I see that everybody are quoting 2 resources on the net for hiss.. one is Aiken other is Valve wizard.. and their own opinion is? Which?
Will be better to use soldering iron and find by themself in maybe 2-3 minutes, draw some conclusions. CC like non inductive resistor is premium resistor, so grid stoper fro high gm devices is first option.
Put metal film in gate stopper in cascoded CCS with mosfets and let me know results. Oscillating?
Cartman wrote: ↑Fri Jan 12, 2024 3:27 pm
In a push-pull power amp with grid stoppers, is it necessary or beneficial to match the resistor values? The reason I ask is I have a Marshall 18W clone with EL84s that has 5.1K grid stoppers, but they have drifted to about 6.4K and 7K. I'm trying to decide if I should change them or it's really nothing to worry about. Thanks.
Matching resistor values in a push-pull power amp with grid stoppers like your Marshall 18W clone is beneficial for optimal performance. If the 5.1K grid stoppers have drifted to 6.4K and 7K change them to maintain circuit balance and ensure reliable amplifier operation.
The resistors in the original Marshall were probably 10% tolerance and I highly doubt anyone was sorting for matched pairs. The output tubes also likely had a bit of a mismatch, same with the PI. I wouldn’t sweat that amount of difference, but “optimal” is subjective. If the amp sounds good to you, don’t worry about it.
Quick footnote about noise.
Whatever the source of the noise, it matters most the closer to the amp inputs it gets generated. Noise at the jacks input or reverb tank input is amplified by the full gain of the amp. Noise at the output tube grids is only amplified by the gain of the output tubes. That can be a difference of 60-100db of gain.
I like metal film on input stages. It gets less important as you go toward the output stage.
Another funny thing about CC. CC has a high voltage coefficient of resistance. That is, the resistance itself changes with the voltage across the resistor. This makes for a small single-sided distortion. I believe that this is responsible for the "carbon comp sounds better" mythos, as humans seem to perceive slight second order distortion not as distortion, but a "sweetening" of the sound. This only comes into play when the signal swing is large - probably a minimum of 70V peak to peak and up.
maxkracht wrote: ↑Mon Jan 15, 2024 2:47 pm
The resistors in the original Marshall were probably 10% tolerance and I highly doubt anyone was sorting for matched pairs. The output tubes also likely had a bit of a mismatch, same with the PI. I wouldn’t sweat that amount of difference, but “optimal” is subjective. If the amp sounds good to you, don’t worry about it.
the pihers could be 10% perhaps, at least they vary at least 5-6% often. But the iskras vary less
Roe wrote: ↑Mon Jan 15, 2024 6:16 pm
the pihers could be 10% perhaps, at least they vary at least 5-6% often. But the iskras vary less
Fair enough. I still hold by my point that a little mismatch won't hurt anything or make a drastic difference. Whether someone finds a bit of asymmetry good or bad is entirely subjective.
The answer IMHO, from the extensive reading I have done on guitar amps over the past few years, is "It depends..."
I believe a Dumble style amp would greatly benefit. My understanding is that an ODS depends on a clean, "almost HiFi like", power section to produce its clean tone when the overdrive section is not used. Distortion is created in the OD section when desired.
On the other hand amps that depend on distortion in the power section I don't think matching is as critical and may even add some "character" to an amps personality.
grid stopper on g1 on ot.tubes is there to protect against the oscillations, and nothing more, and there is no difference and can be in operaton between 5k and 7k, maybe only in extreme clipping but nobody using and playing like this and even then IMO is impossible to detect some differences
bepone wrote: ↑Tue Jan 16, 2024 4:19 am
there is no zero grid current.. grid current measures in uAmps and in case of clipping going higher..
Yes. But what is the difference in voltage across two mostly matched resistors - say, 10% different - when each lets through a microamp of current? It's calculable.
And the signal voltage on output tube grids is up to twice the bias voltage before grid clipping. When the amp goes into grid clipping, grid stoppers make the clipping really abrupt as they stop the grid from getting as much current as it needs to move with the signal. If you want to allow less-distorted grid-positive drive, you have to drive the output grids from a very low impedance like a follower. Grid stoppers in this case directly work against the grid drivers.
Since I have a simulation of a 100W 6L6 power amp handy, I ran it with the power tube grid stoppers offset by 100%: 5k1 on one side and 10k2 on the other. I see no significant difference in the balance of the outputs, even when it's pushed into clipping and producing grid current spikes.
