[Amps] Amp question

Tue Jul 26 10:46:23 EDT 2005

A "conventional combiner" is more often than not a "zero-degree combining network".

For a through discussion of power dividers and combiners take a look at:
http://www.minicircuits.com/appnote/an10006.pdf, and also pp 498 - 510
in "Single Sideband Systems & Circuits" 2nd edition.

It is the designer's choice if the combining system should be zero-degree or 180-degree,
but as it simplifies construction and testing somewhat the zero-degree 
approach appears to be usually taken.

A special case is when you use 90-degree hybrids to absorb load reflections, but
they seem to be not so very common for multi-octave combiner networks, as those encountered in HF
amplifiers.

73/

Karl-Arne
SM0AOM

----- Original Message ----- 
From: "Will Matney" <craxd at engineer.com>
To: <amps at contesting.com>
Sent: Tuesday, July 26, 2005 12:23 PM
Subject: Re: [Amps] Amp question

> Karl,
> 
> Please see below;
> 
> > 
> > The reason why this works is that the output combining network in 
> > most solid-state PA's are
> > "zero-degree hybrids", which means that they expect the voltages 
> > (or currents) driving them to be
> > in-phase to provide a combined output. Output transformers in p-p 
> > stages can on the other hand
> > be classified as  "180-degree hybrids", which require 
> > opposite-phase driving voltages to operate properly.
> 
> 
> By looking over the schematic of this amp, everything looks the same as one for 180 degrees phase difference except it being minus a splitter in the front end. To me if it were to be zero degrees you wouldn't need a splitter or combiner, just merely paralell the two modules the same as paralelling tubes to add up the power output. They have two conventional RF output and input transformers per module. These work in push-pull to provide a linear signal being it's class AB1. Then each output from the RF transformer goes to either end of the combiner transformer just like all the others being 180 degrees out of phase. However, on the schematic, they dont show any phasing dots on the combiner. Still though, I would think it would still have to be out of phase if it's wound the same as a conventional combiner. The only way it could be different is the two coils be in phase instead out of phase? It has to be one of the two by the schematic they show. I still just can't see how it's
>   doing it by having them both on at the same time and not buck each other at the end. I'll have to do some research on this as I've not seen that arrangement in the Motorola app notes, nor in any book I have on the subject.
> 
> > 
> > As the phase shifts in the two amplifier halves can be assumed to be equal,
> > it remains to drive the halves with in-phase voltages.
> > 
> > The RC-network on the input side just acts as a resistive power 
> > dividing network with its outputs in-phase,
> > and as the amplifier power gains decrease with frequency, capactive 
> > shunt elements
> > can compensate for this decrease, as the insertion loss of the 
> > network also will decrease with frequency.
> 
> 
> Agreed.
> 
> > 
> > Finally, the reason for using 200 ohms for the balancing resistor 
> > is simply that the designer has chosen a 100 ohm
> > input design impedance for the combining network, which requires 
> > the amplifier outputs to be 100 instead of 50 ohms.
> > This is "perfectly legal" in an environment where you are at 
> > liberty to choose the design impedances freely, and do not have to 
> > take long runs of interconnecting RF cables into account. But the 
> > stray and winding capacitances present can be more difficult to 
> > handle in a system with a higher design impedance.
> 
> 
> Exactly. I figured it was because of it going into the low pass filters the way they had it designed, and allowing them to finish stepping it down to 50 ohms. But, like you said, using 100 ohms increases the windings on every transformer by X2.
> 
> > 
> > In professional circles this is not very common, as most 
> > solid-state professional PA's are built from modules
> > having a design impedance of 50 ohms.
> 
> Agreed. That just didn't make a lot of sense to me.
> 
> Best,
> 
> Will
> 
> > 
> > 
> > 73/
> > 
> > Karl-Arne
> > SM0AOM
> > 
> > ----- Original Message -----
> > From: "Will Matney" <craxd at engineer.com>
> > To: <amps at contesting.com>
> > Sent: Tuesday, July 26, 2005 6:57 AM
> > Subject: [Amps] Amp question
> > 
> > 
> > > All,
> > >
> > > Ok, most know how splitter/combiner transformers work in solid 
> > > state amps. Here's the question, how can an amp work using two 
> > > sets of push-pull amps, having a conventional combiner, but not 
> > > have a splitter in the front end? What this amp has is a divider 
> > > network made up of a RC network using several resistors in 
> > > paralell with a capacitor on either side of the RF input. In 
> > > other words, when the RF comes into this RC divider, it applies 
> > > RF equally to both sets of push-pull amps at one time. There is 
> > > no splitter transformer there. How can this work by having both 
> > > push-pull sets on where they should be 180 degrees out of phase, 
> > > and one being off while the other is on? To use a combiner 
> > > transformer, the RF from one set makes 1/2 wave and so from the 
> > > other set. The combiner then adds these two halves together 
> > > making a full sine wave. So, how can this work having RF going to 
> > > both sets at once instead of using a splitter which would put 
> > > them 180 Deg out of phase? This absolutely has  me stumped. By 
> > > the way, it's for an Ameritron ALS-500M.
> > >
> > > One last question, most combiners/splitters use a 100 ohm 
> > > resistor in paralell with it's inputs for a 50 ohm load. How can 
> > > this work if the resistance is 200 ohms? That would be for a 100 
> > > ohm load wouldn't it?
> > >
> > > Best,
> > >
> > > Will
> > >
> > > -- ___________________________________________________________
> > > Sign-up for Ads Free at Mail.com
> > > http://promo.mail.com/adsfreejump.htm
> > >
> 
> 
> 
> 
> -- 
> ___________________________________________________________
> Sign-up for Ads Free at Mail.com
> http://promo.mail.com/adsfreejump.htm
> 
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
> 
> 
> -- 
> No virus found in this incoming message.
> Checked by AVG Anti-Virus.
> Version: 7.0.338 / Virus Database: 267.9.5/58 - Release Date: 2005-07-25
> 

-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.5/58 - Release Date: 2005-07-25