Conjunctive filters (not?)

[Alex R]

Does anyone know how to work out the frequency response of those simple resistor-capacitor filters you get in the signal path between stages in a lot of high gain amps. Not going to ground, just in parallel in the signal path?

I used to call them 'conjunctive filters', but I don't think that's what they are...

I have in a MV Marshall which has a 470k/470p filter chaining the signal from the 'high' input first stage into the 'low' first stage. The amp loses way too much bass just there and the quick fix might be to change the filter (lazy).

[Bruce / Mission Amps]

Does anyone know how to work out the frequency response of those simple resistor-capacitor filters you get in the signal path between stages in a lot of high gain amps. Not going to ground, just in parallel in the signal path?

I used to call them 'conjunctive filters', but I don't think that's what they are...

I have in a MV Marshall which has a 470k/470p filter chaining the signal from the 'high' input first stage into the 'low' first stage. The amp loses way too much bass just there and the quick fix might be to change the filter (lazy).

Make the cap 680pF to 1500pF. Try it in stages.

[stokes]

There is a formula in the RCA Receiving Tube Manual for figuring the values of a conjuctive filter,but I think it is called a corrective filter in the manual.They are described as being used across the OT primary,it might work in your application.

[mooreamps]

Does anyone know how to work out the frequency response of those simple resistor-capacitor filters you get in the signal path between stages in a lot of high gain amps. Not going to ground, just in parallel in the signal path?

I used to call them 'conjunctive filters', but I don't think that's what they are...

I have in a MV Marshall which has a 470k/470p filter chaining the signal from the 'high' input first stage into the 'low' first stage. The amp loses way too much bass just there and the quick fix might be to change the filter (lazy).

Yes, I can tell you a few thing about this. . . First of all, a resistor and cap in parallel is not what has been coined as a "conjuctive filter". The cap in parallel will shunt the higher frequencies around the resistor beginning at what we call the cross-over point. Now, just to make things simple to understand without going in too deeply, in analog electronics in this case, we are going to call the cross-over point between these two components ; at which ; their impedance becomes equal at a certain frequency. . . It can also be "coined" as the 3db down point ; at that frequency......


Now, for example. With the case of a 470 pF cap in parallel with 470 K, that cross-over point is pretty close to 700 hz. 440 hz on the electric guitar is concert A, so that's going to be pretty high for a cross-over.

Instead, lets select a lower cross-over of, well, lets say 83 hertz. The value of bypass cap in parallel with 470K ohms for this frequency is in the neighborhood of 4.7 nF. So, if this value works out to be too much bass, then you can walk this value down to 2.2nF, or 1 nF at your discression.

-g

[daz]

Gary........Is there a combo you could tell me that would allow a attenuation of all but the very top frequencies so that the signal would have more top end sparkle?

This gave me an idea by the way that i think would be quite helpful in determining the perfect combo quickly. I'm sure many have done this before, but i'm thinking of a pot with 4 alligator clips, 2 off one side and 2 off the wiper. (one pair on about 1' wires) Connect the wires to the place in the circuit and use the other clips to try different caps. Between turning the pot (probably a 1 meg would be best) and clipping various caps you could quickly try every imaginable scenario. Think i'll try that tomorrow.

[Alex R]

the cross-over point between these two components ; at which ; their impedance becomes equal at a certain frequency. . . It can also be "coined" as the 3db down point ; at that frequency......

That's the missing piece of my jigsaw! Now I get it. I'm always happier if I understand the damn thing!

Many thanks to stokes for the ref and to Bruce for the practical suggestions.

It's definitely this point - the 'one-wire mod' place if you like - that the amp drops all its bass and half its mids, you can see it clearly on the scope. Some MVs/JCM800s sound fine with it (toppy but fine) but not this one. I will report back.

[Steve Conner]

A parallel RC in the signal path will interact with the impedances of the rest of the circuit to add one pole and one zero. So there are two time constants and two 3dB points. A control engineer would call it a lead network. You can play with the applet on this page to see what the frequency response looks like.

http://www.facstaff.bucknell.edu/mas...Comp/Comp3.htm

So at low frequencies, the gain will be determined by R and the input resistance of the following stage, which I'll call R2. If R was 9M, and you were driving a stage with a 1M grid leak, you'd be 20dB down.

At a frequency such that C has an impedance equal to R, the response is 3dB up from this, so -17dB. It now starts to rise at 6dB/octave, until you get to a frequency where C has an impedance equal to R2, when it flattens out again.

[mooreamps]

Gary........Is there a combo you could tell me that would allow a attenuation of all but the very top frequencies so that the signal would have more top end sparkle?

I do not know the answer to this. I design new builds, I don't work on fixing them. A better source for this might be KOC himself, or others who have a better working knowledge of existing equipment. But I think you can add a 47pF to 100pF cap bridged across most pre-amp master volume controls that would give you the same effect.

-g

[Bruce / Mission Amps]

Gary........Is there a combo you could tell me that would allow a attenuation of all but the very top frequencies so that the signal would have more top end sparkle?

This gave me an idea by the way that i think would be quite helpful in determining the perfect combo quickly. I'm sure many have done this before, but i'm thinking of a pot with 4 alligator clips, 2 off one side and 2 off the wiper. (one pair on about 1' wires) Connect the wires to the place in the circuit and use the other clips to try different caps. Between turning the pot (probably a 1 meg would be best) and clipping various caps you could quickly try every imaginable scenario. Think i'll try that tomorrow.

Here is the approximate -3dB points of a 470K resistor in parallel with a few different caps:
360pF = 935Hz
470pf = 720Hz
680pF = 495Hz
1000pF = 340Hz
1500pF = 226Hz

[daz]

Thanks. I know this is pathetic, but i don't understand that exactly. When you say -3db at X Hz, does that mean it drops 3 DB only at that point. or does it drop 3 DB there and roll off even more as you go down in frequency? in other words, if a given R/C is -3DB at say 500Hz, would it be like taking a graphic EQ and creating a progressive slope from 500Hz towards the lowest guitar frequency?

[mooreamps]

Thanks. I know this is pathetic, but i don't understand that exactly. When you say -3db at X Hz, does that mean it drops 3 DB only at that point. or does it drop 3 DB there and roll off even more as you go down in frequency? in other words, if a given R/C is -3DB at say 500Hz, would it be like taking a graphic EQ and creating a progressive slope from 500Hz towards the lowest guitar frequency?

-3 db is also known as the half-power point. So, if at a certain frequency, where the impedance of the cap is equal {roughy equal} to the impedance {or resistance} of the resistor, half of the signal voltage is dropped across the cap, and the other half of the signal voltage is dropped across the resistor. Now, since they are both driving the same input impedance of the following stage, that impedance has little effect on how these two parts interact.

-g

[Enzo]

When something starts to roll off, to slowly fade down, however you want to describe it, you have to set SOME point at which you say the thing has happened. For example, as the sun sets, the sky gets less and less bright. At SOME point we start to call it night time. So we could pick some level on a light meter.

Jusy so with things like audio response. AS the rolloff response curves lower and lower, we picked as a standard the point on the curve where it is 3db lower as the point.

[Steve Conner]

My previous post already answered daz's questions. To recap, there are two time constants and two 3dB points. As you go down in frequency, it starts to roll off 6dB/octave, then it flattens out again at some lower frequency, which isn't necessarily the lowest guitar frequency.

This lower frequency is what the resistor across the cap sets: the higher one where the rolloff starts is determined by the other resistances in the circuit.

[Rob Mercure]

Steve,

Ever thought of becoming a copywriter for a cobbler?

Rob

[hasserl]

The formula is F = 1/(2*p*r*c)

I've got an Excel spreadsheet with the formula plugged in, all you have to do is plug in the r and c values and it will spit out the F. Anyone want a copy shoot me an email I can reply to with the spreadsheet attached.