[TenTec] TenTec 228 ATU

[Dr. Gerald N. Johnson]

On Fri, 2008-06-27 at 15:19 +0800, Marinus Loewensteijn wrote:
> Earlier I invited on the list input in a proposed change to the TT
> 228 ATU.
> 
> I was considering placing an impedance transformer at the output 
> of the transmatch in an effort to lower the losses when matching 
> into a low Z.
> 
> Reading in the ARRL Antenna Compendium Vol 3 about the losses 
> and problem in a T-match when matching a 50 : 50 Ohm coax 
> connection has me made rethink placing this "un-un" transformer at
> the output. The main problems are associated with reactances that
> the antenna can present.
> 
> The objective is to prevent the melting of the plastic on the 
> toroid / switch. Since I intend to work purely into "matched"
> antenna's extreme matching capabilities are not paramount. 

Do you know that switch was melted feeding a reasonable load or was it
used to feed a short wire with a very low radiation resistance with more
than the tuner's rated power?
> 
> I am targeting to match to a 1:1 SWR to prevent damage to the
> transmitter. Secondly I like to obtain a larger bandwidth on 40, 30
> and 20 meters with a minor interest in 80 meters.

You can increase the bandwidth by selecting the tuning components with
the lowest loaded Q, probably going for the maximum inductance for the
band. Otherwise converting to an L match guarantees the lowest loaded Q
for the impedance transformation because its the only match possible.
> 
> Extensive simulation is leading me towards using a 1:9 transformer,
> stepping up the 50 Ohm transmitter impedance to 450 Ohm and
> then use a low pass L-network to match to 12.5 - 200 Ohm.

I think you will find the 1:9 transformer bandwidth way less than any
comparable sized 1:4 transformer and that matching the 12.5 to 200 ohm
load range to that 450 ohm impedance will result in further narrowing of
the band. The loaded Q of the L network rises when the I/O impedance
ratio rises. A higher loaded Q means narrower bandwidth and greater
losses.

I've tried to build a 1:9 transformer with large toroids and found it
was barely useful on one band where I was wishing for 450 ohms on the
high side. 1:9 is quite practical for 5.5 to 50 ohms, but very difficult
for 50 to 450 ohms because the high impedance takes so many turns to
achieve an inductive reactance a few times larger than 450 ohms that the
wire approaches a quarter wave long which sets the high frequency side
of the transformer pass band. Again something achievable in a quarter
inch toroid, but hard in a 2" toroid.
> 
> This setup has values for the capacitor and inductor that are of
> practical values. In a pinch I can add on the back an extra connector
> in which I can plug in extra capacitance (vitramon capacitors) for
> matching on the "top" band. If inductance proves to be an issue then 
> this can added in line with the antenna.

The trade off in raising the operating Z of the tuner is that at the LF
end, it will require more inductance which is fundamentally lossier than
more capacitance at high frequencies.
> 
> I appreciate feedback on this changed approach, many thanks in advance,
> 73, Marinus, ZL2ML
> 
When your antennas are resonant, your ordinary tuner will do quite well,
and on bands with the bandwidths of 40, 20, and 15 meters, the PA will
perform the same without across them without a tuner. You can improve
the antenna bandwidths by increasing the effective diameter at the end
insulators or by applying a parallel LC tuned circuit at the antenna
feed point.

73, Jerry, K0CQ