As my 4ele vertical array for 40m at my contest station broke down due to a tremendous winter storm, I recently repaired it using my home-brew vector network analyser (VNA). Here is an update of the current status - still there is some work to do, which I hope to finalize during the summer period.
Important to note, that all measurements where taken with the indivudual elements one by one. No mutual coupling of the elements took place, as when measuring an element all other where took down. Although this seems to be a rather exhaustive procedere, it could be easily done, as I anyway had to refurbish all antenna elements. Also I have a simple mechanism to bring each element down very quickly, laying on the ground

Here is the procedere which I followed to align the 4ele phased array properly:

Repair of all antenna elements for same frequency resonance

The frequency of antenna being in resonance is defined by the fact that all impedance is real (Rs) and complex impedancs (Xs, capacitive or inductive) is zero. I measured Xs of all individual elements over the whole 40m band. While 3 element are tuned very close to the same frequency there is one element still probably 5-10cm too long to match the same frequency. I will correct this asap.

Try to get the resistive Impedance of all 4 antennas to the same value

This is a rather difficult requirement, as nothing must be done to change the resonant frequency of the individual elements. Probably I will never be able to exactly meet this requirement, it is very desirable for symmetry of the array. As it is swiched electrically into 4 different directions (30°, 120°, 210° and  300°), to meet this requirement would assure the same pattern in any direction.
I measured Rs over the entire 40m band with the individual elements. Again one element is rather off in impedance. I must check asap, what the reason is.
I guess, that there might be an earth system problem with that element which introduce additional losses resulting in much higher in impedance. 

I also checked SWR curves of each antenna elements - what can we learn from these curves ? SWR tells tou nothing, if you don't know the impedances in magnitude and angle....

I love my new homebrew vector network analyser, which shows me the details reliably.

Things to do:

- fine tune as mentioned above
- modify the elements with current probes at each element
- directly measure the antenna current in magnitude and angle following the suggested procedere of Greg Ordy (W8WWV). Greg recently presented a novel way to align multi element phased arrays during the Dayton hamvention 2008. Visit his website to see his Dayton presentation about it