Raymond Koonce wrote:
Patrick Turner wrote:
"Mikkel C. Simonsen" wrote:
Raymond Koonce wrote:
Hi RATs,
Here's a link to a project I've had on the bench for a while and got
finished today. I learned quite a bit on this one.
http://www.timebanditaudio.com/47%20Amp/Carmen47.html
It sounds as good as it looks.
Comments, questions?
Looks great - as usual 
Best regards,
Mikkel C. Simonsen
The specs indicate 1.5mV of hum.
In horn speakers with 100dB/W/M that 1.5mV would plainly be audible
if the horns were that sensitive at 100Hz, and the hum was 100Hz hum.
Has any attempt been made to reduce the hum?
Patrick Turner.
Hi Patrick,
My speakers are fairly sensitive, Klipsch Cornwalls, rated 101dB/W/M.
The hum (60 Hz probably from filaments) is not audible to my 50+ year
ears from closer than a normal listening distance. If I stick my ear in
the woofer, I can hear some, but it's not objectionable. I am running
my 47 filaments on AC with hum pots and all I've done is to adjust the
hum pots to the lowest AC at the speaker terminals.
My PSU is LCLCRC and is very quiet, no 120Hz hum at all.
Suggestions?
Best,
Raymond
A guy here built a pair of tratrix profile horns with Fostex full range
drivers
and the result was high sensitivity and horrible sound, as though it was
fighting its way out
of a wooden box.
He tried his own generic SS amps which didn't make any difference, then
my class AB triode amp
using PP 6CM5 in triode, and that made no difference either; the main
problem wasn't the amp,
it seldom is, it was with the speakers.
But my triode amp was at that time silent with normal speakers because
hum was just below 1mV.
But it was clearly and anoyingly loud with these horns, and if any amp
has power supply switching noise
creeping in somehow from the diodes, and this is very common, then you
will hear
a buzz of some sort with horns.
If the sole harmonic component of the hum is 60Hz, then it will not be
as noticeable as 120Hz or
180, 240, 300Hz etc, because as F rises above say 20Hz the noticeablity
to our ears increases.
Anyway, My triode AB amp hum became noticeable and I was forced to
improve the B+ filtering
by making alterations to the regulated B+ supply; it has regulation
rather than CLC.
Having made this alteration the hum became much lower by about 12dB, and
I can only assume its noticeablity with horn speakers was
reduced to below audibility; it was not re-tried again with horns.
Any amplifier with a low amount of NFB is especially prone to having
some noise at its output
even with the input grounded.
NFB generated within the amp circuit enclosed by NFB which would include
any heater leakage from filaments to
cathode, or any unbalanced/uncancelled heater voltages across any dh
cathodes is all reduced by the NFB
in a similar way as the THD/IMD etc is reduced by NFB, so if there is
say 3mV of hum before any FB is applied,
then 20dB of NFB reduces that hum to 0.3mV, a respectanble figure for
any power amp.
If a tube amp is run without any loop NFB, and hum from heating
filaments or dh cathodes is the culprit, then
one answer is to use well filitered floating dc supplies for all and
ground these supplies well to 0V
with large electrolytics, which is easy because with low dc cathode
voltages involved,
a 470uF rated at 10 to 63V isn't a large cumbersome item.
Needless to say, SE amps need good CLC filtered anode supplies to the
output and driver stages.
But usually dc applied carefully around cathode circuits and anode
circuits should reduce
hums and buzzes to neglifible levels. I say should, and sometimes the
amp is still noisy
with well smoothed supplies.
Then is a case of revising earth paths, reducing magnetic induction and
reducing
stray capacitances until noise is minimized. This is most difficult to
achieve if the PS is on the same chassis as the
amp. But not impossible, and sometimes some trial and error with leads
and item positions
need to be done with a CRO connected to confirm any noise reductions
that are achieved.
Eventually one learns never to have critical leads running under power
trannies, and that no ferrous chassis
like copper and Al and brass are worst for magnetic bothers, and that
screening cables only works against
electrostatic stray C noise pick up.
One star earthing point is a must, located near the input.
The chassis should be grounded directly to the earth lead of the 3 wire
mains input lead,
and it also should then connect to a 0V buss by say 22 ohms, so that
preamps and the like
don't form earth loops.
If it is impossible to have all DC heating, then the AC heating should
have an adjustable
hum balance pot; a wire wound 250 ohm pot across the 6.3V with wiper
grounded or taken to the
R&C cathode bias network will allow hum adjustments.
Patrick Turner.