From Patrick Turner below: "because they are substantially class A. Not
totally class A, but more class A than nearly all other amps in the output
stages."

This is called Class A/B1.
Just cause a tube conducts for 360 degrees when it's delivering some amount
less than full power (below the onset of clipping) doesn't mean that it's
operating as Class A at that point.

Lloyd


in article , Patrick Turner at
wrote on 2/22/04 3:31 PM:



Miles O'Neal wrote:

Patrick Turner responded to WakyAmps, who said:

Yes, how it is set up makes a huge difference, no a Vox AC-30 isn't
really class A.

The ones I have serviced were class A, with absolutely no feedback.
They had a high bias current, and each of the 4 EL84 draws about 12 watts
at least at idle.

Now this doesn't make any sense at all.
In Class A, full tilt isn't different
enough from idle to make any difference.

The input power to the 4 x EL84 is around 48 watts.
In an ideal class A amp, a max of 24 watts in class A is available.
But usually, class A max efficiency is about 40%, so
expect about 20 watts of class A from 4 x EL84,
with a few extra AB watts to make it to about 30 at onset of clipping.
Such operation is substantially class A.
In a Musical Reference hi-fi amp made in the US, Ea = 700v, Eg2 = 350v,
and power output at clip from a pair is 36 watts, but its very definately
mainly class B, with only a couple of class A watts.
Idle power is about 5 watts per tube. During low level operation,
there is much more thd than for amps with gobs of class A power..
But nevertheless, the use of a high value RL seen by the tubes of 14k,
and 5% of CFB from a winding on the opt, and global FB, the amp
measures ok.

Many EL84 amps are less ambitious about power,
with a quite a few good for 21 watts, such as Mesa Boogie,
some Fenders with 6V6, where the B+ is high enough, as well as
the load value, to get such power, which is well in excess of what you'd get
if every watt was class A.
The Fender I fixed last week got 21 watts AB1 from 6V6, with 450v supply.
BIG voltage swing, low current swing, and mainly class B operation.
The cut off behaviour of tubes is such that the crossover distortion
of such amps is not entirely awful, certainly not in a guitar amp,
where distortions are prized, and there is no such thing as really
clean operation, because it sounds so plain dull and boring.

One is forced to allow some harmonic distortions, principally 3H,
and then boost the hell out of all the treble, add some reverb,
and maybe it sounds well.
Most guitar amps are PP mainly class B in their output stages.
Very few *pure* class A PP output stages are ever used, and
*extremely* few single ended class A output stages.
But all the input stages are nearly all SE stages using 12AX7,
so much of the warmth comes from the SE triode input stages which
are over driven a bit, and the distortion happens to be musically appealing,
( with a high level of 2H and 4H ), even to old smoothies, like George Benson.
But don't ask me what Mark Knoffler uses in his amp circuit to give him his
trademark sound.



So, just curious - why do you say they
were Class A? What do you think that
means?

because they are substantially class A.
Not totally class A, but more class A than nearly all other amps
in the output stages.



...
As I see it, all the 6550 from yesteryear about which ppl rave as being
just so superior to
anything else are all actullay prone to failure from hard use, unfortunate
events with the wrong speaker impedance, etc, since all the ratings
indicate that they are what they are, mortal tubes, and like us mortal
human beings,
they are gonna die one day.

While this is true, some of the tubes made back
in the far exceeded their published specs, and
hence are worth far more than the common tubes
of today, to some people. If it'll last five
times as long when being punished, it's worth at
least five times the price - more if you calculate
the lower likelihood of failure during a gig.

But none of the tubes I have seen will take 5 times the
power dissipation of an average spec tube.
There is *no* 6550 with an anode diss rating of 200 watts,
and nealry all turn red at just over 45 watts, and then its time for
prayer, or to reach for the off switch.

I actually like NOS US made tubes.
I have a pair of ancient 40 yr old Sylvania 6CA7,
which in UL still pump out 28 watts of class AB at
0.125% thd with 17 dB of NFB in an ancient ex school PA
amp which I use regurlarly in my shed to test speakers etc.
It withstands the occasional red plate session, when left running into
a load which is too low. Done that a few times.
Accidently had the leads come together for awhile with a signal,
and sure, it got hot, but good as new after being turned off for
5 minutes.
These tubes were old and tired when I bought 7 years ago sh for $3 each,
but they keep truckin. I know I ain't likely to
ever blow up a nice pair of customers speakers I happen to be testing,
simply because I'm usin a transformer coupled amp, and there is no nasty
DC to come a leapin outa the box to fry a driver.

I see no reason why today's tubes won't survive the "industrial use".

I have had one sovtek KT88 get a broken heater connection on a tube pin,
and this made the other in the PP circuit do all the work, and it went red
for quite some time before the customer noticed, ( customers are
non techs, and notoriously slow to realize something is wrong ),
but after a resolder, the overheated tube seemed undamaged.



I'm merely suggesting
that, while for Rich there's no other tube than the GE and no other amp
than the PS400, the OP is looking to retube a reissue SVT and the OP may
be quite satisfied with the Sovs (at half the price no less). And it is
the OP's needs we're trying to address here, right?

Sure. I agree with you here. At the same time
I understand where Rich is coming from. He's a
firm believer in the value of the old tubes over
anything built today (so far), and wants to help
others understand that.

In some case I agree with him on the state of the
art and the value; in others I don't. Most of the
time we're in different threads, because I mostly
deal with 6BQ5s and 6V6s, which he dismisses as
"radio tubes". 8^)

The littlies have a very nice sound though, and its why dudes use them.
For a small venue for a jazz scene, maybe its all that's needed.




I assure you that other folks make big tube amps. As new old stocks of
favourite old tubes dry up, all we are gonna have is sovtek, EH, Svetlana
etc.
But asking more than 75 watts per pair of bottles is asking for trouble.

With the new tubes this may be so - I don't know.
But I have known lots of folks getting 100W or more
per pair (think SVT!) with no problem.

Back in the 1960's, quite a few amps using EL34 with
800 volt B+ supplies were used, mainly for PA, because
100 watts was available, and low THD wasn't a priority.
I probably sat thru many a sermon ( boring ) at the church
where I was taken each sunday as a kid with such amps.
Where are all those amps? all gone to the Yonder!
And these were all mainly used for just a bit of pious speach,
not kickass bass with serious overdrive.
Tube and socket replacements due to arcing were common,
so one reason they all were replaced was the spectacular
failures, costs of maintenance, and anyone selling the then very
new, and even rather expensive solid state amps had little trouble.
I don't believe they sounded all that much worse in the PA field.
Most of those have all expired and been thrown out.
Nothing is forever.

...
Remember, Rich claims that the GE and Sov "aren't even close" sonically
or electrically. Maybe so in his tweaked 400PS and to his ears and
playing style; there may not be as strong a case in the OP's RI-SVT and
to the OP's ears and style.

How anyone can tell much difference between the sound of tubes when the
thd is at around 25%, and IMD at around 50% is beyond me.

Just curios - are you a musician?

Not now; I was in a band many years ago,
and played acoustic 12 string. I did some time in coffee lounges
with folk music; I can stil play, but I have little time.
I often have to fix guitar amps, and I hear the musician's problems
and the sound, during testing....

Do you care about
tone?

Yes, and mainly at low levels, because at gross overload levels,
the waveform is mainly a square wave, and the difference in tube brands
does not make an enormous difference to the sound
of such energy being supplied to a speaker.

Do you consider your ears good?

Good enough.

Because the
sonic differences in tubes are ell documented. My
wife and kids, none of whom play guitar (Josiah keeps
saying he wants to, but never practices much) can hear
these differences, and they aren't tube freaks by any
stretch of the imagination!

I hear differences, sure, but mainly at low levels.
The "dark metal" guy who likes his 6550 in triode runs them at
gross overload all night. I don't believe the sound would change
if he changed from Sovtek to GE.
To me it would still sound like a continuous aeroplane crash
happening right in front of me.



[Story of 0 ohm tube PA into 1 ohm bins snipped]

In this case, its better to err on the load value by choosing a value too
high, rather than too low!

Why not try your friend's PA head into, say, 64
ohms at full blast and let us know how that goes?

If you have a 400 watt amp capable of 400 watts into 64 ohms,
then using 64 ohms would give only a fraction of the power,
and the sound level, because the voltage swing from the tubes is
nearly maximum even with 8 ohms, at 56 vrms, and
with 64 ohms, only 20 more swing is available, ie 67 volts,
so one gets only 70 watts. But the class A % would be quite high.
Pentode/tetrode operation is almost universally used for output stages
in the music industry in tube amps.
Distortion in such amps with a high RL is not necessarily
all that much lower than some lower value RL which is the optimal for
low thd. But high value RLs mean the output stage has much higher gain.
If there is a feedback loop, then there is effectively much more applied FB
with a high value RL.
The gain of a 6550 in class A with a 2 kohm plate load is about 19.
But with 16k its about 80+, or five times the gain, which is + 14 dB.
If there was 12 dB of FB applied with 2k, then there is about
5 times the feedback applied when the load is changed to eight times the load
value.
Thus with 16k, there is an effective 26 dB of FB applied.
One has to be careful about stability if using a high value load, lest the amp
begin to oscillate at some LF or HF.
In practice, since most amps operate mainly in class B,
the open loop gain is about 1/2 what I have described, so the max amount of
FB with a 64 ohm load would be 20 dB at high power.
But at low power, the gain is higher, since it is class A.
This explains why some poorly stabilised amps oscillate just a little
and no more, because they do so only in the operating region where the tube
gain
is high.

Speaker impedances rise to high values at HF, because of their
inductive character.

So, I would never recommend that anyone use a 64 ohm load with
a pentode/tetrode amp which was designed for 8 ohms.
16 ohms would be OK though, but 4 ohms could be deadly.
To ensure stability, even with no load at all, when the output tube
gain is perhaps 150+ then the feedback loop has to be carefully
considered, and some way of damping, ie, loading the output stage
at F above 20 kHz should be employed,
say by having an RC network with HV rated C and large wattage R
strapped across each 1/2 primary.
Some ancient amps using V1505 power triodes with 2 kV B+,
good for 1,100 watts class AB2 used adjustable spark
gaps to allow arcs to occur if the anode to anode signal voltage exceeded a
certain
level. Little FB was used. That was the primitive approach.
But with pentodes or tetrodes, unlike triodes, the gain changes with load,
and often some FB is used. Well concieved RC gain limiting networks are
good practice.

The idea of using multiple parallel KT88 or 6550
isn't new for high power.
There was a design described in full in a 1957 edition of Wireless
World, which easily tested at 400 watts with 10 x KT88.
All the parts were mounted on a piece of 3/4" plywood,
about 24" x 9", breadboard style. Class A % of the power was
substantial.
Probable use was for a stadium, theatre PA.
Poor struggling musos could never afford such gear, and all the speakers
that went with the amp, and the van to carry it all.
There were 10 ohm resistors to each cathode of the KT88, with a meter which
could be switched
to each 10 ohms to check the bias currents.
The tube idle power dissipation was kept low, and to allow for
uneven bias drift; such amps often have only one fixed bias voltage,
and one has to allow for one or more tube's idle current to
mysteriously escalate during its life.
Nevertheless, the tests of the 1957 amp showed it could sustain
the use of lower load values briefly, and that it was capable of
well over 500 watts.

I believe the prudent designers then would never have considered the use of
only

6 x 6550 to do the job.

It'd be like flogging good men to death.

Patrick Turner.



-Miles

-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----