I built a copper pipe colinear jpole antenna (http://home.comcast.net/~buck0/Kb1dig_j.htm
). I tried quite a few things ... moving the lead points up and down quite a bit, adding some copper wire to the top to see if that would affect it, and trying a different coax. They all seemed to yield the same results. (Mind you, I haven't tried it with the new SWR meter... it may be worth revisiting at this point possibly).
J-poles are quirky. One reason is that the coax shield is connected directly to the antenna
at a point that is NOT at RF ground - that almost guarantees you'll get common mode currents
on the coax, and antenna performance (and SWR) become dependent on how the coax is
routed, coiled, etc. If you check the SWR with a short coax jumper, then put a longer cable
on when the antenna is mounted on the roof or on a mast, that will change the SWR. It can
be quite frustrating. (Not to mention that, with some combinations, the radiation from the
coax cancels that from the antenna, and you can get nulls in the pattern where you were
wanting to have maximum radiation.)
For a standard J-pole I cut the radiator (the portion above the top of the stub) to 1/2 wavelength
and adjust the feedpoint and the length of the stub for best SWR. Unfortunately most
copper pipe J-pole designs don't allow for adjusting the stub length at the bottom, and if you
adjust the top then you are changing the radiator length at the same time. I haven't made one
using copper pipe, but if I did I'd probably make the stub longer with a sliding short circuit across
it to adjust the length, and use some sort of a balun on the feedline as well.
But try the antenna as it is and see if it works well enough for you. A 2 : 1 SWR isn't likely to
hurt the radio. Or remove the upper section and phasing stub and see how that works, and if
you can get a better match.
One point that concerns me is your comment that moving the tap points and other changes
didn't seem to affect the SWR:
What type of feedline are you using, and how long is it?
Making such adjustments SHOULD change the SWR, even if it doesn't lower it to 1 : 1. But
a long length of relatively lossy coax can make the SWR low even with no antenna connected.
For example, 50' of RG-58 coax would give a maximum SWR on 2m of 2.7 : 1. So just because
the SWR doesn't look too bad doesn't mean that your antenna is anywhere close to being
tuned properly. An SWR of 5 : 1 at the antenna would look like 1.9 : 1 at the rig end due to
the losses in the coax.
So if you are measuring the SWR through a significant length of coax (particularly RG-58) then
it may be difficult to see the effect of changes in the antenna matching.
And that comes back to why we rarely use antenna tuners in the shack on VHF: it can
match the 2 : 1 SWR to the rig, but you still may be losing 3/4 of your power in the feedline.