[TenTec] TenTec 228 - TenTec 229 ATU

[Dr. Gerald N. Johnson]

On Tue, 2008-06-10 at 17:40 +0800, Marinus Loewensteijn wrote:
> 
> The 228 uses a 42 position switch for switching the inductance, the 229
>  has a roller inductor. When the power amplifier stage consisted of tubes 
> the roller inductor with its "infinite" adjustment was a very fine thing to have.
> Another advantage that a roller inductor has is that the self capacitance 
> is very low which means a far higher self resonance for a given inductance.

That can be true for the parallel resonance, but the series resonance of
the unused portion is mostly due to the length of the wire. That
resonance is very close to where the wire length is 1/4 or 3/4
wavelength (if open or 1/2 and full wave if shorted) and how its wound
has little effect. 

I'm not sure that the distributed C of a coil on a ceramic form (fairly
high dielectric constant) is lower than that of an air wound coil
supported on three or four polystyrene bars (ala B&W coil stock).


While the used roller coil will have its bad spots when adjusting, the
switched coil may still go open while switching. Ah, you NEVER ever
switch the coil with the key down?
> 
> I wonder if a roller inductor is still an optimum solution with 
> semiconductor output stages. I view a switch as self cleaning and 
> a roller inductor imho as requiring (little) maintenance to prevent 
> arcing. Arcing (in my book) is a sign of not making proper contact 
> which means that the output stage is not properly loaded and this 
> is potentially damaging the semiconductors in the output stage.

Which points to our common technique of tuning with RF power may not be
the best technique. It produces QRM, and as you have noted can be
hazardous to the RF power source if it isn't well protected. Now I've
owned two Kenwood TS-120 and a TS-130 and tuned my manual tuner maybe
thousands of times with a brick on the key and with full drive and never
hurt their PA. Many radios look at SWR and cut the drive to the PA.
Tentecs don't do that. They built a sturdy PA and protect it when the
current gets high, by the fast breaker or power supply over current shut
down. Which makes it a bit less tolerant of the mismatches that always
happen when adjusting the tuner.

QRM is much less and the PA is very well protected when we use a noise
bridge, tuner tuner, or antenna analyzer to adjust the tuner for a 50
ohm input resistance (no reactance). And that device will also give a
strong hint of any roller coil or switch connection problems. Tuning at
low power, maybe milliwatts or less, doesn't apply enough voltage to the
silver oxide to break it down to make a good connection. That if not
disturbed will pass microvolts, but if not broken down it holds off a
significant fraction of a volt.
> 
> Would the 228 with its T circuit and switched inductor be a better 
> ( mechanically, contact wise )  solution than the 229 with a roller 
> inductor? Or am I missing something here?

The roller coil is more versatile. The T circuit has many combinations
of inductance and capacitances that will perform the impedance match.
There is an optimum one that has the lowest loaded Q that gives the best
tuner efficiency, and many that because the higher loaded Q causes
greater circulating currents that heat the tuner more. We rarely search
for the lowest Q operating point.

L match tuners guarantee that minimum Q because with only two elements
there is only ONE combination that will match a specific load to 50
ohms. There the closer to 50 ohms that the load is, the more efficient
the tuner.

The bigger variation in tuner loss comes from the wire size in the coil,
the fatter wire always offering lower loss, but greater tuner bulk.
> 
> PS This is not to replace my S-Match for balanced output ( twincom 
> / open wire ),  purely interested in driving a (unbalanced) coax 
> system ( some phased verticals ).
> 
> Thanks for your input, 73, Marinus, ZL2ML
> 
73, Jerry, K0CQ