[Phil Allison]

** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395

Grid voltage = - 78

Cathode voltage = - 56

Cathode current = 3 mA

Plate dissipation = 1.35 watts.

Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

What happens to the 6550s when the heater cathode insulation fails during
warm up ?

( The above IS what happened with this one. )

Also - what 12AU7 is the most likely to be OK here ?


...... Phil

[Patrick Turner]

Phil Allison wrote:

** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395

Grid voltage = - 78

Cathode voltage = - 56

Cathode current = 3 mA

Plate dissipation = 1.35 watts.

Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

Not very long with Ea = 460V and more during warm up.


What happens to the 6550s when the heater cathode insulation fails during
warm up ?

If the 3 x 6550 grids on each side of the PP circuit are direct coupled
to the CF cathode,
then a short between cathode and heater at 0V biases the 3 tubes at 0V
thus causing
about 400mA Ia flow per 6550 and probable fuses blowing if you are
lucky.
But worse is where the 'AU7 ****s itself fully and B+ becomes connected
to
6550 grids... aaahhh.



( The above IS what happened with this one. )

Also - what 12AU7 is the most likely to be OK here ?

None are suitable with Ea so high.
What is needed is that the 12AU7 anodes are connected to about +180Vdc
so that Ea = sum of the +180V and the -50Vdc
cathode voltage, ie, about 230Vdc. Tubes can also arc internally with Ea
too high
and ARC fuctards have 6922 set up similarly with Ea = way too high in
VT100 and the result
is a stream of customers who suffer silence and smoke and who then hate
ARC.

But such multiple direct coupled grids is always asking for trouble
unless active protection
circuits against excessive dc cathode currents.
I think I am the only person who fits active dc protection in tube amps.
The occasional high anode current of say 400mA at peak power during AB2
bursts
of music don't damage the tubes, but its the sustained high wrong dc
bias and
Pda that quickly wrecks OP tubes if it isn't prevented.

I do not like direct coupled output tubes.

Its done to get
the extended anode load voltage swing
by means of driving grids positive and into grid current regions during
class AB2 operation.
McIntosh also use the idea.
I have had to fix a few re-issue McIntoshes.

But if you wish to keep the direct coupling, then use a much more rugged
tube
such as a 12BH7 with the same pin out as a 12AU7.

It should just drop straight in without adjusting anode loading on the
gain triode halves, and gain should be similar or slightly less than the
12AU7.


The right way to drive output tubes with direct coupling
is to have a biased ct choke to bias the grids, and driven by the CF,
and a heater
winding biased to the -ve ct voltage and a safe Ea at say +150V from a
supply
with a series R so that is there is a short, the B+ is pulled down,
resistor fries open,
but grid bias remains healthily negative. Bean counters
would never do it this way though.

If the AMPEG has C&R coupled grids, so that class AB1 is used rather
than AB2,
then the problems of the 12AU7 you have won't root up the outputs.
But the sound will be horrible of course, so reduce Ea like like I said
above
and use a 12BH7.

To reduce Ea, use a dropping R from the main B+ and a 470uF cap to
ground.
A string of 3 x 5W zeners can be used to regulate the B+ and prvent it
starting off high at turn on. Make it so Idc in the zeners = Ia in
normal op mode.

With CR coupling of 6550 grids from the CF cathode, you should have 6
caps and six pots and six
bias R with 10 ohm Ia sensing R between each 6550 cathode and 0V.
if there is only one bias voltage applied to all 6 x 6550, some will be
running hot
and others running cool, and the hot ones fail first.

I've never had any success telling the yanks how to build their amps.

I get a lot of work rewiring their junk.

Patrick Turner.






..... Phil

[Phil Allison]

"Patrick Turner"
Phil Allison wrote:

** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate
about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s
with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395
Grid voltage = - 78
Cathode voltage = - 56
Cathode current = 3 mA
Plate dissipation = 1.35 watts.
Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

Not very long with Ea = 460V and more during warm up.


** During a normal warm up ( ie amp on standby) plate voltage = zero, but
cathode supply = -195.


Also - what 12AU7 is the most likely to be OK here ?


But if you wish to keep the direct coupling, then use a much more rugged
tube such as a 12BH7 with the same pin out as a 12AU7.

It should just drop straight in without adjusting anode loading on the
gain triode halves, and gain should be similar or slightly less than the
12AU7.


** Thought of that, just after making the post.

Original 1970s Ampeg SVTs used a 12BH7 to do the job.

Tried a ( Soviet) 12BH7A already, worked like a charm.

Specs say it has a -200 volt cathode / heater rating too.



....... Phil

[Patrick Turner]

Phil Allison wrote:

"Patrick Turner"
Phil Allison wrote:

** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate
about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s
with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395
Grid voltage = - 78
Cathode voltage = - 56
Cathode current = 3 mA
Plate dissipation = 1.35 watts.
Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

Not very long with Ea = 460V and more during warm up.

** During a normal warm up ( ie amp on standby) plate voltage = zero, but
cathode supply = -195.

Also - what 12AU7 is the most likely to be OK here ?


But if you wish to keep the direct coupling, then use a much more rugged
tube such as a 12BH7 with the same pin out as a 12AU7.

It should just drop straight in without adjusting anode loading on the
gain triode halves, and gain should be similar or slightly less than the
12AU7.

** Thought of that, just after making the post.

Original 1970s Ampeg SVTs used a 12BH7 to do the job.

Tried a ( Soviet) 12BH7A already, worked like a charm.

Specs say it has a -200 volt cathode / heater rating too.

Plenty of NOS 12BH7 around because it is not the darling tube of the
hi-fi cognescenti.

But its a beautiful tube and used in countless arduous conditions as
line oscillators etc in TV sets.

Many CF have a resistor from the cathode to some B- supply, and -200V is
typical,
but if a short happens to the heater fed by ct 6.3 at 0V then the
cathode immediately
goes to 0V. But maybe only during warm up as the tube will still begin
to conduct
and micro thermal movement may break the short and the cathode falls
back to say -60V approx,
and maybe not enough to make the tube short while its warm.

The better practice is to have the CF heater tied to a dedicated heater
winding with CT
via 220k resistor and have a 10uF cap to 0V, so the heater winding is
biased
at the same voltage as the cathode, but has its ct grounded for AC.
The low output resistance of the CF 1k, so noise via stray C from PT
doesn't get in.
The 220k loads the cathode, but its OK because probably there is already
68k to -200V there
already, and an extra 220k won't make much difference.

But in countless guitar amps, I have seen cathodes of CF stages where
the cathode is at +200V
and there's a 12AX7 CF driving a tone stack etc. And its very common to
see an
LTP in a Mullard 520 type of design with the cathodes of the
12AX7/12AT7/12AU7 LTP pair at +150V
while the heaters are at 0V.

The insulation used on the heating wire is only a thin layer of oxide,
and can get thinned dangerously by careless workers in Chinese
sweatshops
when they insert the folded heater wires into the cathode tubes.
It all gets down to QC.

Patrick Turner.






...... Phil

[Jon Yaeger]

in article , Phil Allison at
wrote on 7/20/08 11:07 PM:


** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395

Grid voltage = - 78

Cathode voltage = - 56

Cathode current = 3 mA

Plate dissipation = 1.35 watts.

Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

What happens to the 6550s when the heater cathode insulation fails during
warm up ?

( The above IS what happened with this one. )

Also - what 12AU7 is the most likely to be OK here ?


..... Phil



A qualitative answer:

I have an Allen mono amplifier consisting of (4) 6L6GTs wired together
through an odd output transformer arrangement.

The output stage is direct cathode coupled using 6SN7s. With this
configuration the driver selection seemed to be critical.

In my limited experience, I found that the 6SN7s had to be matched and in
perfect working order, or else the plates of the output tubes would glow
orange or red. Some Russian tubes I tried caused problems. Emissivity and
gain seemed to have a lot of impact.

The best results I had were with NOS American made 6SN7s (RCAs in my case).
There are a whole lot of 12AU7s that should work for you, e.g. Telefunken,
Sylvania, etc. I would recommend Telfunkens because they seem to be stable
and their gain is relatively stable over time.

jon

[Patrick Turner]

Jon Yaeger wrote:

in article , Phil Allison at
wrote on 7/20/08 11:07 PM:


** Have an Ampeg SVT-2 Pro ( bass guitar head ) on the bench - it uses
six 6500s in parallel push pull ( class AB) with grid bias to generate about
300 watts. Drive signal to the 6550 grids comes from a pair of 12AU7s with
half of each wired as a direct coupled cathode follower.

The 12AU7's operating conditions a

Plate voltage = + 395

Grid voltage = - 78

Cathode voltage = - 56

Cathode current = 3 mA

Plate dissipation = 1.35 watts.

Heater / cathode voltage during warm up = -195 volts.

QUESTIONS:

How long will a cheap, Chinese short plate 12AU7 survive ?

What happens to the 6550s when the heater cathode insulation fails during
warm up ?

( The above IS what happened with this one. )

Also - what 12AU7 is the most likely to be OK here ?


..... Phil



A qualitative answer:

I have an Allen mono amplifier consisting of (4) 6L6GTs wired together
through an odd output transformer arrangement.

The output stage is direct cathode coupled using 6SN7s. With this
configuration the driver selection seemed to be critical.

In my limited experience, I found that the 6SN7s had to be matched and in
perfect working order, or else the plates of the output tubes would glow
orange or red. Some Russian tubes I tried caused problems. Emissivity and
gain seemed to have a lot of impact.

The use of 6SN7 would be ideal but matching triode sections is not
critical.

But Ea should not be in excess of +250Vdc even though 6SN7 is rated for
a higher Ea.
Don't tempt fate with a too high Ea.

But what is critical is the bias current in the output tubes, and if the
output works in class
AB2 with too low an RL to extract every ugly morsel of power available,
then expect red/orange anodes, especially if the idle dc is too high.
AB2 is rough PA sound. Pda at idle for Ab2 should not be more than 10
watts for 6L6.

In many amplifiers designed by optimistic fuctards, there is no way to
easily check idle bias current
in EACH output tube and no means to adjust it properly; ie, each tube
individually.
And the same OFs seem to think nobody will ever over drive the amp with
a load that's too low,
and there isn't any active protection against Ik going way too high
and causing tube thermal runaway.

Its possible to get 80 watts from a pair of 6L6 in AB2 with Ea =
+600Vdc, and Eg2 = +300Vdc
and some high RL value to allow a wide Ea swing of about 500V peak at
each anode.

The maximum anode efficiency for tubes in AB2 with low idle bias Ia is
about 67%.
If you have 80W of output, there must be 120W of input power, so there
is 40W dissipated in each output tube.
So 20W each tube, 2W less than the rated max, but if somebody has a load
too low, say a 4 ohm speaker instead of 8 ohms,
then Pda easily exceeds the 6L6 22 watt Pda rating, and at suswtained
levels the tubes will fail.

On the same basis, you can get 140W from a pair of 6550/KT88 in AB2.
Or 100W from a pair of EL34.

As long as you have plenty of spare tubes its OK and you don't mind them
blowing rather often.




The best results I had were with NOS American made 6SN7s (RCAs in my case).
There are a whole lot of 12AU7s that should work for you, e.g. Telefunken,
Sylvania, etc. I would recommend Telfunkens because they seem to be stable
and their gain is relatively stable over time.

12AU7 is only what it is and I wouldn't trust one with Ea at over 400V
as in Phil's Ampeg repair.
12BH7 is better, but Ea should be reduced.
And 6SN7 is OK, but is damned octal, and can't be used in Ampeg unless
the
socket is changed which isn't necessary because other suitable tubes are
available.
6CG7 long plate are also OK as these were made sometimes with the *same*
anode/cathode/grids
of 6SN7 but crammed into a nine pinner tube, and slightly de-rated for
Pda.
But 6CG7 wouldn't suit the Ampeg because its heater pin out is different
to 12AU7.

BTW, The NOS Oz made 6CG7 is a really beautiful sounding tube. Don't
tell anyone.

Patrick Turner.






jon

[BretLudwig]

"With CR coupling of 6550 grids from the CF cathode, you should have 6
caps and six pots and six
bias R with 10 ohm Ia sensing R between each 6550 cathode and 0V.
if there is only one bias voltage applied to all 6 x 6550, some will be
running hot
and others running cool, and the hot ones fail first.

I've never had any success telling the yanks how to build their amps.

I get a lot of work rewiring their junk."

The original SVT used a different driver tube that was well specified and
6146 outputs. Bill Wyman got 10+ years out of the original set.

The current production is a St. Louis Music/Gene Kornblum bodge. Done as
cheaply as possible.

We can't even straighten them out,and it's unlikely you'll succeed
either. But you are right, a lot of crap is built here-along with
everywhere else.

--
Message posted using http://www.talkaboutaudio.com/group/rec.audio.tubes/
More information at http://www.talkaboutaudio.com/faq.html

[Phil Allison]

"BratLudwig"


The original SVT used a different driver tube that was well specified


** The 12BH7 - as previously noted.


The current production is a St. Louis Music/Gene Kornblum bodge. Done as
cheaply as possible.


** The SVT-2 Pro retails here for a cool $A 5500.

The amp is full of crappy dodges that no tube instrument amp should ever
have.

Eg :

Ribbon cables with IDC plugs connect grid bias to all the OP tubes.



...... Phil

[Patrick Turner]

Phil Allison wrote:

"BratLudwig"


The original SVT used a different driver tube that was well specified

** The 12BH7 - as previously noted.

The current production is a St. Louis Music/Gene Kornblum bodge. Done as
cheaply as possible.

** The SVT-2 Pro retails here for a cool $A 5500.

The amp is full of crappy dodges that no tube instrument amp should ever
have.

Eg :

Ribbon cables with IDC plugs connect grid bias to all the OP tubes.

..... Phil

Phil, the Chinese invasion is all but complete,
and mopping up operations are underway for the period of occupation for
the
forseeable future.

Nearly all musician amps are now festooned with all manner of effects
boards and do-dah switches
for gee-whiz effects to make up for the lack of any real playing skills
of performers and the
godawful meningless lyrics of the songs being sung.
The medium is the message now.

In many amps there are second generation PCBs with circuitry full of
tiny opamps and surface mount parts and one dare not try to repair
anything
except by replacing a whole board, but unhitching the ribbon cabling
and installing a new board, and plugging cables back.
Somehow I get the feeling that NONE of what is made now when
ppl of 20 turn 60 in 2048 will have any collector value at all.
Who will treasure a CP player in 2048?
Who will undertsand how one works?
Who will be able to repair a CD player?
Same will go for the thousands of guitar amps now made.

A computer with enormous complexity is remarkably reliable,
but in guitar amps and other gear consumers are treated like suckers.
Just about all electronics is passe, outdated, obsolete well before it
"wears out".

A late model Behringer combo amp I have here for repair is made in China
to a standard enforced by Germans. $399 retail here.
Lotsa features, and it does nearly everything except give a BJ.

But why the **** is this stuff in my workshop for a fix?

They just don't think of everything they should have.
Like, if there is a way a dude can **** it up, the dude will find a way,
even when he isn't tryin.

But the really bright electronics designer folks wouldn't be
seen dead designing guitar amps; no kudos. Only dumb counts
design such lowly crap, one step up from a toaster.

Its impossible for any one of us to have the slightest effect on the
quality
of globally produced kraperology.
They will all do it *their way*, and not ours.
Its always been that way.

Patrick Turner.