F8DRA recently posted some details of an internal flashover problem with a 42,7m (140′) dipole, 6.7m (22′) of 300Ω ladder line and a MJF-986 ATU. Though he did not mention the frequency (that I saw), let’s analyse the antenna system at 7.1MHz.
We might reasonably assume that at resonance around 7MHz, the feedpoint impedance would be of the order of 2200+j0Ω, lets assume that for this study.
The MJF-986 ATU is a differential T tuner.
Though the MJF-986 ATU manual does not specify the capacitance, the above extract from the MFJ catalogue describes it as a 300pF differential capacitor.
Now let’s model the scenario in SimNEC.
Above is a SimNEC model of a similar scenario, the components are not exactly the same as F8DRA’s, but similar and allow discussion of the risk of flashover for that system if is as he describes and there were no faults.
The first thing to note is the dipole is approximately a full wave and the center feed impedance will be very high, perhaps as high as 4400Ω depending on proximity to ground, 2200+j0Ω is assumed here.
That feed point impedance is transformed by the two wire line, the input impedance to T1 is 43.34-j124Ω. This is the load seen and matched by the ATU.
C1, L1 and C2 make the MFJ-986 ATU. It is a differential T circuit, so C1+C2=300pF. The ATU has been adjusted for a match.
The model G block is set to supply 1000W to a matched load so that voltage calculations apply to 1000W input. Note that all voltages and currents shown are RMS. The model also calculates vc1 and vc2, the voltages across C1 and C2.
Analysis
Some key points:
- vc2 is 1060RMS, or 1500Vpk which is quite high;
- power lost in the roller inductor is 96W which is a lot of power to dissipate over the small part of the roller inductor that is in use (perhaps around 6 turns);
- 49W is lost in the quite short feedline, 0.24dB loss, open wire lines are often thought of as (nearly) lossless; and
- 851W reaches the feed point, somewhat less will be radiated due to conductor and ground losses.
Note that these numbers come from a scenario of just one third of the power rating of the MFJ-986 ATU.
Conclusions
The assumed feed point impedance is mid range, real antennas could be higher which would imply higher system voltages.
This is just an analysis of one band, albeit a likely flashover case. Other bands may be worse. The analysis is of the stated scenario, and even small changes might alter the results.
Without knowledge of just where the flashover occurs, it is difficult to make comment. However the voltages inside this ATU are very high and may be higher in F8DRA’s system.
Because the capacitance values are small in this ATU, voltages are likely to be higher than would occur with an ordinary T tuner using 500pF capacitors. The convenience of the differential T circuit and this implementation come at a cost of efficiency and voltages impressed.
Update 08/1124:
F8DRA advised online a correction to his scenario, the transmission line is 450 ohm windowed ladder line.
Above is the revised SimNEC model. The outcome is a little different, a little better loss figures, but voltage vc1 is almost the same, and the conclusions stated remain applicable.