What range of impedances do you want to match?
The old ARRL books used to have link-coupled circuits for matching balanced line.
One of the early 6m solid state rigs published in QST included a small tuner for
matching to a long wire antenna.
The small tuners sold for CB use may match as high as 6m, depending on the
load impedance. You could reduce the coil size and probably make one work.
But there are some important principles that you need to understand.
(1) The exact values of the variable capacitors and the coil aren't critical as long as
they are adjustable over a sufficient range. For example, you may find that you
usually need 10 to 30pf for a capacitor to match the range of impedances you expect
to encounter. You could use a variable with a maximum of 30pf, or 50pf, or 70pf, as
long as the minimum in each case was less than 10pf. All would give you the ability
to tune the required range.
(2) The range of values required will depend on the range of impedances you want to
be able to match. This can also affect the choice of circuit.
(3) The parasitic reactances in the internal leads become significant as you go higher
in frequency. The inductance of the wire from the coil to the switch and back may be
more than you need to get a match. This makes tuner layout much more critical.
(4) Tuners aren't used as often on VHF because coax losses are more signficant, especially
when the SWR is high. That means it makes more sense to adjust your antenna for a
low SWR (which minimizes feedline losses) rather than using a tuner at the shack end.
That said, here are some links that may help you:
You can determine the range of values you need for a T network tuner by feeding W9CF's
tuner simulator a range of impedances and seeing what values it uses to match:http://fermi.la.asu.edu/w9cf/tuner/tuner.html
W4RNL's discussion on designing tuners for the higher frequencies (specifically for 10m,
but applies even more for 6m.)http://www.cebik.com/content/a10/ant10.html
PA0FRI has a pi-network tuner designed for 20 - 10m. A smaller coil and capacitors
would work on 60m - possibly as low as 100pf for the capacitors.http://pa0fri.home.xs4all.nl/ATU/PiATU/pifiltereng.htm
The PI network is a reasonable choice because the shafts of the capacitors are
grounded, and getting reasonable capacitor values is not as difficult as it is at
80m or 160m. One such design I saw used a half-wave coax balun on the
output, feeding two SO-239 coax connectors. Balanced line was connected to
the center pins of the two connectors using banana plugs. A single coax
feedline was connected to one jack and tuned in the normal manner. (Though
a good 1 : 1 current balun may be a better choice.)