GerryE123 wrote:
>
> Hi:
>
> I have heard or owned a number of single-ended 45, 2A3 and 300B amps.
Of
> these amps, I liked the 300B ones the least. With that in mind, I just
> recently came across a website for a foreign amplifier manufacturer that
has
> some interesting thoughts about driving the 300B.
>
> I would appreciate your comments about their ideas. Here's the exact
text
> from the manufacturers website (NOTE: I am copying it as-is, it loses a
bit
> in the translation):
>
> What points we are interested in?
>
> Perhaps there are so many 300B single ended amps that we could
> not discern good ones from bad.
>
> For its natural and mellow mid-range, 300B single ended amp
> is loved so much but its anemic bass and weak high frequency
> response drive out many tube amp enthusiast from entering in the
> Single Ended World.
The bad ones lack a decent full range response, usually because of
a poor OPT, such as many used in ancient cinema amps.
Combined with a poor load and op conditions, sound wasn't wonderful.
But with decent OPT, the 300B or any other triode in SE mode is just
fine
and the the idea 300B amps and other triodes just cannot do bass,
midrange and treble
all simultaneously and as well as each other; ie, do any type of music,
is a complete myth,
ie, a pile of ideolgical bull****.
However, the performance from a single 300B is limited to what around 7
watts max can do.
It was designed and produced originally for theatre amps which powered
horns with
typical 100dB sensitivity, and without a full F range.
But most ordinary dynamic speakers in normal boxes are 90dB/W/M,
and you cannot expect the same dynamics as you would get if you had a
horn speaker.
With today's low sensitivity, you'd need multiple parallel tubes.
THEN there is never going to be a problem with dynamics and people
ba****ng
huge drums and so on.
Try a 7 watt transistor amp. They will flounder just like a lone 300B
if you drive it beyond its capability, which so many people do
when making a silly comparison.
I just auditioned a pair of 845 in parallel for 60W max.
Utterly no clipping, oddles of bass and treble, and the OPT
allows 8Hz to 50kHz, -3dB points, even without NFB.
300B will do just the same!!!
>
> What are the problems?
>
> Basically there has been no good driving force for 300B.
> First, Western Electric's 91B circuit was improvised for
> easy manufacturing (for delayed production of 86B) and higher gain
for
> a theater system.
>
> It's driving tube 310B ( or 6SJ7 or 6SH7 ) is pentode, notorious for
> its distortion (too sensitive of its screen grid condition ) and
cotton
> sugar bass response and too high internal impedance (about 100K?!!
> bad frequency response in low and high spectrums).
Forget the old ancient ways of driving 300B, 845, 2A3, et all.
Use a 6SN7 with both triodes paralleled.
Have Ea = 250V at least, and Iadc for the two triodes totalling 8mA at
least.
Have a dc carrying resistance of 20k and a 70H choke in series to the B+
of 520V approx.
As F rises, the L ensures the 6SN7 does not have to work hard
putting AC power into the anode feed resistance.
The reactive properties of the choke are isolated from the anodes
because of the 20k.
The 300B being driven should be cap coupled to the driver anodes and
should be cathode biased in the normal way
with R&C, with C > 330uF. Bias R should be around 220k, which will be
the
lowest load the 'SN7 has to drive when ZL is high in the mid F.
Since the load is so high compared to the low Ra of the 'SN7 = about 5k,
you'll get a wide Vswing,
at very low THD.
>
> Second, there are two stage driver circuit. Some use 6SN7 twin
triodes
> one triode for 1st stage and 2nd stage for driving stage. Alas! 6SN7
> is not so powerful for 300B and it's high gain of two stage is not
> suitable
> for 300B power amps, so they sell them as integrated amps.
> Its major defect come from capacitor coupling, which can transfer a
few
> percent of driving wattage to 300B grid, almost of its energy is
> converted
> into useless heat of 6SN7's plate resister.
Use a second paralleled 6SN7 for an input tube if you want sensitivity
and to be able to apply some loop FB, usually 10dB is plenty.
>
> There are also a power drive circuit using EL34 or 6V6 or we349A (triode
> connection).
>
> They are not really 'power' drive, why? They use big plate resistor
for
> EL34 about
> 10k? 20w, but it generates big heat only and could not drive 300B
> sufficiently due
> to the next coupling capacitor, in theory it transfers only few
> percentage of its
> power.
>
> There are only two methods to consider.
>
> First, there is interstage transformer coupling and then plate choke
> drive method
> (which will be discussed in our 300B integrated amplifier). A
interstage
> coupling
> method is a ideal one for transferring driving power to 300B grid
for
> its full
> operation.
The driving "power" is negligible, because there isn't any grid current
in 300B operation.
You just need a linear voltage source with low Ra to overcome Miller C.
Se above about combined L&R to get dc supply.
Interstage trannies are very difficult to make with 8Hz to 60kHz BW.
The usually don't allow loop NFB because of te phase ****ft of the Cshunt
and Lp.
>
> Its driving wattage could be transferred about 93% (less some coil
and
> core loss of
> interstage transformer). We need sufficient power to make 300B work
of
> its full
> capacity.
>
> You will be surprised to the difference between under driven
operation
> of 300B and
> full driven one. But it is really difficult to get a good interstage
> transformer.
We all know that....
>
> There three points to check for interstage transformer.
>
> 1. Must be wide range. Minimum 20Hz - 20KHz FLAT at designated
source
> impedance.
> 2. Good response to 10KHz square wave for flexibility and
musicality.
> 3. Low distortion at low frequencies.
>
> What do you think? Thanks!
The whole business of an IST is difficult.
The parafeed I mentioned above is the best I have used,
wide BW, with slight shelving of response at LF and HF,
and no added iron cause distortions at any F.
One can also use a µ-follower which also works well,
or use a pentode for the dc anode supply with extremely high AC
impedance.
I like the parafeed choke plus R method because it works, and sounds
well,
and has electronic simplicity of elements, even though
you would need a choke with 25mm square core section and have
7,000 turns of fine wire, and just the right gap size.
There is one other way, just make a B+ suply at say +750V, and have the
anode
dc supply resistance for 8mA to be about 70k, well rated for power, made
up of
3 x series R of about 22k each, so the R won't suffer over voltage.
The higher R load will be more easily driven then 20k.
I have used EL84 in triode which is the best IMHO, because
it has gain = 18 when loaded with choke + RL, and with Ia = 15mA,
Ea = 300V, with supply voltage about 500. R would be approx 12k,
and choke = 50H.
Ra will be about 2k2, or equal to about 5 x 1/2 6SN7 triodes paralelled.
An input tube is still needed, so use a 6CG7.
I have 3 x EL84 in parallel to drive the 845, which needs 127vrms max.
The 3 x '84 can produce 160Vrms at 2% THD with parafeed, and at 127Vrms
THD is about 1.4%
At normal listening levels they make about 20Vrms, with THD less than
0.4%,
all of which naturally cancels output tube THD because its all mainly
2H,
so all listening can be done with THD < 0.05%, even with only 8dB loop
FB.
There are several very good ways to build triode amps.
Use the brains and the soldering to find out.
Buy as much good iron from Sowter or Lundahl or someone else as you can
afford.
Don't be afraid of global NFB. Its not a monster, but a technique
to allow the beauty of the music pass without the added artifacts to
spoil anything,
and mainly to ensure the output resistance of the amp gives good control
over speakers because they all have been designed to be driven by a
voltage source < 1 ohm.
My experience with NFB is positive, but many dislike global NFB around
triode amps
and its often because they have no clue how to make sure the amp remains
stable
under any load condition.
I have heard audiophiles say after I have demonstrated my SE amps that
"This proves NFB is OK around SE amps".
Patrick Turner.
>
> Gerry
|
