The weak link is the series capacitor feeding the top element. The stock antenna uses a short piece of special coax trimmed to the correct length. Ever wonder why some people report blowing the cap up with just a couple hundred watts while others insist the can put 1 KW into it? This all depends on how clean the end of the coaxial capacitor was cut when it was tuned. There have been some improvements in this area since the first production runs. If it was a perfectly clean and flush cut it may have an RF breakdown voltage around 1500 volts. If there are any imperfections in the cut causing the end to not be concentric or any of the braid to come unraveled, the breakdown voltage can fall well below 1000 volts. Unfortunately there are no commercially made capacitors of the correct value that can fit in this fiberglass radome and the reason this part must be custom fabricated. The stock capacitor is made from RG-303 coax cable. It just so happens that RG-393 coax has the same velocity factor, and capacitance per foot as RG-303 but it has well over twice the RF breakdown voltage at 5000 volts. This makes RG-393 the perfect Teflon coax cable to fabricate the custom upgraded capacitor from. The bad news is it's very expensive costing over $100 for a new 10 foot section to experiment with and I'm not selling any. If you want to try this modification on your own, the most important thing is how you prep the open end of the coaxial capacitor. You'll notice the braid is pulled back over the Teflon by about a 1/4 inch to prevent arcing. What is not so clear is that the center conductor inside the coax has also been pulled back inside the Teflon by the same distance. The hole in the end was filled with clear silicone and plugged with a piece of Teflon cut from the insulation to completely captivate the center conductor in Teflon. The length of the new capacitor is determined by the length of the braid on the stock cap. The distance the braid overlaps the center conductor determines the capacitance value. The new cap will have a longer Teflon center but the length of the outside braid should be identical to the stock one. I unwrapped the braid and formed it to fold under the board to solder flat. Using a Dremmel tool I cut the bottom of the board through the center of the two holes the wires originally went through so the new braid could be soldered in the exact same position. The top black wire is soldered to the top braid on the cap since it does not fit well inside the fiberglass when the connection is at the bottom. The copper trace on the bottom of the board is spaced too close to ground and can arc. This is why the center conductor are soldered on the top of the board. I Used clear silicon to insulate this connection from the board and seal it. The unused copper pad on the bottom of the board was removed to prevent arcing. The lower circuit board that connects the matching stub to the two orange cable also needs its center pad modified in the same way. All three center conductor should be pulled from the board and soldered on top with the old pad removed. Seal both boards in heat shrink tubing that has glue inside and you should be good for 1500 watts PEP in intermittent communications duty cycle. I wouldn't go past this point due to the orange coax being the next weak link in the chain. It's average continuous duty cycle ratings indicate it should have no problem at 1500 watts PEP for intermittent communications service. Beyond this and you risk melting the coax or the connector in the center of the harness. Using larger coax prevents you from wrapping enough turns around the base to form the choke. Almost forgot: The length of the top black wire must be shortened by the length of the braid on the capacitor that was added in series with it. It is the difference between soldering the black wire to the board or soldering it to the end of the braid so it fits inside the antenna. Trim it with a VSWR meter checking the bandwidth ends to insure you get it where you want it without passing the point.