Propeller Head Plaza Technical and scientific discussion of amps, cables and other topics.

---

[ Follow Ups ] Thread: [ Display  All  Email ] [ Propeller Head Plaza ]

---

[ Alert Moderator ]

---

Re: I fixed a bias problem in my Mosfet follower (experiment #1 continued)

Posted by mfc (A) on June 30, 2003 at 20:02:34

In Reply to: I fixed a bias problem in my Mosfet follower (experiment #1 continued) posted by kurt s on June 30, 2003 at 17:43:49:

Hi,

I thought I'd put these results up here since the
other thread below was starting to get branch rot.

The Cathode follower simulation was similiar to the
Common cathode simulation done earlier. A 6c45 Spice model,
ILoad = 19.778 ma. To choose the B+ for the Cathode follower
I used this formula:

B+(of common cathode stage) - RLoad(common cathode)*ILoad
=> 270 - 3K*.019778 = 210.666 and plate is tied directly
to B+.

This formula provided an operating point for the CF that was as
close to the operating point of the Common cathode stage as
possible (for comparison purposes).

Trying to look at the influence of feedback (in large amounts)
on a Cathode follower is kind of silly. With a max gain of 1,
feedback will reduce this. For instance, just 6db of
feedback will reduce the gain to .5.

Something that is interesting is that in a Common
cathode stage with an unbypassed cathode resistor, feedback
is increased by raising the resistor value. In a Cathode
follower, the opposite is true. The 0 feedback case occurs with
an infinite current source (19.778 ma), and you *LOWER*
the cathode resistor value - just the opposite of the other stage.
This provided a bit of an AHA described later.

The experiment used these values of cathode resistors:

infinite (ideal CS), 900, 700, 500, 300, 200, 150, 100, 50.

corresponding to a gain of:

-0.18,-0.49,-0.53,-0.58,-0.73,-1.09,-1.52,-1.94,-2.79,-5.66 dB

and a feedback factor of:

0.00,-0.31,-0.35,-0.40,-0.55,-0.90,-1.34,-1.76,-2.61,-5.48

and an input signal of:

1.4,1.4508,1.4572,1.4654,1.4916,1.5535,1.6326,1.7145,1.8909,2.6304
volts peak (not peak to peak)

Looking at the raw data, the harmonics increase (get worse) as you
apply small amounts of feedback (lower the cathode resistor). Thus
a CF will typically operate in the range of increasing harmonics as
load is increased.

Here are some of the results for the CF with a
100 ohm cathode resistor:

Fundamental (dB):0.00
2nd :-30.27
3rd :-46.44
4th :-61.29
5th :-75.69
6th :-90.05
7th :-104.64
8th :-119.75
9th :-135.95
10th :-154.71
11th :-181.83
12th :-181.91
13th :-187.41

Here are some of the results for the Common cathode with a
100 ohm unbypassed cathode resistor:

Fundamental (dB):0.00
2nd :-32.10
3rd :-46.53
4th :-60.46
5th :-73.62
6th :-86.69
7th :-99.66
8th :-112.89
9th :-126.40
10th :-140.53
11th :-155.42
12th :-173.92
13th :-184.99

You can see the Cathode Follower (CF) has a bit more 2nd harmonic,
and also a bit more upper harmonics, but in both cases the
results are pretty good. I don't think these results would rule
out use of a cathode follower.

The infinite current source results proved interesting. Here are
the results for a CF with an ideal current source and no load:

Fundamental (dB):0.00
2nd :-117.92
3rd :-174.72
4th :-188.35
5th :-183.13
6th :-182.87
7th :-184.55
8th :-194.35
9th :-186.02
10th :-182.23
11th :-184.33
12th :-186.55
13th :-184.62

Essentially no harmonics!

Here are the results of a CF with an ideal CS and a 100 ohm
resistor load:

Fundamental (dB):0.00
2nd:-30.35
3rd:-46.57
4th:-61.45
5th:-75.89
6th:-90.28
7th:-104.89
8th:-120.00
9th:-136.24
10th:-154.64
11th:-181.63
12th:-184.87
13th:-181.01

Notice that they are the same (very close) to the 100 ohm resistor
with no load case. This is as it should be since they have the same
equivalent AC circuit.

AHA discussion
--------------

At this point I realized that the load of a cathode follower is
part of the feedback loop. Perhaps this is where problems can
occur. Usually a CF has a big ugly output coupling capacitor...
it is part of the feedback loop! If this cap has any reactance,
the feedback is not going to be exactly 180 degrees at that
frequency. Trouble.

The Spice model I used is very linear. Since real CFs are less
linear than this model, you have to remember, that a CF is
operating at a place in the feedback curve where increasing the load
causes an increase in harmonics. I also noted that in the case
of a CF, 2nd harmonic *gets worse* (increases) with increasing
feedback not gets better (decreases), as it does with a Common
cathode stage.

Summary
-------
A CF by itself, is not a recipe for disaster. However, loading
them very lightly gives the best performance. As load is increased,
since the load is part of the feedback circuit, it pays to make
sure the load a CF sees is well behaved. DC coupling would be a
recommendation.

Further Thoughts
----------------
How lightly loaded are the output stage CFs in a Solid State
amplifier with respect to this discussion?
Since CFs are used as output stages, and speakers are the load,
how well behaved can this arrangement be?

This post is made possible by the generous support of people like you and our sponsors:

Schiit Audio

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---