Fleetie wrote:
"Nick Gorham" wrote

tony sayer wrote:

which leads you
to directly heated triodes, then you have to consider the fillament
supply, DC hums less, but AC sounds better.


Do they still make directly heated valves and surely AC would hum like
no ones bizz, or was this some wind up?....


What I don't understand was someone's earlier assertion
(I paraphrase): "DC heaters hum less but AC sound better."

HUH?! Why would an AC heater supply make the amp sound better
than a DC?

Well, first thing, this is with directly heated valves, secondly it just
does seem to do that in practise, and thirdly I will see if I can
explain my theory about this, but its just that my theory.

If you look at a typical output stage using a DHT, where the fillament
is also the cathode, you will see the power to heat it applied, then a
couple of resistors to create a center point, and this is then taken
down to ground or through a cathode resistor.

Now firstly you can see from this that the fillament supply is directly
connected to the output stage of the amp, so any components in the power
supply to the filament can play a part in the amp sound. AC supplies are
normally just a transformer secondary connected to the fillament, but DC
supplies are normally a TX, maybe some plain slow diodes, and a big old
electrolytic, maybe better diodes and maybe a PI filter or matbe a
regulator. but whatever is used its connected to the output stage, so
any switching noise or inductances of the caps all can get into the act.

Secondly, if you look at the voltages on the fillament, one side will be
at a higher voltage than the other, by the heater voltage, no this means
that there is uneven bias along the length of the fillament, so I would
expect a uneven electron flow. Now I don;t know if thats actually taking
part in the difference, but I can't see how it can help. If you take a
extream example, of GM70 running at 100ma current, and the fillament
being connected via the normal two 100R resistors, the 20v filament
supply seemd to me to mean that o current flows in one leg, and 200ma in
the other.

This is not to say that you can't have a DC supply that doesn't suffer
from the first problem, just that it will take a bit more care.

Another interesting area thats being investigated is to use hi frequency
(20k) ac to driver the heater.

--
Nick