Low Pass Filter: Math vs Practice

[tmknight]

In this schematic (highlighted green), do we think there is a practical reason for R=470k, C=20p vs. R=20k, C=470p (both equal fc=16.9 kHz)? Or is it just the components values they likely had most readily available/on-hand?

low-pass-filter.png

[pdf64]

The effect of Miller capacitance (about 120pF) with that very high value grid circuit resistance will be somewhat more significant.
High value grid stoppers can help to mitigate blocking distortion. That shouldn’t be an issue there though.
http://aikenamps.com/index.php/grid-res ... -they-used

[bepone]

470k here is mostly for creating distortion, over the g1 positive grid current (creating clipping on positive side). on negative side of the input signal clipping is created with valve current cut off. plus with input capacitance (150pF+20pF) 470k is forming low pass filter over 1.5kHz..

20pf is for fine tuning and oscilation protection.

with low R (20k) there is no effect of clipping this much..

1M also can be used for more creamy sound.

[John_P_WI]

I'll be a little more cynical and say that the 470k r is simply a carry over from the isolation resistors used in this position of a 2 channel amp, think old bassman or plexi. In reality, it adds some input impedance isolation to the following gain stage as the preceding volume pots are varied making the tone more consistent.

[bepone]

schematic is from dual rectifier m. boogie, and 470k is not isolation resitor.. because you have GAIN pot selectors before over the LDRs which are isolation resistors in Mega ohms....

[roberto]

That 470k together with the 20p has multiple functions:
- it does indeed a LPF with the 20p;
- it does a LPF with the miller capacitance of the following stage;
- it increases the input impedance of the grid of the following stages, affecting how it is (over)driven;
- it dampens thin-tone effect at low gain settings.

That 470k is taken from the SLO, as other parts of preamp values.
I generally prefer lower values, and distribute the LPFs on more stages, where I find it is needed.

[tmknight]

Good input! Yes, this is part of a Mesa DR pre-amp. I figured the high R value was directly in response to the previous bright caps (0.001/0.002u) which will, in a high gain circuit, introduce a lot of hiss (which I believe tends to be in the 12-20 kHz range) and setting up the next stage to not be subject to driving that high freq to oscillation. Though seems there would be competing (maybe additive is more appropriate) effects between the 20p cap and the 12ax7 capacitance. Wonder if this contributes to the "muddiness" some attribute to DRs (I've never played one, so really don't know)? Wherein a lower value (though, maybe not 20K) might achieve more specific results (assuming very high frequency cutoff is what they're after and not tone shaping) with less darkening effect of overall tone?

[bmx]

The 20p is noticeable when you remove it, but not all that much. And certainly there are no oscillations when you remove it. You can also jump the 470k with a 10k, which is much more effective due to Miller capacitance-still no oscillations. It will basically sound like a DR, but much less dark, 90's tone with more apparent saturation and sustain. Noise doesn't go up much. Most of the noise is from the heaters not being elevated- that is a guess though. I have built the DR turret style and, with elevated heaters, it is quieter than stock. Additionally, I have heard of people lifting the artificial center tap resistors from the board and connecting it to an elevated voltage with good results (Someone on GroupDIY). That's my two cents.

[bepone]

Good input! Yes, this is part of a Mesa DR pre-amp. I figured the high R value was directly in response to the previous bright caps (0.001/0.002u) which will, in a high gain circuit, introduce a lot of hiss (which I believe tends to be in the 12-20 kHz range) and setting up the next stage to not be subject to driving that high freq to oscillation.

why you think that this 470k will lead to hiss?

Though seems there would be competing (maybe additive is more appropriate) effects between the 20p cap and the 12ax7 capacitance. Wonder if this contributes to the "muddiness" some attribute to DRs (I've never played one, so really don't know)? Wherein a lower value (though, maybe not 20K) might achieve more specific results (assuming very high frequency cutoff is what they're after and not tone shaping) with less darkening effect of overall tone?

when you finish the amp you can find all those answers by yourself or it is better to do SLO100 /50 which was born before

[roberto]

why you think that this 470k will lead to hiss?

I think he refers to the Johnson–Nyquist noise.

[bepone]

why you think that this 470k will lead to hiss?

I think he refers to the Johnson–Nyquist noise.

there is no current passage or it is in micro-amps during overloading (signal ratio is max in that moment and hiss is masked with the signal).. i was putting al kind of resistors there , in good amp there is no hiss also in high gain....so hiss is coming from different source is'nt it?

[roberto]

there is no current passage or it is in micro-amps during overloading (signal ratio is max in that moment and hiss is masked with the signal).. i was putting al kind of resistors there , in good amp there is no hiss also in high gain....so hiss is coming from different source is'nt it?

Thermal noise is not linked to the current.
On top of that, if you want to change the voltage of one capacitor, a flow of current is always necessary, and there we have around 200 pF.

[bepone]

ok, yes, ok is your point.. but g1 resistor is (almost) negligible for hiss in guitar amp.. think how many hundreds of kilo-ohms you have in tone stack and still there is no hiss..

[pdf64]

I think it’s the equivalent source impedance in the grid circuit that matters for such noise.
Grid stoppers add directly to that, whilst for other stuff like volume controls and tone stacks, there’s parallel paths.

[roberto]

It's a long time that I don't build hi-gain amps (even if one is in its early stages) because during the last years I focused more on hi-fi, but I clearly remember the difference between different pots values and grid stoppers as well.