[TowerTalk] Re: L-Networks, Stacked Beams

Jay Terleski wx0b@arraysolutions.com
Mon, 25 Dec 2000 13:59:13 -0600


Thanks for the good contribution Michael,

Ken would like us to believe he has a $10.00 solution to the tuner I
think, we have not seen what it is he is selling yet. But it seems to be
a tuner of some type.

As you point out even a ceramic RF switch can arc easily at 1 kW levels.
The last RF switch I purchased at Dayton was a $150 RF switch with 6
positions. It cost me $45 and I thought it was a bargain. It went into a
Donated SB220 that had arced when we "beefed up" the HV supply. Alot of
modification was needed to move all the tuned input coils, (work done by
a good friend who loves working on amps). But we now get about 1800
watts out on all bands with it.  Not bad for a couple of tired 500Zs.

The T network solution vs L netork is moot when the values are
adjustable, Since you can tweak the L or C.  But if you have just fixed
values, a T network can have more useful SWR bandwidth if you choose
values to compensate for the frequency.

Merry Christmas, 

Jay, WX0B




Michael Tope wrote:
> 
> Jay and Ken, please see my comments below.
> 
> 73 de Mike, W4EF...................
> 
> > K7GCO@aol.com wrote:
> > >
> > >  >    Jay: I described my "Universal L-Network Tuner" in a previous Post to
> > >  > match just about anything and it would be a good construction article for
> > > one
> > >  > of the mags or some Mfg to make.   I have 4 L-Network combo's, Series L&C
> > > and Shunt L&C "instantly selectable".  (No mfg has a tuner like this)
> 
> Ken, take a look at the Ten-Tec 4229/229 Tuner. This tuner uses a single air variable
> and 12uH roller inductor to form an L-network. A ceramic switch is used to re-configure
> the tuner into two configuration for matching Hi-Z (>50 ohms) and Low-Z (<50 Ohms).
> For each configuration there are a number a settings which switch in varying amounts
> of fixed capacitance to extend the frequency/matching range. My experience has been
> that this tuner will match just about anything. The only weak point in the design is the
> size of the ceramic switch. Ten-Tec probably should have used a bigger switch as the
> one they choose arcs under some circumstances (I actually destroyed one of the
> switches with an SB-220). Of course my unit is around 15 years old, so perhaps they
> have made some improvements since then.
> 
> >
> > Lets see if you can do the above excersize
> >
> >
> > >
> > >  This I know, a well designed T network will outperform an L network for
> > >  bandwidth even with fixed Resistive loads.
> > >  ********That I have not observed in comparisions.  You also have "3 KNOBS"
> >
> >
> > Here again we are not talking the same thing. I am talking FIXED
> > networks. Now do you agree? Let me rephrase my statement.
> > A fixed network T can be designed to have a wider frequency range of
> > function over a fixed L network.
> >
> >
> > Its now becoming very obvious to me that you are talking about a GCO
> > adjustable network.
> >
> >
> > > (2 C's-1L) to adjust with a T which is a big pain, a big time waster and a
> > > step in the WRONG  direction.  A pi and perhaps the T will attenuate
> > > harmonics better due to the higher Q which generally means LESS BANDWIDTH.  I
> > > have just "ONE KNOB TO TURN IN EITHER THE UNBLANCED OR BALANCED L NETWORK"
> > > (with 2 variable L's&C's) I HAVE.  I call that the "Ultimate in Simplicity"
> > > for Tuners.  I didn't say it was simplier than your no adjustment UN UN's as
> > > you have no adjustements.  I have the least tuner adjustements and loss of
> > > any tuner.  There is a reactance sensing circuit maintaining "1:1 SWR" I'm
> > > going to use that makes adjustments--automatically.  How do you like them
> > > design apples?
> >
> I have both T type (Dentron) and L type (Ten-Tec) tuners in my shack. In either
> case, searching for match point is a pain in the butt, so I generally put a piece
> of paper behind the knobs and mark each band I going to use with a preset.
> Both the T type and L-type Tuners have eaxactly three knobs to mark and adjust.
> The T type has two air variable caps and a ceramic switch to select the
> appropriate tap on the air core inductor. The L-type has one air variable cap, a roller
> inductor, and a ceramic switch to select the L configuration and the amount of fixed
> padding capacitance. In either case, once the presets are marked, it takes a few
> seconds to switch bands. In both cases losses seem to be minimal as I generally
> observe little if any self heating of the network components when running 1 KW.
> Based upon my limited anecdotal observations of the two tuners, my preference
> is towards the L-network. It seems to have a wider matching range (there are some
> loads I just can't seem to match on the high bands with the T network), but this
> is easily explained by the fact that the L-network has the roller inductor whereas
> the T uses a finite number of taps. The L-network seems to be a little better in
> terms of bandwidth, but admittedly, I have never set them down side by side
> and done a good scientific comparison test.
> 
> > >   >
> > >   > Your doubts of practicality and time predictions of working out the
> > > design or
> > >  > whatever you were trying to say on what I have, are unjustified as you have
> > >  > not seen what I have or how it works.
> > >
> > >  I have made lots of L and T networks. They work on one band and must be
> > >  adjusted to work over an octave yet alone 4 octaves of frequency.
> > >  ********You still haven't seen what I have.  I have said serval times I need
> > > only ONE configuration for the application discussed here with a slight touch
> > > up over the range but I can always obtain "absolute 1:1" at any
> > > frequency--you can't.  The configuration is the same REGARDLESS OF FREQUENCY.
> > >  Take a 25 ohm load and using the correct L-Network configuration and match
> > > 25 ohms over 4 octaves.  The configurations stays the same, only the
> > > reactance values of each component change.  Case closed.
> > >  >
> 
> In a contest, if I had the choice between living with a VSWR slightly greater than 1:1
> and having to touch up an knob every time I changed bands, I would go for the
> former.
> 
> > >
> > > Jay you have been sounding like W8JI who nit picks everyones sytems just to
> > > intimidate them to get them to give him more information at their expense and
> > > time because he can't understand what he's told and doesn't know how to
> > > derive it himself.
> >
> > I am aware you have run off JI from this reflector.  But I am not trying
> > to nit pick on you Ken, you started this by nit picking on me. I am
> > still waiting for an apology or at least a comment on the accusation
> > you made concerning my StackMatch product being compared to "an indian
> > guide who stole somebody's money by taking them for a ride".  I asked
> > for an explanation from you giving you the benefit of the doubt. But
> > none came.
> >
> >
> >   He bad mouth the Match Box and I defended it justifiably
> > > as had didn't know what he was taking about.  Everytime I corrected him he
> > > called it a "Persoanl Attack".
> >
> When I suggested a simple test that would settle this argument, you never
> responded.

-- 
Jay Terleski
WX0B - Array Solutions
www.arraysolutions.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/towertalk
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com