Re: The single ended myth



Bret Ludwig wrote:


To compare single ended versus push pull amp topologies we must have
a similar power supply for both. Some years ago Manley/VTL built a
two-300B amp that could be jumpered for parallel single ended or push
pull operation. A homebrew predecessor used a single 6AS7. In every
case push pull operation was unquestionably better.

Its a very old tired argument to state that SE always sounds worse than
PP.

Well, this amp sounded better PP than SE, although if not pushed hard
the two were indistinguishable in many situations.

Depending who is at the tiller to steer the design parameters the PP/SE
outcome
can be variable. Nearly all PP designs are actually class AB, and don't
have the theoretically
exact same class A power. The class A pwoer of SE is *always class*
ONLY.
But PP can be ALL class A to clipping above a certain high load value
that is rarely used,
and below this value the po is AB up to clipping, and depending on
driver
arrangements with 6AS7, the PO can be surprisingly high especially in
class AB2.
But at 1/2 a watt of average power into each speaker, very few ppl would
pick the difference,
if the designer has done a good job. Then there is the issue of NFB and
how much and how its used.

The clients I have sold SE amps with 25 watts or more are not
complaining, and find the SE
give them very fine sound. The 13E1 amps use global NFB, and measure
less than 0.5% at 25 watts,
and 0.05% at the 1 watt level, mainly 2H. The SE35 with a quad of
EL34/6CA7 have far lower thd,
and less than many PP amps because the driver thd cancels the output
stage thd, and
thd is overall much lower than the 13E1 despite quite low levels of
global NFB.
If anything, ppl like the SE35 better than the 13E1, but when I have
been present during comparisons,
there isn't much of a difference; both give superlative sound. The power
is THREE times what a lone 300B would give,
and a lot more than a pair of 6AS7.

Once you get over 25W the issue of power ceases to be much of a concern
unless you are the head banger type.
None of my civilised clients are like this.



And it is NOT automatically true that PP sounds better than SE.

One example of a unique design which gives extremely low THD without
using
shirtloads, trouserloads, and singlet loads of NFB but which is purely
single ended is at

http://www.turneraudio.com.au/se35cfbmonobloc.html

Choke input PS are not included in any of my designs because they are
all sunstantially
class A in operation with little change to Idc during normal operation.
I have learnt to make fine sounding and very well measuring CLC type PSU
with large amounts of cheap and easy to use/source C and a small
size/value non expensive choke
to get ripple voltage to less than a few millivolts at the CT of a OPT
in a PP amp
or one end of an SE OPT.

The secret in getting low interference levels of spuriae caused by high
peak charge currents in CLC
is to carefully set out the earthing paths and FB wiring, and to keep
mains and other PSU
wiring well away from input terminals.

It helps to use a lot of copper and to follow the cloaca principle,
the ground impedance must be a fraction of PS impedance.

Not sure I follow your cloaca principle.




Choke inputs convey no particular inherent benefits, but can be useful
to use
in re-engineering exercizes to reduce the B+ using the original PT.

It is much cheaper to buy another power transformer than to use a
choke filter instead of capacitor except in junkbox projects. I like
to buy new properly made stuff for serious use: junkboxing of course
is fine for experimenting.

In a recent re-engineer of a Woodham CR Audio developments amp, I would
have had to get a
toroidal just the right size to suit the case or wind a new PT to suit.
I already had a suitable spare choke.

It wasn't hard to fold a box and pot with molten pitch.


An engineer, it was said, is someone who can do for tenpence what
any fool could for ten pounds. (Obviously, in pre-"decimalisation"
Britain.)


CR Audio of the UK are fuckwits whose work I was forced to re-engineer.
Or else the owner would have been stuck with a very unreliable amp on
which he'd paid
about USD $4,500.00: anyway, a full re-build point to point wiring and
redone internal metalwork
fetched me a grand, worth every cent, or "tenpence" or whatever.

I am currently doing a 1996 ARC VT100, maybe a months full time work so
far,
but the guy will be pleased, and won't have such a ridiculous number of
smoky sound lounge events.

OPT are very good, shame about the silly load matching, absurd
complexity,
terrible layout, difficulty to service, etc.

The more I see from world renowned fucking "engineers" of hi brow tubed
audio leads me to believe
these makers have lost the plot; its never enough to just settle for
enough power,
simplicity, and reliablity; they always go for absurdly high power,
tubes teetering on the edge,
badly loaded driver and output stages, and too many bells and whistles.

The so called fool who charges 10 pounds to do what an engineer does for
tenpence may
not be such a great fool, of what I see of so called engineers is any
guide.



Choke filtered PS was used because of its better load regulation. As
long as the critical current was pulled the voltage is pretty well
constant to the design limit. I suggest you may wish to read, horrors,
a uniquely American document-the old ARRL Handbooks.

I have several ham radio Handbooks, including the British one which in
my opinion is superior to
the US books.

I know how to do a decent choke input when i think its not out of place.


Peter Millett has
at least one up for free (expired copyright). Capacitor input even
with oil filled caps was highly discouraged because of poor load
regulation and consequent distortion. These were all Class B
amplifiers as the popular mod tubes were a pair of 811s.


I normally like to use CLC, with mild C1, enough L, and humungous C2.

Ripple is in mV, and resonance is less than 7Hz, and any amp, SE/PP,
can work from what is a very low supply impedance.



Usually the RF finals and the mod ran off the same B+ supply. By the
time popular audio types could put out enough power on lower voltages,
SSB had come in. That was a mistake but is also another story.

I'm stuck in the AF band have not moved up to RF yet; maybe I never
will.

Patrick Turner.
.