Dave Platt wrote:
Does any one know what is inside these tuners as I would like to build
one for my Jeep here in Australia

SGC is probably the best-known brand of these tuners (although I
understand that they did not create the basic design). If you go to
http://www.sgcworld.com/PubInfoPage.html you can download the SGC-237
manual, which has a complete schematic.

There's also the original LDG design, which was featured in a QST article.


The impedance-matching network in these tuners is a pi network, with
two switchable sets of capacitors making up the input and output shunt
legs, and a set of switchable inductors making the series (top) leg of
the pi.

or as an L network with series L and shunt C, which is switched to
either input or output end. (that's what the LDG tuners do)


This sort of tuner seems to have several quirks or limitations.
Programming the search algorithm to find the right match (or *a*
correct match - often, many are possible) apparently isn't easy. The
SGC tuner I have seems to be easily "confused" if the radio varies its
power during the tuning process (e.g. as part of a high-SWR power
foldback).

This is the heart of the tuner and whether it's "good".. what's the
tuning algorithm.. The mechanics of switchable Ls and Cs (or motor
driven Ls and Cs for that matter.. another QST or QEX article had one of
these) is fairly straightforward, it's the fast tuning.





Frankly, I wouldn't recommend a project like this as a one-off
hobbyist project, even for a real enthusiast. I can see it becoming
an incredible time-and-effort-sink.

Especially since you can buy the already built unit for less than it
would cost to buy the parts to build it...



A somewhat similar tuner, which would be much easier to build, would
be one which implements a "pi" or "T" tuning network using
motor-driven air-variable capacitors and a relay-switched multi-tap
inductor (or perhaps a motor-driven roller inductor). You wouldn't
get the very rapid bandswitching of an automated ATU, but would
probably have lower losses, much less physical complexity, and a
better chance of getting (and keeping) it working.

The losses might be about the same. (For example, the SGC tuners use
air core inductors.. same as you'd have with your multitap inductor..
and getting suitably high Q ceramic caps is pretty easy)


A "tuner plus whip" setup for a vehicle is often very electrically
inefficient, at least on the lower-frequency HF bands - you can end up
with most of your transmitter power being lost as heat in the tuner,
due to the very high circulating currents flowing between the pi-match
inductance and the very-capacitive impedance of the whip.

Well... as long as the wire from tuner to whip isn't something like coax
with the shield grounded, the efficiency isn't all that bad. Of course,
it's possible to configure the tuner for a "bad" configuration (as you
mentioned, there's multiple "solutions" that generate a match, some are
better than others, and that's where the cleverness in the tuning
algorithm comes in)



My
understanding is that a whip/tuner setup is often a couple of
S-units poorer than a decent screwdriver antenna, and even worse than
that compared to a well-designed-and-built single-band "bugcatcher".

I don't think so...

2 S units is about 12 dB..

If that were the case, the tuner would be dissipating more than 90-95%
of the Tx power, and it would melt.

There's a fair number of measurements out there on tuner losses, etc.,
and none of them are credibly at the 90% level... maybe a couple dB,
except in a pathological case.

Your big hit is just from having a short antenna and a crummy ground (in
the vehicle case).. Easily a 6-12dB hit over a dipole over the same
ground, but there's nothing you can do about it, because, especially on
lower bands, you can't drive around with a 60 foot dipole 30 feet up.

The single band bugcatcher has higher efficiency AT A FEW FREQUENCIES
and is appalling once you get very far away, which means it's not really
an apples to apples comparison with a frequency agile system. If you
operate on a few select frequencies (e.g. you have a 40m or 75m net that
you participate in on a regular basis), then the bugcatcher approach is
more attractive (but I might still contemplate using a tuner.. just put
it in bypass mode when you're "in-band")... and the bugcatcher is still
going to be down 6dB from the dipole.





On the "plus" side, though, you do get a lot of physical ruggedness,
since the whip itself is sturdy and easily replaceable if it has a
serious disagreement with a tree branch.