Off centre fed (OCF) dipoles are very popular with Hams on HF.

Many claims are made for variations of the theme, but the essential common feature is that the offset feed point provides a feed point impedance that is not too unsuited to 50Ω coaxial feed line at the fundamental resonance of the dipole and half wave harmonics.

“Not too unsuited” does not imply a ‘perfect match’ to the coax, but that a moderate length of coax can transport RF energy reasonably efficiently, but some impedance transformation may be required to deliver the transmitter its rated load impedance over each band.

Most other claims made of OCF dipoles in various forms are bogus and will not be discussed here, fantasy of devotees and sellers mostly.

The challenges in implementation of an OCF dipole are to tune the dipole so that VSWR minima fall in-band, and minimisation of feed line radiation / noise pickup for use of effective baluns (common mode chokes).

## A model for discussion

For the purposes of discussion, an NEC-4 model of an OCF dipole was created. The dipole is about 40m in length and at 10m height above average ground, the model includes the common mode current path that exists on a vertical feed coax, a 4:1 current balun with choking impedance of 1000+j1000Ω at the top, and a 1:1 current balun with choking impedance of 1000+j1000Ω at the transmitter. The dipole length and tap point were tuned for consistently low VSWR across 3.5, 7, and 14MHz.

The figure above shows the geometry of the model, and the green curve is the current distribution on the conductors at 3.6MHz. The feed point is at a point of high current, though not the current maximum. In fact, the feed point impedance is 42+j25Ω (looking into the 4:1 current balun).

Above is the current distribution at 7.1MHz, and again the feed point is at a high current point. Feed point impedance here is 42+j17Ω.

Above is the current distribution at 14.2MHz, and again the feed point is at a high current point. Feed point impedance here is 47-j12Ω.

Above is the VSWR plot over the range 3.5 to 15MHz. It can be seen the there are minima which are quite low (about 1.3) at 3.5, 7, 14.2MHz and they might suggest an ATU would not be needed. Some optimisation was done to get the model to deliver this characteristic which depends to some extent on all parameters of the model. In the real world, you might obtain a similar result with trimming, quite a lot of trimming! The bandwidth on each band is limited, and in any event an ATU would be advised to ensure the transmitter sees its rated load, and delivers its rated power at low distortion.

Regarding 28MHz, yes it does ‘drop in again’ at 28.3MHz with low VSWR, but the pattern has lots of lobes and intervening nulls and it does not constitue a good antenna for that band. Pressed, it could be used on 18MHz and 24MHz, though the VSWR isn’t quite as low.

## System view

Taking a system view, we need to consider the losses in the feed line, baluns and ATU.

Balun losses are significant due to common mode current. The strong tendency of the asymmetric antenna to drive common mode current means that even with high choking impedance, there is sufficient common mode current to drive moderately high dissipation in the baluns. This is one of those cases where a high R balun may be poorer than a high X one (contrary to the teachings of some). This is the Achilles heel of OCF dipoles, either suffer the loss in choking common mode current or suffer the noise and EMC/EMR issues.

Lets assume that you have used 10m of RG213 and that the worst case VSWR within any band is 4, the line loss under mismatch will be worse than matched line by 0.2, 0.1 and 0.25dB for the three bands respectively (Duffy 2001). This is not a large penalty for the benefit of multi band operation.

ATU losses can be expected to be less than 1dB (Schmidt 2003) for a reasonably efficient ATU. Again, quite modest loss.

All up, reasonably optimised for a site, such an antenna system should deliver efficiency (PowerRadiated/PowerIn) of better than 50% on each of the bands. Note though that most of the power lost might be in the current balun, and it might need to safely dissipate up to half the average transmitter power… a big ask at high power.

Key elements of this design are:

- Copper wire dipole;
- effective 4:1 voltage balun used at the dipole feed point;
- effective 1:1 current balun used at the transmitter;
- short feed line; and
- dipole length and tapping point trimmed to steer VSWR minima to desired frequencies.

Note than the Ham world abounds with bad balun recipes!

## References

- Duffy, O. 2001. RF Transmission Line Loss Calculator (TLLC). VK1OD.net (offline).
- Schmidt, K. 2003. A T-Network Tuner Simulator. http://fermi.la.asu.edu/w9cf/tuner/tuner.html (accessed 13/05/14).