Push pull vs parallel finals

[kb3ouk]

I'm trying to decide which one I might go with for my 814 deck, push pull with a link coupled output, or parallel with a pi net. Is one better than the other in some way (simplicity, efficiency, etc.)? It's going to be an 80 meter only rig, I'm going to drive it with a Millen 90800 exciter. Antenna is a coax fed dipole, no antenna tuner. Either one should work with that setup.

[N2DTS]

Push pull was mostly used with triodes.
It would depend on what parts you have or can get.
Push pull coils and dual caps are hard to find in good shape.

No reason to do it one way or the other, same power output, and push pull needs close attention to the
balance and parts layout.

[W7TFO]

Harmonics are higher in parallel.

73DG

[steve_qix]

Parallel is _WAY_ easier than push-pull !! Having built both over many years, hands down - go parallel.  Everything is easier; the capacitor is simpler (split stator not needed); it's smaller; if you bandswitch - that is MUCH easier.  Just DO it   

[N2DTS]

I built up a nice push pull 812A rig modulated by 811a's with separate supplies for the rf and mod decks, one of the best rigs I ever built, tuned up very nice and neutralized better then anything else I ever had.
KW coils, link tuning, balanced output if you wanted, who needs a tuna?

I do not see any reason to do it that way with tetrodes though...

[WD8BIL]

The 2nd harmonic attenuation advantage of push-pull is totally out-weighed by the ease of multiband construction in parallel. As has been pointed out, go parallel. Harmonics can be controlled easily with proper tuning and other methods.
 

[K1JJ]

A push-pull linear tube amplifier can make for a cleaner (IMD)  and more efficient output, like when p-p for a modulator or audio amplifier or even for RF compared to a single-ended stage. (Pure class A single-ended is an exception for IMD)   But for AM class C, p-p is really not worth the extra effort.

I look at an RF p-p transmitter as a novelty - something like driving a model T. It can be a hassle to build but can be a lot of fun and gets lots of attention, but that's about it.

An RF single ended / parallel tube(s) stage with a pi-network is the popular time-proven method that is used everywhere these days, for good reason.

T

[W1ITT]

To echo W7TFO's comment, the Pi-Net will give better harmonic suppression and, for the cost of just one more inductor, a Pi-L output will be even better.  It was mentioned that this will be a single band 80m rig, so the Pi-L can be implemented without the complication of a second wafer on a bandswitch.   Just make sure the spacing on the output capacitor in the Pi-L is sufficient.  One of the Cardwell  1000 pf variables is a better choice than the oft used broadcast radio variables.

[WA4WAX]

Parallel is the way to go, unless you are on a vintage/nostalgia kick.

Some comments:  Do not use a single ended Pi network for tubes in push-pull.  Please know that  a symmetric Pi would work!

I had a long chat about this subject with W2DU.  The odd harmonic cancellation bears fruit when you use push-pull in an audio circuit.  The utility in an RF circuit is marginal; many other ways to keep the odds down.

The push-pull link circuit was popular before the Pi, or "Collins Coupler" came into vogue.  This happened a few years after World War Twice.  While losing push-pull was not a big deal, we made a mistake when we got away from balanced output, at least on HF.  With a proper tuner, this problem is easily circumvented.

Push-pull did have one big selling point in its day.  Most transmitters in the 1930's were rock bound.  Money was scarce.  Guess what?  A crystal controlled 40 meter P-P rig can run on 20 meters with the same crystal and the same output circuit.  Just tune for the 2nd, and voila!  20 meter Dx.  Who  could resist?  Two bands for the price of one.

Have fun!

Matt

[W2NBC]

Interesting thoughts about push pull/parallel

http://amfone.net/Amforum/index.php/topic,12173.0

This site is a treasure trove..   

[WA4WAX]

Senior moment!  I meant to say EVEN cancellation!

Specifics on the old sets: A pair of grid driven, push-pull 809's (Or 812, 100TH, etc.) would have the input tank (grid) tuned to 40 meters, but the plate tank could be tuned to 20 meters.  The rig would run as a push-pull class C doubler.   Common practice in the day.

That brings up another subject: Neutralizing a grid driven single ended triode can be difficult once you get up to around 14 or 15 Mc.  On the other hand, a push-pull pair is fairly easy to neutralize.

You may get a little more power out running a pair push-pull instead of parallel, but not much.

For tune up, you dip with the split stator and load with the swinging link.  If you cannot dip and load, you antenna impedance is "out of bounds".  Do not force things.  :-)

Push-pull: Nifty factor.  Direct compatibility with balanced feeders.  Easier neutralization with grid fed triodes.  Second suppression? Yes, but you better have the match and symmetry spot on!

[w1vtp]

Senior moment!  I meant to say EVEN cancellation!

Specifics on the old sets: A pair of grid driven, push-pull 809's (Or 812, 100TH, etc.) would have the input tank (grid) tuned to 40 meters, but the plate tank could be tuned to 20 meters.  The rig would run as a push-pull class C doubler.   Common practice in the day.

<snip>

Didn't they call that "push-push?"

[Opcom]

Harmonics are higher in parallel.

73DG

This is a universal truth as anyone who has ever been seated in a restaurant next to a family with two or perhaps three small children will attest. 8-)