
The tube grid is over 100Megs over most of the audio band.
Look at it like a Voltage Divider. Huge variation of the resistor make "NO" difference to the voltage on the grid.
A high-Mu triode grid may come down past 200k at the top of the audio band. Still a few-k resistor makes about no difference.
Anyway carbon-comp is not THAT bad. It's all we had for decades.
I realise there is no DC current through this resistor yet I wonder about contribution of noise from this resistor, 12AX7 is a high mu tube. Contact, thermal, Shot and or Johnson noise?.
Would a ferrite bead be a better solution?. 9 pin tube socket, cut the grid lugs down the centre, slip ferrite over the socket pin, solder to the lug tip.
If all we are doing is creating a LP filter, why not small value cap in parallel with the grid leak resistor?..
Anyway, I'm interested to see where this goes.
Cheers,
HK
I am still not getting why the carbon comps are BETTER as a grid stopper resistor than a metal film one. WHAT makes them SOUND better?
Grid stopper is needed only and only if the stage is unstable without it. Sometimes it even does harm, from example in a first stage of a power amp that has feedback to it's cathode, where a capacitor around 100 pF to the ground my work better. Sometimes it plays also other roles, like much bigger value resistor at input of a guitar amp, to work as a filter. Or, in the grid of output tubes where it increases time constant of a coupling capacitor when it's being charged through the grid current on overload.
Last edited:
100R + ferrite keeps high Gm and Mu from oscillating, I don't know how or why.. but it works, take away 100R and it oscillates. I'm not sure its strictly LP filter, suspect there is more to it than that. Low level <200mV.. as always could be wrong. Just saying from real world experience.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.