[TowerTalk] KT34A M2 Upgrade 15 meters Issue

[bills stuff]

I upgraded my old triband (~1984) KLM KT34A that worked great, with the 
available M2 kit, to the (upgrade version of the) KT34M2.  The 15 m 
performance (pretty much a dipole pattern with F/B a few dB at best 
using my modest test range) and SWR (> 2 on low end of band) were not up 
to expectations.  Tests of DC continuity of all the joints were done, 
all seemed fine. (Poor contacts in these joints are a well known cause 
of KT34 problems, especially for the older hardware – see reflector 
archives.) Email exchanges with M2 first recommended replacement of all 
8 of the 15m capacitor tubes (16”) - the thinking of Mike Staal at M2 
(designer of the antenna) being that the Q of the traps was degraded. 
This purchase and replacement of new tubes (from M2) made no difference 
in performance (in the replacement process all joints were checked for 
tightness and all dimensions verified yet again).

I asked M2 if there were any simple (no high end test eqpt) diagnostics 
that might be helpful and was told that measurement of the element 
resonant behavior could be definitive.  The high Q of the 15 m traps 
produces relatively sharp resonances. M2 provided a copy of a 1980 test 
notebook page giving resonant frequencies for the individual elements 
(for the KT34XA, which is, I believe, the same as the KT34A except for 
an additional director element). Here is the M2 (really KLM) table of 
resonant frequencies (MHz):

Band                    20                       15 
10
Reflector           13.8                   20.9 
27.8
Rear Driven       14.03                 20.925 
28.75
Front Driven      15.075               21.86 
29.4
1st Director          14.4                  21.9 
29.35
2nd Director          14.6                  21.65 
29.15-29.20

These values were measured with the individual elements removed from the 
antenna and fed with a 1:1 balun, in the clear (at 28 ft).

I carried these painful measurements with my antenna elements and at 15 
m (did the other bands too) and found these resonant frequencies (min 
VSWR ~ 1.3 with RG-8, 50 ohm feedline):

Reflector           21.05
Rear Driven       20.9
Front Driven      22.1
Director              22.0

The most notable result is for the reflector which is high and in band. 
Mike indicates that the high reflector resonance value would be expected 
to kill the F/B.

So now what?  The resonances are all fairly narrow indicating that the 
high Q traps are working so it looks as if the sole issue is resonance 
location.  However, tinkering with this antenna is not for the 
faint-hearted since it has a novel, even unique, design with low loss 
traps, dual driven elements and linear loading.

My pleas for further recommendations from M2 for the final fix are no 
longer answered.  Getting the info I did required playing the role of 
the informed pest (and buying some parts) but the undertone of some 
responses indicated that a lot of the performance complaints to M2 are 
rooted in customer’s inability to follow directions (which is likely 
true).  Not too surprising that they are not too interested in fudging 
the dimensions to make up for what appears to be some assembly problems. 
However, IMHO the assembled dimensions are not the problem.  The problem 
is not known.

For entertainment,  and with low expectations, I put together a model of 
the KT34A using MMANA, recognizing that, for free, you cannot get high 
quality results.  I can supply the input file if anyone is interested. 
M2 indicates that more sophisticated models are need to handle the novel 
loading.  Using capacitive loads in the model for the two traps, I 
experimented with values of capacitance that would produce resonances 
that were either my observed ones or M2 measured ones for the individual 
elements.  Somewhat surprisingly, the required Cs are not terribly 
different from what you get by calculation for a co-cylinder cap of the 
actual dimensions.  Note that the inductances for the traps are 
determined by the dimensions of the loops in the elements and these are 
probably modeled okay.  The model has all the segments of the actual 
antenna, aside from capacitors.

The model of the assembled elements produces patterns on 10 and 20 
pretty much like you might expect for a yagi – in spite of the fact that 
on 20 m where there was no resonance data fitting done.  On 15 m the 
pattern is very sensitive to frequency and gives a F/B that actually 
goes negative (in dB) at frequencies below the resonance of the 
reflector (21.05 MHz) agreeing with expectations.  If I use the M2 
resonances, the F/B is still not impressive, less than 5 dB.  In both 
the 15 m cases, the current flowing on the DIRECTOR is very low in the 
model, in contrast with the other bands.  All this may not be very 
meaningful.

So finally I get to the point - are there any folks out there who have 
experienced similar problems with the KT34 on 15 meters and was there 
any resolution?

  (BTW, it is reported that KLM during some of its history sent out 
capacitor tubes with a wall thickness, and inner diameter, that was not 
quite right – thus giving the wrong capacitance and bad 15 m 
performance.  This is not the case for my problem.)

Has anyone modeled this antenna?

I believe it is possible to tune the resonances for the elements by 
adjustment of the shorting bars that determine the length of the 15 m 
and 10 m inductance loops but I was hoping for some experience for 
guidance before venturing into unknown territory.

Bill, N6MW