[BC] "Cusp Rotation"

[Phil Alexander]

On 16 Mar 2007 at 14:52, Thomas G. Osenkowsky wrote:

> The optimal sideband cusp rotation will be different
> for every transmitter design. 

Seems to me that should be all the more reason for
some manufacturer coming up with a way of measuring
the load(s) on the ACTUAL PA(s).

> For a DA system, we have the issue of pattern
> bandwidth.

True, but that is a somewhat different issue, although
there can be some interaction IOW Tx not "liking" the
load because pat b/w is poor even though rotation is
correct.

> For the simple non-D case, usually a Tee network
> is employed in the ATU. The rotation can be measured
> at the ATU input terminals. The phase shift of the
> transmission line (considering Velocity Factor) can be
> calculated or the rotation measured at the tx antenna
> terminal.

At best, all you will get is the load phased where the
manufacturer approximates it should be to present a 
"horns up" load at the actual PA output(s). This is the
reason we need a method of reading the match from the
perspective of the PA(s).

> The transmitter output network is more complex than
> a simple Tee network. It has the task of matching the
> final amplifier(s) to the antenna and attenuating spurious
> and harmonic energy. 

In some cases that is a simple Tee with shunt branch 
series resonant at the worst harmonic. Consider the
Harris "Gates" models or SX models which set this to
the third harmonic. Not very complex but hard to tune
for most people.

> Most new transmitters combine
> multiple PA stages to produce full power. The optimal
> cusp rotation must appear at each PA device. The phase
> shift between this point and the antenna terminal is the
> unknown that is important. 

But of course, and that has been my point from the 
beginning. We need direct measurement of this for what
should be very obvious reasons.

> The manufacturer can provide this data. 

No, they can only approximate it unless they measure
it in test. Some will do that upon request, provided
the buyer knows they can ask and the factory does it
while the Tx is in test. This is a definite minority.

> Not all transmitters are the same! In the older
> plate modulated rigs, it was common to use a -225 deg
> network. This provided isolation between Plate Tuning
> and Plate Loading. 

And it was an easy, cheap network to build. 3 or 4k
to 50 works nicely in a Pi net.

> Some newer transmitters employ
> 90 degree, 135 degree or even 180 degree output
> networks. 

And others are even more strange. <g>

> Placing a symmetrical load to  the antenna
> terminal can be the worst possible load to present to the
> PA for a -225 degree network, but perfectlu acceptable
> for a 180 degree network.

This is exactly my point, thus there needs to be a way to
measure at the actual PA(s). This is especially true where
a slight change in phasor adjustment may have very 
significant effect on rotation.
> 
> Tom Osenkowsky, CPBE
> 
> 

> From: "Kent Winrich" <kwinrich at gmail.com>
> 
> > You are expecting the manufacturers to provide this info??? 
> > HAHAHAHA  They dont know half of what they put out.  Heck I couldnt 
> > even get a simple thing like BTU output from Continental for their 
> > 816HD (you know so I could spec out an HVAC system??) let alone a 
> > bunch of measurement points on their own transmitter.

This is OT for this thread, but since I started this, 
I'll throw it in anyway. <g> There is an easy way to 
calculate BTU output from any box. Subtract the 
comesoutta from the goesinta and the balance is heat 
in kW. Usually, they will give you the box power input 
for a given output, especially if they are touting the 
overall eff. Convert the kW to BTU and you have heat 
load for HVAC.

> > I like the idea of your Vector analyzer.  Sure would be easier than 
> > going through the OIB and what have you that I did for my two HD 
> > installs.  UGH.

I would think it would be possible to put a pickup 
upstream of the combiner/match in an SS box that 
could feed a RF network analyzer. It would need 
effectively a directional coupler working at the 
very low Z of the PA(s), but otherwise it would 
not be so different from the measuring rig that 
Ron Rackely uses for sweeping DA's except that 
the Tx would act as the generator and since 
modulation would be randomly hitting frequencies 
you would need a dynamic method of selecting and 
storing frequencies for presentation as a chart 
on the 'scope. At the very least, the manufacturers 
should be able to provide the pickup which would
probably be not much more than pickup loops in 
proximity to the output of the PA(s).


---------------------------------------------
Phil Alexander, CSRE, AMD
Broadcast Engineering Services and Technology 
(a Div. of Advanced Parts Corporation) 
Ph. (317) 335-2065   FAX (317) 335-9037