I'm putting together an 807 push-pull amp. Desired power out is 25W or
so.

Plate supply (5AR4, choke-input) is about 370V-390V as current drawn
goes from 150mA to 90mA. It goes up to 470V with no load. The
transformer and choke are rated at 150mA but I've run this supply at
150mA for a whole day and it's just getting lukewarm, so I think I can
draw a little more on peaks without any big problems.

Grid bias (-20V or -25V) will come from a separate transformer
winding/pot.

For the screen supply, it looks like my options and
advantages/disadvantages a

0. Ultralinear. Won't work because plate voltage is much higher than
rated rated screen voltage (300V). This is where an 807 differs from,
say, a 6L6GC type it seems. Elsewhere in some 807 data sheets they rate
triode-mode operation (plate strapped to screen) to 400V, can this be
extrapolated to saying that it's OK to run from ultralinear taps at
400V? How about at the 470V that my power supply puts out with no load?

1. Dropping resistor from B+. Not good, because variation with plate
voltage and screen current drives up distortion considerably.

2. A stiffish voltage divider from B+ supply. Appears to not be
particularly stiff (screen current will probably go back and forth from
5 to 20mA depending on signal) although some electrolytics can probably
clean this up. But "stif" here means probably 60mA (three times the
peak screen current) through the divider chain and that's eating up a
lot of my power transformer's capacity. Is 60mA too stiff, not stiff
enough, ???

3. Separate transformer winding/rectifier/filter. Could do it if I had
such a transformer and filter, but no more space on the chassis...

4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing.

#4 looks like like the way to go. The delta in screen current I read
from the 807 curves make it look like 5-20mA is the right amount for
two tubes, and this just happens to be the range over which some VR
tubes hit the application exactly.

Are there any gotchas with using VR tubes here? If one of the VR tubes
fails to fire then they don't regulate at all, but a good choice of
resistor to B+ will limit the screen dissipation to a survivable amount
if this happens, right? Or am I looking for a complete lightning storm
if a VR tube doesn't light?

I don't see a lot of other designs using VR tubes for screen supply in
audio amps, although in my long experience as a ham this is very often
seen in RF stages.

Are there gotchas with power-on voltage sequencing that my choice of
tube rectifier (5AR4) is going to cause?

The B+ current drain goes up by another 25 or 30mA or so by getting
screen voltage from B+, but I think I can handle that much more current
drain there.

Did I miss the boat completely in terms of guesstimating required
divider stiffness in #3?

Tim.

wrote:
I'm putting together an 807 push-pull amp. Desired power out is 25W or
so.

Plate supply (5AR4, choke-input) is about 370V-390V as current drawn
goes from 150mA to 90mA. It goes up to 470V with no load.

That's what choke input supplies are supposed to do.
You need a bleeder resistor or something else to keep
a minimum load on the PS to keep that from happening.
If you have an old ARRL handbook, I think there's something
in there about calculating the resistor value.

For the screen supply, it looks like my options and
advantages/disadvantages a

0. Ultralinear. Won't work because plate voltage is much higher than
rated rated screen voltage (300V). This is where an 807 differs from,
say, a 6L6GC type it seems. Elsewhere in some 807 data sheets they rate
triode-mode operation (plate strapped to screen) to 400V, can this be
extrapolated to saying that it's OK to run from ultralinear taps at
400V?

Yep. Screen ratings for ultralinear are usually about the same
as triode. In this case, I don't recall any published specs
for UL for 807's, but you can figure they'd be about the same as
5881 or 6L6-GB. There have been production amplifiers using
807 as ultralinear with similar voltages.


How about at the 470V that my power supply puts out with no load?

Like I said, if you've got a bleeder in there, that won't happen,
and it'll go away as soon as the 807's warm up, anyway.


3. Separate transformer winding/rectifier/filter. Could do it if I had
such a transformer and filter, but no more space on the chassis...

That's been done before. The advantage is that the screen voltage
won't be affected much by the plate transformer load.

4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing.

You can do that like RCA (VR tubes as a shunt across the screen
to ground), or like a Leslie (VR tube in series with the screen).
In case 2, the screen voltage will track the plate voltage, but
since yours is pretty well regulated, that shouldn't be a problem.

Are there any gotchas with using VR tubes here? If one of the VR tubes
fails to fire then they don't regulate at all, but a good choice of
resistor to B+ will limit the screen dissipation to a survivable amount
if this happens, right? Or am I looking for a complete lightning storm
if a VR tube doesn't light?

Nah, shouldn't be a problem. Just remember not to bypass
the VR tube directly with a big capacitor

I don't see a lot of other designs using VR tubes for screen supply in
audio amps, although in my long experience as a ham this is very often
seen in RF stages.


Here's a couple:
http://www.triodeel.com/mi9377.gif
http://www.triodeel.com/a340a.gif



--
Ned Carlson
SW side of Chicago, USA
www.tubezone.net

Ned Carlson wrote:
wrote:
I'm putting together an 807 push-pull amp. Desired power out is 25W or
so.

Plate supply (5AR4, choke-input) is about 370V-390V as current drawn
goes from 150mA to 90mA. It goes up to 470V with no load.

That's what choke input supplies are supposed to do.
You need a bleeder resistor or something else to keep
a minimum load on the PS to keep that from happening.
If you have an old ARRL handbook, I think there's something
in there about calculating the resistor value.

Or, of course, the VR tubes with dropping resistor will draw a good
constant
30-40mA out just as well, as would the stiff voltage divider I
postulated.

0. Ultralinear. Won't work because plate voltage is much higher than
rated rated screen voltage (300V). This is where an 807 differs from,
say, a 6L6GC type it seems. Elsewhere in some 807 data sheets they rate
triode-mode operation (plate strapped to screen) to 400V, can this be
extrapolated to saying that it's OK to run from ultralinear taps at
400V?

Yep. Screen ratings for ultralinear are usually about the same
as triode. In this case, I don't recall any published specs
for UL for 807's, but you can figure they'd be about the same as
5881 or 6L6-GB. There have been production amplifiers using
807 as ultralinear with similar voltages.

I look around and I see lots of warnings to not simply translate 6L6GB
designs to the 807 because the 807's screen is so wimpy. That said I've
heaped lots of abuse onto 807's in ham stuff over the years and only
occasionally did they ever explode with blue sparks!

4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing.

You can do that like RCA (VR tubes as a shunt across the screen
to ground), or like a Leslie (VR tube in series with the screen).
In case 2, the screen voltage will track the plate voltage, but
since yours is pretty well regulated, that shouldn't be a problem.

It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Here's a couple:
http://www.triodeel.com/mi9377.gif
http://www.triodeel.com/a340a.gif

Thanks!

Back in the 70's and 80's with my ham stuff that used VR tubes (usually
OA2's) I discovered that a lot of them were quite unreliable in
starting up. These were mostly consumer-type tubes that were either
original equipment or came from the TV repair shop in town. Lately I've
amassed some collections of surplus military JAN OA2's and other
regulators and have had MUCH better luck with these reliably firing. I
don't have much experience with the octal VR bottles, do they follow
the same trend with regards to reliability of firing/starting up?

Tim.

I look around and I see lots of warnings to not simply translate 6L6GB
designs to the 807 because the 807's screen is so wimpy. That said I've
heaped lots of abuse onto 807's in ham stuff over the years and only
occasionally did they ever explode with blue sparks!

Don't forget that ham radio transmitters on average have low duty
cycles. Long periods of time listening to a ham receiver interrupted by
occasional transmissions. At worse it would be about 50% duty cycle
when a ham is calling "CQ" (ham jargon for an invitation for anyone who
wants to reply to do so). Audio amps are obviously running at 100% duty
cycle (when considering the several hour period of listening in an
evening). So hams could get away with overloading their tubes as they
would have a chance to cool down during the off periods.

On 24 May 2006 10:25:25 -0700, wrote:


4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing. snip

Don Leslie built tens of thousands of cheap 6550 amps with OC3s in
them using this same idea...trouble-free (unless the OC3
croaks...which is never) and keeps the screens right where they ought
to be.

In your case, two in series (OC3 + OD3) would not only be the best
practical solution, but would give you a very nice violet glow, thus
making your 807 amp the envy of every audiophool on the planet. OC3s
are dirt cheap, I dunno about OD3s. I think I still have 20 OC3s
around here...somewhere.

The ideal way to use the 807 is with a tertiary screen winding and a
lower screen supply or in regulated screen tetrode mode, but regular
tapped ultralinear works fine if you have a decent sixzed resistor on
the screen tap. You want to solder this resistor on the socket itself.
It acts as a stopper against parasitics also.

VTL did this on their very successful early amps.

If you use a separate screen supply you must ensure you never have
screen voltage applied before the plate voltage or you will have
excessive screen dissipation.

A pair of push pull 807s with a reasonably good OPT will provide the
optimum in listening at reasonable cost. Most 807 amps sound good and
give good tube life.

DeserTBoB wrote:
On 24 May 2006 10:25:25 -0700, wrote:


4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing. snip

Don Leslie built tens of thousands of cheap 6550 amps with OC3s in
them using this same idea...trouble-free (unless the OC3
croaks...which is never) and keeps the screens right where they ought
to be.

In your case, two in series (OC3 + OD3) would not only be the best
practical solution, but would give you a very nice violet glow, thus
making your 807 amp the envy of every audiophool on the planet. OC3s
are dirt cheap, I dunno about OD3s. I think I still have 20 OC3s
around here...somewhere.

Several organ amps ran regulated screen using gas discharge tubes. I
once had a half dozen Leslie amps sitting around after a company I
worked for solid stated all the Leslies in a huge black church in
Oklahoma. (Actually the church was red sandstone: the congregation was
black.) We used the transformers in guitar amps, particularly the
outputs. When they went broke the chassis and relays all got
dumpstered. I think there was a mod to replace the gas tube with a
solid state affair using a zener string and a transistor.

On 24 May 2006 19:08:44 -0700, "Bret Ludwig"
wrote:

Several organ amps ran regulated screen using gas discharge tubes. snip

The later Leslie 6550 amps did. The first ones did not, and the
screen potential would rise above the plates under hard drive, and
then would lunch a pair of T-S 6550s. The next year (1956) Don Leslie
put the OC3 regulator circuit in the design to get rid of this
problem. His 6L6 amps never needed screen regulation.

Some moron rockers would get rid off the OC3 to "make it louder, man"
which meant "more distortion!" Usually, a fried pair of 6550s would
result in a couple of hours, and would usually take the fuse with it.
These were NOT very good amps fidelity wise...the transformers were
cheap, cheap, cheap...the design was a Muntz sort of exercise...least
amount of elements to make it work. At 40 watts, a Leslie amp in good
shape will yield about 10% THD, and that's with matched 6550s.
Leslie's previous 40 watt amp (actually 35), a quad 6L6 that was
probably copied right out of the RCA tube manual, was actually a
cleaner amp. At the time, it was considered, like with gee-tawrs,
proper to add as much THD, hopefully even ordered, to "juice up" the
otherwise dull, boring tone of the tonewheel organ. 10% of it, plus
$2 tweeters for the treble horn and a cheapie Jensen P15LL woofer, and
the standard in aftermarket Hammond organ speakers was born.

With NOS Tung-Sol 6550s going for $250 a pop, it pays to keep a couple
of spare OC3s around for a buck a pop. In the 807 proposal, he'd need
an OC3 and an OD3 in series to get what he needs.

I think there was a mod to replace the gas tube with a
solid state affair using a zener string and a transistor. snip

Three 35V zeners in series, period, to replace the OC3. With an OC3
usually being cheaper than the zeners, and OC3s generally never going
bad, it never did make too much sense.

On 24 May 2006 19:08:44 -0700, "Bret Ludwig"
wrote:


When they went broke the chassis and relays all got
dumpstered. snip

It might pain you to realize that good Leslie 6550 amps now go for
around $200 apiece, more if rebuilt.

wrote:


Or, of course, the VR tubes with dropping resistor will draw a good
constant
30-40mA out just as well,

Yeah, I though of that after I posted.

I look around and I see lots of warnings to not simply translate 6L6GB
designs to the 807 because the 807's screen is so wimpy. That said I've
heaped lots of abuse onto 807's in ham stuff over the years and only
occasionally did they ever explode with blue sparks!

Not only that, like a gazillion AM commercial transmitters used 807's
as modulators and RF drivers. That kind of service is CCS on steroids,
commercial stations typically like as close to 100% modulation
as they can get, so the 807's are running constantly right to
design limits, if not further.



It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.


Back in the 70's and 80's with my ham stuff that used VR tubes (usually
OA2's) I discovered that a lot of them were quite unreliable in
starting up. These were mostly consumer-type tubes that were either
original equipment or came from the TV repair shop in town. Lately I've
amassed some collections of surplus military JAN OA2's and other
regulators and have had MUCH better luck with these reliably firing. I
don't have much experience with the octal VR bottles, do they follow
the same trend with regards to reliability of firing/starting up?

If they're constantly taking a dump on startup, SOMETHING is
causing them to draw way too much current. Someone mentioned
Leslies, I've never heard of a VR tube croaking on a Leslie
unless one of the 6550's had a dead screen short. Never heard
of a VR tube in an RCA amp crapping out unless it was worn out
or just bad, there's resistors in series that keep them from
drawing too much current (them RCA injuneers was thinkin', wasn't
they?).




--
Ned Carlson
SW side of Chicago, USA
www.tubezone.net

DeserTBoB wrote:
OC3s are dirt cheap, I dunno about OD3s.

OD3's are currently $2.30 NIB from AES. That is cheaper than dirt
cheap, man!

In any event, now that I'm putting this all together I'm finding that
the tube costs are completely inconsequential compared to the iron
(transformers+chokes). I happen to be using some transformers and
chokes that I already own so I don't have to pay "new" prices, but
still the tube costs are tiny. Hey, the electrolytics (450V Nichicons)
are only a few dollars each too but they start adding up to more than
tube costs pretty quickly.

Thanks for some of the reassurances about VR tubes not crapping out
very often. As it is I think the choice of the dropping resistor at the
top of the chain can be made such that the screens don't actually arc
over or melt down in case the VR's don't all fire up.

Tim.

Ned Carlson wrote:
wrote:
It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.

Now that I'm drawing up the schematics and comparing it to bleeder
values from the ARRL handbook it looks like I'm still not drawing
enough current to really replace the bleeder resistor entirely. But...
with a little luck the 807's will warm up at just about the same rate
as the 5AR4 rectifier and they'll start drawing current just as the B+
comes up.

Back in the 70's and 80's with my ham stuff that used VR tubes (usually
OA2's) I discovered that a lot of them were quite unreliable in
starting up. These were mostly consumer-type tubes that were either
original equipment or came from the TV repair shop in town. Lately I've
amassed some collections of surplus military JAN OA2's and other
regulators and have had MUCH better luck with these reliably firing. I
don't have much experience with the octal VR bottles, do they follow
the same trend with regards to reliability of firing/starting up?

If they're constantly taking a dump on startup, SOMETHING is
causing them to draw way too much current.

The problem I always had is that the OA2's (and similar) would start up
very unreliably after just a couple years of service. I had to replace
them way more often than the finals or any other tube.

I don't think the Heathkits etc. abused the OA2's all that badly. I
think I just had really crappy quality tubes from the local TV repair
shop.

Now that I'm getting back into tube radios/amps again after a couple of
decades on the sidelines I'm amazed at how much I was paying for crappy
tubes at that TV repair shop in the 70's and early 80's. Today I can
find JAN/milspec/industrial tubes for a fraction of the price.

Tim.

said:


Thanks for some of the reassurances about VR tubes not crapping out
very often. As it is I think the choice of the dropping resistor at the
top of the chain can be made such that the screens don't actually arc
over or melt down in case the VR's don't all fire up.


When using regulator tubes in series, place a resistor of about
100....220 kohm over each tube.
That way, they'll start up for sure.

--

- Never argue with idiots, they drag you down their level and beat you with experience. -

On 25 May 2006 05:09:41 -0700, wrote:

The problem I always had is that the OA2's (and similar) would start up
very unreliably after just a couple years of service. I had to replace
them way more often than the finals or any other tube. snip

MIL/industrial gas regulators in mini form are far more reliable than
the "consumer grade" crap. Octals seemed about the same all around.

I don't think the Heathkits etc. abused the OA2's all that badly. I
think I just had really crappy quality tubes from the local TV repair
shop. snip

Remember, most "TV shops" would mark up tubes around 250%.

Now that I'm getting back into tube radios/amps again after a couple of
decades on the sidelines I'm amazed at how much I was paying for crappy
tubes at that TV repair shop in the 70's and early 80's. Today I can
find JAN/milspec/industrial tubes for a fraction of the price. snip

Yes, and far better quality!

In article ,
admin-n-att--tubezone-dott-net wrote:

wrote:

It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.

I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value, the peaks in the screen current go higher than that
and you must allow for that in the design of your screen supply. Do not
connect a capacitor directly across the VR tube to accommodate the peaks
or you may shorten its life, and if you are really unlucky end up with a
relaxation oscillator. I think this is the reason that in screen supplies
you always see VR tubes used as a fixed dropping element from the main B+,
or simply as the reference element in an active series regulator. Check
the VR tube spec sheet for design guidelines.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

John Byrns wrote:
In article ,
admin-n-att--tubezone-dott-net wrote:

wrote:

It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.

I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value

I think you are right. Although I would be surprised if the
instantaneous AC currents ever got to twice that.

the peaks in the screen current go higher than that
and you must allow for that in the design of your screen supply. Do not
connect a capacitor directly across the VR tube to accommodate the peaks
or you may shorten its life, and if you are really unlucky end up with a
relaxation oscillator. I think this is the reason that in screen supplies
you always see VR tubes used as a fixed dropping element from the main B+,
or simply as the reference element in an active series regulator. Check
the VR tube spec sheet for design guidelines.

Ah, that may be why I don't see these VR tubes often used as shunt
regulators for AF tetrode screens, and when they are used as shunt
regulators for the screen I now know that I should carefully inspect
how they handle this issue.

For RF tetrode screens, of course, the 5000pF or so bypass is large
enough to be a good bypass at RF and not large enough to cause
oscillation with the VR tubes.

With this tidbit I think I understand the common practices in screen
supplies for AF power amps better than before. Thanks!

Tim.

John Byrns wrote:
I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value, the peaks in the screen current go higher than that
and you must allow for that in the design of your screen supply. Do not
connect a capacitor directly across the VR tube to accommodate the peaks
or you may shorten its life, and if you are really unlucky end up with a
relaxation oscillator. I think this is the reason that in screen supplies
you always see VR tubes used as a fixed dropping element from the main B+,
or simply as the reference element in an active series regulator. Check
the VR tube spec sheet for design guidelines.

I am referencing the schematics Ned supplies in his reply:

http://www.triodeel.com/mi9377.gif
http://www.triodeel.com/a340a.gif

Ah, you are right, in the Leslie it's a dropping element from the main
B+.

In the RCA design two VR tubes in series (OC3+OD3) develop 255V or so
that is used to supply screen voltage to 4 6L6's (each with 100 ohms in
series). The same 255V goes through a 3.9K resistor, gets filtered by a
40uF capacitor, and this gives a presumably filtered B+ to the preamp.

Now 100 ohms in series with a screen current going up and down between
(to pick per tube AC peak currents that are twice as large as the DC
ones) 0 and 20mA gives a fluctuation of only two volts.

That sounds like a perfectly reasonable amount of screen regulation to
me. Now actual regulation may not be that good because 40mA is a pretty
broad current range to get out of the VR tubes, but it's gotta be
better than no regulation at all, right? Or is crappy regulation that
chops out on the peaks perhaps worse sounding than no regulation (say a
beefy divider) with good bypassing?

My impression is that the RCA design is not exactly known for its high
fidelity (it was more of a PA amp application, am I right?)

Maybe over the weekend I can build all these variations and sound them
out!!

Tim.

wrote in message
ups.com...
John Byrns wrote:
I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value, the peaks in the screen current go higher than
that
and you must allow for that in the design of your screen supply. Do not
connect a capacitor directly across the VR tube to accommodate the peaks
or you may shorten its life, and if you are really unlucky end up with a
relaxation oscillator. I think this is the reason that in screen
supplies
you always see VR tubes used as a fixed dropping element from the main
B+,
or simply as the reference element in an active series regulator. Check
the VR tube spec sheet for design guidelines.

I am referencing the schematics Ned supplies in his reply:

http://www.triodeel.com/mi9377.gif
http://www.triodeel.com/a340a.gif

Ah, you are right, in the Leslie it's a dropping element from the main
B+.

In the RCA design two VR tubes in series (OC3+OD3) develop 255V or so
that is used to supply screen voltage to 4 6L6's (each with 100 ohms in
series). The same 255V goes through a 3.9K resistor, gets filtered by a
40uF capacitor, and this gives a presumably filtered B+ to the preamp.

Now 100 ohms in series with a screen current going up and down between
(to pick per tube AC peak currents that are twice as large as the DC
ones) 0 and 20mA gives a fluctuation of only two volts.

That sounds like a perfectly reasonable amount of screen regulation to
me. Now actual regulation may not be that good because 40mA is a pretty
broad current range to get out of the VR tubes, but it's gotta be
better than no regulation at all, right? Or is crappy regulation that
chops out on the peaks perhaps worse sounding than no regulation (say a
beefy divider) with good bypassing?

My impression is that the RCA design is not exactly known for its high
fidelity (it was more of a PA amp application, am I right?)

Maybe over the weekend I can build all these variations and sound them
out!!

Tim.

Hey Tim,
If you have a later RCA tube manual around, check out the 50w amplifier
schematic in the rear circuits section. There is a nifty adjustable series
regulator circuit to consider in that amp. It uses a dissimilar dual triode
6GF7 and an 0A2 (OD3 in octal) VR tube as a reference voltage. I am using
this regulator in at least 2 functioning home builts to date. You might have
to screw around with the voltage divider on the output to get the output
voltage range where you want it. You can use the 13 volt tubes like a 13GF7
or 13EM7 with a Rat Shack 273-1365 heater transformer. The whole thing can
be done for pretty cheap, I think a 13EM7 from AES is about $4, and
Ratformer about $6+ change. The separate filamant trans is necessary because
the cathodes of both triodes are at a significant potential to ground. You
need to have the heater somewhere between the 2 cathode voltages. The single
VR still can function as your bleeder. Good luck with your amp!
MarkS

wrote:

I'm putting together an 807 push-pull amp. Desired power out is 25W or
so.

Plate supply (5AR4, choke-input) is about 370V-390V as current drawn
goes from 150mA to 90mA. It goes up to 470V with no load. The
transformer and choke are rated at 150mA but I've run this supply at
150mA for a whole day and it's just getting lukewarm, so I think I can
draw a little more on peaks without any big problems.

Grid bias (-20V or -25V) will come from a separate transformer
winding/pot.

For the screen supply, it looks like my options and
advantages/disadvantages a

0. Ultralinear. Won't work because plate voltage is much higher than
rated rated screen voltage (300V). This is where an 807 differs from,
say, a 6L6GC type it seems.

807 is a 6L6 with an anode cap.

Elsewhere in some 807 data sheets they rate
triode-mode operation (plate strapped to screen) to 400V, can this be
extrapolated to saying that it's OK to run from ultralinear taps at
400V?

Yes. 40% UL taps are OK with Ea = Eg2 = 450V.
Be safe and have Ea about =400V.

How about at the 470V that my power supply puts out with no load?

Choke input supplies will have a soaring B+ with no load. You should have a
bleeder from OPT CT to 0V so that about 15mA flows at all times.



1. Dropping resistor from B+. Not good, because variation with plate
voltage and screen current drives up distortion considerably.

No need for series R.



2. A stiffish voltage divider from B+ supply. Appears to not be
particularly stiff (screen current will probably go back and forth from
5 to 20mA depending on signal) although some electrolytics can probably
clean this up. But "stif" here means probably 60mA (three times the
peak screen current) through the divider chain and that's eating up a
lot of my power transformer's capacity. Is 60mA too stiff, not stiff
enough, ???

I know when things sure are stiff enough around here but in UL there is no
need for
any divider, stiff or not.



3. Separate transformer winding/rectifier/filter. Could do it if I had
such a transformer and filter, but no more space on the chassis...

No need.



4. Couple of voltage regulator tubes in series (3xVR-90's? VR-150 and
VR-105?) with a dropping resistor to B+. VR tubes seem to be cheap and
plentiful and are even kinda pretty glowing.

No.



#4 looks like like the way to go. The delta in screen current I read
from the 807 curves make it look like 5-20mA is the right amount for
two tubes, and this just happens to be the range over which some VR
tubes hit the application exactly.

Only if you are using the tubes in beam tetrode mode with fixed Eg2.
The Eg supply can be R from CT to +300V with series 3 x 100V x 5w zeners,
run about 20mA down through these without the g2 connected, so if B+ at CT
= 400, R = 4.7k, 10 watt rated.
As g2 draws more current, the zeners draw less.
The G2 supply will also double for a bleeder supply to stop the B+ from
soaring.



Are there any gotchas with using VR tubes here? If one of the VR tubes
fails to fire then they don't regulate at all, but a good choice of
resistor to B+ will limit the screen dissipation to a survivable amount
if this happens, right? Or am I looking for a complete lightning storm
if a VR tube doesn't light?

VR are OK but cannot be bypassed with caps or you get an oscillator.
Zeners are quite OK., and should be bypassed with 100uF.



I don't see a lot of other designs using VR tubes for screen supply in
audio amps, although in my long experience as a ham this is very often
seen in RF stages.

Most designs you have seen were designed by accountants.



Are there gotchas with power-on voltage sequencing that my choice of
tube rectifier (5AR4) is going to cause?

No.




The B+ current drain goes up by another 25 or 30mA or so by getting
screen voltage from B+, but I think I can handle that much more current
drain there.

Did I miss the boat completely in terms of guesstimating required
divider stiffness in #3?

A separate screen supply for +300V is OK but really not needed
if the existing supply can deliver say 400V x 40mA for each 807 at idle and
say 20mA
for the Eg2 / bleeder. That's only 100mA and your driver stages will have
say 20mA, so 120mA all up
I bet with music program you won't see much B+ change or huge Ia increase
with hi-fi but with guitar amp
the current draw will go to about 200mA but only occasionally.

Patrick Turner.



Tim.

Hi Bret,

The ideal way to use the 807 is with a
tertiary screen winding and a lower

This kind of FB (essentially UL) doesn't work
good with 807 because of its aligned screen grid
aligned construction - the screen grid is much
less effective as a control grid than with many
other tubes of that class (say, EL34).

Tom

--
MS-DOS is the worst text adventure game I have ever played:
Poor vocabulary, weak parser and boring storyline.

Tom Schlangen said:


The ideal way to use the 807 is with a
tertiary screen winding and a lower


This kind of FB (essentially UL) doesn't work
good with 807 because of its aligned screen grid
aligned construction - the screen grid is much
less effective as a control grid than with many
other tubes of that class (say, EL34).


Is that why they sound so good in triode mode? ;-)

--

- Never argue with idiots, they drag you down their level and beat you with experience. -

I wouldn't fool with regulator tubes or zeners, OR rectifier tubes
which DO NOT sound better and DO NOT constitute a soft start system.

The 807 will work fine in a conventional ultralinear circuit, triode
connected, or a lower screen supply. If you use a split supply with a
centertapped secondary and a bridge rectifier you get two B+ voltages,
one is twice the other. Use a separate heater and bias supply and
sequence them on the primary side.

Put most of your money in a quality NEW output transformer and a good
stiff clean power supply with good transformers and caps. You just
can't go wrong.

On 27 May 2006 17:39:09 -0700, "Bret Ludwig"
wrote:

I wouldn't fool with regulator tubes or zeners, OR rectifier tubes
which DO NOT sound better and DO NOT constitute a soft start system. snip

Uh...well, running a tube rect. DOES slow down the onset of B+, right?
However, your point is that the rect. is usually conducting before the
fils are hot in the outputs, which is usually true for most beam power
tubes. I used to cobble up a timer circuit and change out from tube
to silicon with a dropping resistor and an octal relay for various
applications instead of a 5U4. Seemed to save a lot of 6550s. One
thing that DOES do, assuming running a single 5U4...less B+ sag when
approaching full power, good for audio, bad for the gee-tawr twangers
who thought their amps didn't "go to 11" anymore.

Put most of your money in a quality NEW output transformer and a good
stiff clean power supply with good transformers and caps. You just
can't go wrong. snip

Sage advice...why buy junk trannies off of fraudBay and build a saggy
thing that's going to torch anyway? Go with new, build it tight.

Bret Ludwig wrote:
I wouldn't fool with regulator tubes or zeners, OR rectifier tubes
which DO NOT sound better and DO NOT constitute a soft start system.

Well, to each his own. While I've screwed around a lot with tubes over
the years this is my first foray into the audio (as opposed to RF)
realm. I made a lot of poor choices in the RF realm, but there a "poor
choice" means that the choke caught on fire most likely. If a poor
choice in the audio world means that "it didn't sound better than the
simple way" then I'm going to be extremely satisfied!

While a couple of folks here obviously like their series-pass tube
regulators for screen voltages I am probably not going to go that way
but stick with something simpler like a shunt regulator made out of VR
tubes. Mostly because I don't want to build a tube regulator, I want to
build an amp.

I do have an old Heath regulated bench supply (internally some OA2's
for reference and 6L6's for pass elements) that I might give a try as a
screen supply at some phase.

The 807 will work fine in a conventional ultralinear circuit, triode
connected, or a lower screen supply. If you use a split supply with a
centertapped secondary and a bridge rectifier you get two B+ voltages,
one is twice the other. Use a separate heater and bias supply and
sequence them on the primary side.

Actually I'm not awfully concerned about soft-starting, I just thought
I'd throw that in as a weak technical justification for using a tube
rectifier. Really I just WANT to use some glow tubes and try out some
of these "new" cheap Russki rectifiers. I also WANT to do a choke input
power supply and see what the ripple currents look like on a scope. I
very likely will be back with some gripes about my poor choices, just
wait and see!

As to bias supply, I've been looking at my 50's vintage ARRL handbooks
and found some truly weird-ass ways of getting grid biases for 807/6L6
class modulator tubes. I should probably make those a separate thread,
as some of those circuits completely stump me.

I did find an ARRL handbook article from 1959 where they use 807's in a
modulator and use a shunt regulator made out of VR tubes for screen
supply. In fact a few months ago that particular article probably made
me start thinking about all this!

As I screw around with different phase splitter styles I'm sure I will
learn something.

Put most of your money in a quality NEW output transformer and a good
stiff clean power supply with good transformers and caps. You just
can't go wrong.

Actually, I am going with new Hammond iron for power supply and output
transformer. Also the supply filter choke. As far as choke-input
supplies go, the Hammond transformers and chokes look to be way lower
DC resistance than the stuff that ended up in most of my ham gear (but
of course part of that is the difference between CCS and ICAS and
"beyond-ICAS" ratings.)

Choke-input supplies in the ham world went out of style in the 40's or
50's for the most part. I strongly realize that I am not blazing new
ground in that respect.

A choke-input supply is also a weak justification for using a tube
rectifier instead of solid-state.

As to capacitors, I wouldn't dream of using old stuff from the junkbox
(and in fact I haven't put old electrolytics in the junkbox for years.)
New Nichicons for me.

Realistically I am probably still at least a week away from getting any
sound of this thing, and I'm probably going to shut up until I get it
done and can actually LISTEN. The result will probably sound abysmal to
everyone else but I'm sure I'll be satisfied (not necessarily from the
sound but from the fact I built it myself.)

Tim.

Mark S wrote:


My impression is that the RCA design is not exactly known for its high
fidelity (it was more of a PA amp application, am I right?)

It was a movie theater amplifier, so the fidelity was reasonably
good (and better than average for the time it was built, meaning
opposed to console radios and the like). There's networks built in to
the RCA system to roll off highs to get rid of "ground noise"
from the optical sound system, otherwise it's OK. I'm not that
fond of the RCA drive circuit, but there's nothing wrong with
the screen regulator per se.

--
Ned Carlson
SW side of Chicago, USA
www.tubezone.net

Tom Schlangen wrote:

This kind of FB (essentially UL) doesn't work
good with 807 because of its aligned screen grid
aligned construction - the screen grid is much
less effective as a control grid than with many
other tubes of that class (say, EL34).

Tom

Perhaps that's why there were no Dynaco
amplifiers using 6L6 or 6V6.


--
Ned Carlson
SW side of Chicago, USA
www.tubezone.net

Ned Carlson wrote:
Tom Schlangen wrote:

This kind of FB (essentially UL) doesn't work
good with 807 because of its aligned screen grid
aligned construction - the screen grid is much
less effective as a control grid than with many
other tubes of that class (say, EL34).

Tom

Perhaps that's why there were no Dynaco
amplifiers using 6L6 or 6V6.

I've had lower measured distortion with 6L6s in modified ST70s than
with the stock EL34s, using straight ultralinear operation. Well, it
was a nice theory, I guess.

Tom Schlangen wrote:

Hi Bret,

The ideal way to use the 807 is with a
tertiary screen winding and a lower

This kind of FB (essentially UL) doesn't work
good with 807 because of its aligned screen grid
aligned construction - the screen grid is much
less effective as a control grid than with many
other tubes of that class (say, EL34).

Tom

--
MS-DOS is the worst text adventure game I have ever played:
Poor vocabulary, weak parser and boring storyline.

I had a look at screen grid characteristics a few years ago. My
curiosity was aroused by the so called 'enhanced triode' circuit,
wherein the screen is used as the control grid.

In some instances it is possible to get information right out of the
tube manual. That is the case for the 6L6GC in quite a few of the later
RCA manuals where plate curves with the screen as variable are included.
By inspection the G is around one ma per volt (1000 micromhos). That is,
a plate current change of 50 ma for a screen voltage change of 50 volts.
But that is at zero G1 volts, so not very useful. You could apply all
this to the 807.

Since the plate resistance is the same whether driven by G1 or G2 ( see
the curves) we can calculate the G2/Plate triode mu.
Mu is a dimensionless number so Mu is g*rp or mu = (1) ma/v * 33
v/ma.............or 33, if we take the plate to be at 350 volts & the
screen at 250 volts.

In some other unrelated work I found the 6F6 family to have a G2/Plate
triode mu in the range of 45 to 50.

Quite a few of the tubes such as the 6DQ5 used in horizontal sweep
applications have their G2 curves are plotted. G for the 6DQ5 at the
screen was something like 4 ma/volt in a useable region.

I wired up a simple test circuit in order to get some data on a few
others. The results follow-

TABLE- MEASUREMENT OF Gm WITH GRID TWO AS CONTROL, Plate at 300 volts

CONDITIONS RESULTS

TUBE G1 G2 TP1 TP2 Gm I Plate
volts volts vac vac mA/V mA

6K6GT -9 158.6 13.38 0.5 0.374 22.3
6K6GT zero 158.6 14.50 0.8 0.552 64.0

6V6GT -9 158.6 12.67 0.5 0.395 24.0
6V6GT zero 158.6 12.33 0.694 0.563 72.0

6Y6GT -18 158 12.51 1.51 1.21 63.0

I hope someone out there finds this a bit useful.

Cheers, John Stewart

wrote:

John Byrns wrote:
In article ,
admin-n-att--tubezone-dott-net wrote:

wrote:

It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.

I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value

I think you are right. Although I would be surprised if the
instantaneous AC currents ever got to twice that.

That depends on the loadline line & the program level. In an ideal world the
loadline would run nicely up into the knee of the pentode (or UL) curve. But a
complex load such a lautsprecher will go Hi-Z on it's resonant frequencies. As
the plate voltage bottoms out on those peaks, the screen current will go very
high, indeed. The resulting screen currents are sometimes shown on the published
plate curves.

That is why I pick a loadline of lower Z than what would give max power. That
lower Z load line also decreases 3rd H distortion, something the PP connexion
does nothing for at all.

Cheers, John Stewart

john stewart wrote:

wrote:

John Byrns wrote:
In article ,
admin-n-att--tubezone-dott-net wrote:

wrote:

It looks like I'll be using it as a shunt. With the 807 spec sheet
telling me typical AB1 PP operation has screen currents of 5mA to 20mA,
it looks like 30 mA or so through the shunt regulator will be about
right, and that seems to be their sweet spot. This will also allow the
VR chain to serve as a bleeder before plate current kicks in.

Sounds like a good plan.

I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value

I think you are right. Although I would be surprised if the
instantaneous AC currents ever got to twice that.

That depends on the loadline line & the program level. In an ideal world the
loadline would run nicely up into the knee of the pentode (or UL) curve. But a
complex load such a lautsprecher will go Hi-Z on it's resonant frequencies. As
the plate voltage bottoms out on those peaks, the screen current will go very
high, indeed. The resulting screen currents are sometimes shown on the published
plate curves.

That is why I pick a loadline of lower Z than what would give max power. That
lower Z load line also decreases 3rd H distortion, something the PP connexion
does nothing for at all.

Cheers, John Stewart

The anode voltage bottoming resulting in high screen currents is another reason not
to use
pure beam tetrode operation but then NFB controls the gain and in fact the problems
you mention with G2 currents
don't cause much trouble.

The gm of the screen is much less than the control grid and therefore where the tube
has g1 at Ek
it virtually dissapears and the tube becomes a triode controlled by the screen and
the result can be very
linear if the screen is biased to get about 20 watts of pda in an 807/6L6.
The Ra of such a triode is quite high like the Ra when the tube operates as a beam
tube.
This is because of distances involved between the cathode and screen and screen
anode.
As the control electrode is moved away from the cathode and closer to the anode the
gm and µ both fall
and Ra rises. The electron stream is subject to the net effect of electric voltage
fields resulting from
the distances between electrodes.

I doubt that the alignment of g1 and g2 wires would have much effect other than
keeping the dc pdG2 low.

The g2 is capable of controlling Ia just as much as g1 if the voltage swing applied
is high enough
and this is tha case with the UL connection and with about 40% UL the odd order thd
is very much reduced.
Its a main reason why UL was so eargerly used.

If there was some way of applying the distortion at the anode without the signal
voltage present
then the anode distortion could be more nicely reduced. Such error correction is
difficult.
if achieved it would allow the tube to have the same gain as in tetrode but with
much reduced thd.

Patrick Turner.

john stewart wrote:
wrote:

John Byrns wrote:
I believe that the "screen currents of 5mA to 20mA" mentioned in the spec
sheet is the DC value

I think you are right. Although I would be surprised if the
instantaneous AC currents ever got to twice that.

That depends on the loadline line & the program level. In an ideal world the
loadline would run nicely up into the knee of the pentode (or UL) curve. But a
complex load such a lautsprecher will go Hi-Z on it's resonant frequencies. As
the plate voltage bottoms out on those peaks, the screen current will go very
high, indeed. The resulting screen currents are sometimes shown on the published
plate curves.

Interesting, if I look at the curves in the STC 807 appnote (esp curve
307-341) I do indeed see that the screen current can take off VERY
dramatically as power out for a P-P pair goes above 40 watts. At that
point in fact the screen current is taking off like a rocket while the
plate current is actually leveling off! Plate current leveling off must
mean that we're going into saturation or "bottoming out", right?

But that is after THD reaches the 4 or 5 percent level, and of course
THD only goes up past that point. Isn't the lesson that "all sorts of
nasty things happen at saturation", something we already knew? I
suppose the new point is driving complex loads like a real speaker.

Tim.

While a couple of folks here obviously like their series-pass tube
regulators for screen voltages I am probably not going to go that way
but stick with something simpler like a shunt regulator made out of VR
tubes. Mostly because I don't want to build a tube regulator, I want to
build an amp.

I do have an old Heath regulated bench supply (internally some OA2's
for reference and 6L6's for pass elements) that I might give a try as a
screen supply at some phase.

The 807 will work fine in a conventional ultralinear circuit, triode
connected, or a lower screen supply. If you use a split supply with a
centertapped secondary and a bridge rectifier you get two B+ voltages,
one is twice the other. Use a separate heater and bias supply and
sequence them on the primary side.

Actually I'm not awfully concerned about soft-starting, I just thought
I'd throw that in as a weak technical justification for using a tube
rectifier. Really I just WANT to use some glow tubes and try out some
of these "new" cheap Russki rectifiers. I also WANT to do a choke input
power supply and see what the ripple currents look like on a scope. I
very likely will be back with some gripes about my poor choices, just
wait and see!


Build your power supply first and on its own chassis so it can run a
stereo amp, a linear, or a Collins or Heathkit transceiver or
transmitter. Or as a bench supply. I have seen Heath and aftermarket
ham power supplies used to run audio chassis-hi-fi or modulators for AM
plate modulated transmitters-very well. You can use dummy loads and
look at waveforms to your satisfaction and then build your choice of
amps to tinker with.


The very best tube amps have an outboard power supply-always.
Commercial implementations don't, today (but did in the past!) because
of ill-conceived arbitrary regulations and standards, you as a
homebrewer are limited only by good judgment and common sense.

Choke filtered supplies are the best, provided minimum current is
drawn (which can be guaranteed with zener strings) or the circuit is
proof against the cap-filtered no load runaway voltage. But M-field
radiation of the chokes is a problem unless very heavy shielding is
used or the P/S is off chassis from the signal (space is the best
shielding) which brings us back to "dough"-most high end saloon stuff
is low build cost.