Multiple Coupling Caps+Dropping R's?

[Daver]

http://www.schematicheaven.com/gibson.htm

I've seen on several schematics where there is a coupling cap followed by a resistor to ground followed by another coupling cap and resistor to ground and so on. One schematic I'm referencing is the trem/pentode channel of the Gibson GA20T. What is the purpose of doing this? I've seen this mostly in vintage amps with trem channels, but also in modern designs. Usually the modern designs have different values for the coupling caps and resistors to ground and only two or three sets of them. That, anyway, seems to work to drop signal and change the low frequency roll off. Any input is appreciated.
Dave

[Don Moose]

Cascading filter sections like that (notice the values are repeated) sharpens the filter curve.

A single pole high pass filter of this type (Butterworth) will roll off at 3dB/octave. Cascading 5 stages gives you a 15dB/octage slope. I'm guessing this is to keep the trem oscillator waveform out of the amplified signal.

Hope this helps!

[Daver]

Don,
That indeed does help. You explained it perfectly. I knew there had to be a reason! Thanks a bunch.
Dave

[Earl]

Here's a simulation of that particular filter (AC Sweep) - http://home.texoma.net/~flhh/ga10t.jpg

[d95err]

A single pole high pass filter of this type (Butterworth) will roll off at 3dB/octave. Cascading 5 stages gives you a 15dB/octage slope.

A single pole passive filter rolls off -6dB/octave. The cutoff frequency is defined as the frequency where the signal is at -3dB.

For practical purposes, the signal can be considered flat at approximately 10 times the cutoff frequency for highpass filters and 1/10 for lowpass. This is why e.g. cathode bypass caps often have huge values with cutoff frequencies way into the subsonic region.

[kg]

That, anyway, seems to work to drop signal and change the low frequency roll off. Any input is appreciated.
Dave

incidentally, swap the locations of the Rs and Cs and you end up with a typical power supply rail, which is a beefy version of a low pass filter.

[Paul P]

Why would you want to remove the low-frequency component from
the signal when you just put it there with the tremolo circuit ?

Paul P

[Earl]

d95err - I had read somewhere, that the -3dB point on a cathode follower (bypassed) should be 1/10th, but never actually thought about why. Looks like you answered that!

According to the response sim I did, the -3dB point is ~400hZ, and does become flat at ~4khZ. Seems like an odd high pass circuit for a guitar amp ... ?

[Enzo]

Paul, the point of a trem circuit is NOT to put low freq in the signal path. it is there to modulate the signal at a low frequency.

Imagine someone holding a note while back at the mixer you sit there running the channel fader slowly up and down. That is what a trem does. If they stop playing that note, and you continue to move the fader up asnd down, no low freq sound comes out - unless the fader is scratchy.

meanwhile, the last thing you do want coming out the speaker is a large WUP WUP WUP from the trem. A high pass filter will eliminate that but leave the guitar signal. Most mixers of any sophistication have a low cut button on each channel. That can help eliminate stage rumble coming up the mic stand, some wind noise, etc.

[d95err]

According to the response sim I did, the -3dB point is ~400hZ, and does become flat at ~4khZ. Seems like an odd high pass circuit for a guitar amp ... ?

Hmm... strange indeed. Are you sure you simulated right? Your graph is rather difficult to read since the Y axis is in Volts. Usually in these kinds of graphs, the Y axis is represented in decibels. It would be much easier to read with a logarithmic Y axis (the X axis is already logarithmic as it should be).

[d95err]

...more about the Gibson GA20t tremolo channel.

The first 4 stages have a cutoff frequency of about 34Hz. The fifh and last stage however has a cutoff frequency determined by the position of the volume control.

Since the volume control is wired "backwards", as in a Fender 5e3, the impedance is not constant. At full volume, we get the same 33.8Hz cutoff as the other 4 filter sections. When the resistance is 100k to ground (probably at about 12 o'clock on the volume since it's a log pot), we get a cutoff of 340Hz. So, the amp will be extremely bright at low volume and then fatter as you turn it up.

[Daver]

Enzo,
I think you hit the nail on the head. That makes the most sense, to keep the trem "thump" out of the signal. It's still interesting to me that they used successive roll offs to tailor the signal rather than just use a very small coupling cap.
Thanks to all of you guys that have replied. It's been thought prevoking.
Dave

[Enzo]

But as they explained above, the multiple stages of filter create a sharper rolloff. A small cap would start the rolloff at a much higher freq. This way you keep the bottom end of the guitar signal.

[Daver]

Enzo,
Yes, I got that too. You guys explained it all very well. Thanks.
Dave