Coupling cap breakpoint
From Dr.Distortio–(at)–bs.mhv.net Tue Aug 22 21:42:11 CDT 1995
From: Dr.Distortio–(at)–bs.mhv.net (Dr Distortion)
Subject: Re: BASSMAN amp not giving any bass ?????
Date: 22 Aug 1995 15:44:28 GMT
Chris Johnson (jinx656–(at)–over.net) wrote:
: capacitances involved, as well- however, this could possibly damage
: rectifiers when the amp is powered up, particularly solid-state ones (mind
: you, I’m not sure of that)
Actually, it’s more of a danger when using tube rectifiers. The spec
sheets for many of these specify a maximum input filter capacitance value
which, if exceeded, could cause the tube to fail.
: Note- overkill can lead to problems. My tube stereo amp won’t handle
: over I mf signal-path caps to the power tubes without going onto
: very-low-frequency oscillation.
1 uF is :huge: for a coupling cap in a tube amp, with its high impedances.
0.1uF is more typical of a “large” coupling cap.
: Also, the cutoff point tends to be emphasized for some reason- I have
: found that extending the bass to severely low frequencies actually reduces
: the amount of perceived ‘fatness’ because the LF cutoff is too low to
: resonate (?).
It could be because the amp is wasting energy amplifying sub-audio
frequencies. One possible danger with using coupling caps that are too
big is that the negative grid bias on the tubes could be “swamped out” by
AC signals that are so low in frequency, they look almost like pulsed DC.
A little math saves a lot of trouble. A coupling cap and a grid
resistor, for instance, are in fact a lowpass filter:
What we’re concerned with is finding the “breakpoint” of the filter, or
the frequency at which response falls by 3dB. The formula for this is
For convenience, I use this derivative: 159000/R*C. You can plug in
capacitance in microfarads and resistances in ohms this way…
Let’s take the coupling caps and grid resistors in the output section
of a typical silverface Fender as an example. You usually find .1 caps
and 100K resistors, so…
…equals 15.9 Hertz. That’s a pretty low frequency, when you consider
that guitar speakers generally roll off at about 50Hz, and the guitar
itself probably doesn’t put out fundamentals in that range.
The approach I’ve described is a simplified one. If you wanted to be
hardcore about it, you would factor in the plate resistance and plate
load of the preceding stage. But this is rock ‘n roll, man… 🙂 🙂 🙂
Anyway, my point is that you shouldn’t go through your amp putting in
larger coupling caps at random. The .02 through .1 caps you’re likely to
find in there are more than adequate to pass your lowest signal,
especially since following grid resistances are usually much higher than
100K or so; 220K through 1Meg are common. If your amp doesn’t sound
“bassy” enough to you, it might not be so much that bass isn’t being
amplified, but rather that treble is being over-emphasized. You should
also take your speaker into consideration; some speakers are really
There are many ways to tame an overly-bright amp. You can increase the
negative feedback at high frequencies by bypassing the “upper” feedback
resistor with a small cap. Don’t do this unless you have access to a
scope and know how to look for instability, though, since you’re
basically changing the frequency balance of the feedback circuit and
there could be unpredictable results. A safer way would be to add some
simple capacitor-shunt filters in the preamp; for instance, across the
plate resistors. It’s not the most elegant approach, but I’ve used it
with success on amps that sounded too shrill.
For the adventurous, there’s always the option of adding a conjugate
load to the output section.
This is a resistor and cap in series across the output transformer
primary; you calculate the resistor value by multiplying the
plate-to-plate load by 1.3. The cap can be selected empirically; the old
RCA books said that .05uF should be right in most instances, although
some folks prefer to use caps as small as .002uF. The idea behind this is
that while the reflected impedance seen at the primary rises with
frequency, the impedance of the conjugate load falls with frequency. If
the components are chosen correctly, they should balance each other out,
with the net result that the tubes are seeing something closer to a
resistive load. This will tend to balance out the frequency response of
the output section; and, in many amps, will give you more (perceived) bass.
But this is not a project for beginners… They, perhaps, would be better
off trying small caps across the preamp plate resistors before
considering more drastic measures. Or, simply turning down the treble
control 🙂 🙂 🙂