6L6 to 2x 811A ratio

[W4RFM]

Anyone know the ratio I need for a transformer to link a 6L6 to drive a pair of 811a's in a modulator, like the Globe King did?
Thanks
Bob

[w4bfs]

the old Orr handbooks or maybe RCA xmit tube hb will walk you thru this .... you will need neg feedback to reduce 6L6 driver Z for low distortion ... be wary of phase shifts which can turn your - fb to + fb and attendant oscillations ... it would be easier to use a low Z triode at this power level ...73 ...John   

[W3RSW]

You'll need several watts to drive the 811's.  Referencing the 1957 ARRL hbk, p. 271, If you must use 6L6's, use a pair in push pull, class AB1 with feedback.  At 360 volts on the plates you'll need 9000 ohms plate to plate on the 6L6 outputs.  The 811A side should be abour 1k to 1.5k grid to grid, center tapped, of course.  This will give you about 20 watts audio output.

The transformer driving the 6L6's should have a two separate windings, with a 0.1uf in series with a 100k resistor from each 6L6 plate to the 'ground' side of their respective driving split windings.  The 'ground' side of the driving split windings then need 22k to ground, unbypassed.   I assume these values are for reasonable, communication AM fidelity before the chopped off highs, SSB simulated era.  Don't play too much with the feedback in attempt to get hi-fi or you will run into the phase oscillation previously mentioned.

The cathodes of the 6L6's are connected together and self biased with a common 250 ohm /10w resistor bypassed by 100uf/35v electrolytic.

I guess you could try one 6L6 at 300 volts plate, feedback and a cathode circuit to yield class A operation.  Try 230 ohms in cathode since you'll need larger quiescent operating current.  Output transformer will now be 4.5k plate to B+.
(from tube tables, same handbook, p. v18/19.)   This will get you about 6.5 watts (no feedback) and maybe 5 watts with feedback.  Marginal for a pair of 811's at full strap.

[WBear2GCR]

a pair of 6L6 are more than sufficient, imho.
6W4 do quite nicely... as a cathode follower driver... low Z assured.
No negative FB required.

Get the voltage gain in the previous stage with that scenario...

Also cheaper since you don't need a good driver transformer at all.

And you get excellent bandwidth that way too. 

              _-_-bear

[W4RFM]

All this info is much appreciated.  I found my answer by taking some of the info above and figuring one 6L6 load is 4.5K, and according to the ARRL handbook referenced above the two 811A grids need 2K, so the ratio for my Stancor will be 2.25 : 1.

 I actually thought about copying an Altec 1570 amplifier all the way through, but I do not know the value of that choke feeding the 6W6 cathodes. I am only looking for about 150-175 watts output to modulate a single 4-125A kinda like a Collins 30K running at 250w.
The Altec is at: http://www.triodeel.com/al1570b.gif
Thanks everyone Bob

[KE6DF]

FWIW here is a single 6L6 cathode follower driver for 811A's -- from the Orr 1959 handbook.

[W7XXX]

What xfmr depends on how you connect them. My favorite is the Bill Orr method posted above. I have also used the single 807 cathode fed 813's per Orr. If finding a class B driver transformer is a problem, then use a small multi match modulation xfmr that has a 2:1 ratio. The 1:3 interstage xfmrs are not to hard to find.

[W3RSW]

Good ol' Bill Orr.  What a great circuit. 

The 1:3 interstage as found in the Johnson Ranger is notorious for opening up, marginal for both 12AU7 plates so make sure any 1:3 you find is larger than the 1 1/2 x 1 1/4 x 5/8 inch beasts (rough core dimensions) in the Johnsons and similar of that ilk.

[W4RFM]

I saved the Orr design to my doc's, and that looks like the way to go.  I have a Stancor multi ratio line driver xfmr I can use for T 2 and then I plan to use a Collins ART-13 for the mod xfmr.  Now this brings on more discussion.  What about using a mod reactor to soften the blow to the mod xfmr?  I have read a lot about it, and it seems like a good thing, I know broadcast rigs I worked on during the tube era always had one.  Your opinions??

[WBear2GCR]

FWIW here is a single 6L6 cathode follower driver for 811A's -- from the Orr 1959 handbook.

Sure you know the value of that choke!

Ask urself, how does a choke work at audio freqs?

The first spec is the power - that means the choke needs to be able to handle whatever power level you run at, which is mostly core size. We can come back to that one at the end...

The next thing we know is that a choke has DCR, that is the value that will set the bias, the same way that a cathode resistor would - or if fixed bias is applied then it is the DC cathode degeneration value - you can used combined fixed and cathode bias if needed. You want to run the follower in class A.

After DC, higher in freq, the choke looks like and ever increasing impedance. So, ignoring the DCR (the DCR acts as sort of a minimum level set, like a pot in this example) the choke's impedance at LF creates a "corner" frequency, above which the follower is essentially flat, and below which it drops down to the value set by the DCR... whereas a resistor has the same impedance at all freqs.

So, all we need is a choke with sufficient DCR to work in the cathode.
Good news, it's not critical.
The easy way to do the job is to use the primary side of an output or other suitably rated (power) transformer! That permits sufficient DCR and enough Henrys of inductance. Probably the transformer you have now will do the job just fine.

You can use 6L6 in the same circuit, no changes.

The 1570 drive cirucut is very good, and works better with higher H and higher power handling cores... the core of the stock 1570B choke was a bit undersized, probably in the 4-5watt class. I have replaced them as a matter of course when refurbing 1570s... I use a 10-15watt output iron, whatever I have on hand.

The best part is that a good output tranny is NOT needed at all! In fact a "bad" output tranny in terms of freq response is jes' fine because you want a choke with increasing Z!!

 

          _-_-bear

[W4RFM]

Great comments, thanks BEAR, sometimes we can't see the forest for the trees.

[Ed-VA3ES]

I have some Hammond  "Ham Tips" sheets from the '50's, that describe exactly this situation - driving a pair of  811's with a pair of 6L6's including driver and mod  transformer specs, NFB spec and info, etc.   I have to dig these sheets out from my archives and will post them soon (once I've scanned them).

[Carl WA1KPD]

FWIW here is a single 6L6 cathode follower driver for 811A's -- from the Orr 1959 handbook.

I have been looking at that circuit.
Will it work OK without L1 and L2??
http://amfone.net/Amforum/index.php?action=dlattach;topic=17963.0;attach=9291
Thanks
Carl
/KPD

[WBear2GCR]

yep.

skip everything between the coupling cap coming out of the first tube, and the input to the second tube - that clipper/bandpass circuit.

the "splatter filter" on the output side is NG, you want one of those 3 diode negative limiters as written up elsewhere here on amfone.net

the paralleled 12AU7 could be replaced by a single tube of more snot, assuming there's enough gain, a 12B4 might be nice, or nothing wrong with a 6SN7 or half of one (what is that, a 6J7??)... point is that the tube types are not critical, you can use what you have on hand pretty much...

for example you might make the second section into a LTP and run the transformer in P-P and get the DC off the thing... (better freq response, less critical transformer wise). There's proably a bit of extra gain in the front end because of the bandpass and clipper circuit that is not needed or wanted....

Or you could "parafeed" that coupling iron and get the DC off it that way too... that's a good idea. That's a modified Heising type connection: choke on the plate, and cap going out to the tranny primary... you set the cap value for best response given the Q and H of the primary, but it's not critical and a scope will show you what you've got pretty quickly...

                            _-_-bear

[Opcom]

In that RHB schematic, I like the use of the 2D21 a cathode bias regulator for the 811's. Ahh the days before Zener diodes..
One might wish to bypass the center tap of the 811 filament transformer with a 330uF or larger (1000uF) cap to reduce the modulation of the bias voltage by the 2nd harmonic of the audio signal current coming from the 811 cathodes. It would be the same situation with a Zener.
Orr would not have cared about the small amount of distortion, back when that was designed complete with splatter filter, but today there is no reason to allow any source of harmonic noise in the modulator that could easily be gotten rid of for a couple bucks.

Those are the only considerations for the Radio Hanbook circuit I can think of.

I tend to like the idea of a pair of PP 6L6's with overall NFB to drive the 811's as mentioned, because any speech processing could be done before the 6L6 stage, yet the 6L6 stage and 811 stage would be inside their own FB loop to reduce distorton. The RCA BTA-250 transmitter does a similar scheme.
http://www.bunkerofdoom.com/lit/bta250L/scan0010.jpg

[W4RFM]

I have already built a copy of the Altec 436C compressor, that I plan to use in front of my audio section, and just jump in at the after the mike pre-amp volume control point. An old Behringer MXL 1400 would be great to use for this, but I love building with tubes.