Finally fixed (but knock on wood)

After more than 4 weeks of trial and error with the new antenna… Well, I’m not going into this matter anymore. But it *seems* that I got it fixed now. Finally. You never know, but after 90 QSO on 160m (including W and VE) and 160 QSO on 80m without a single hitch…

Last Wednesday I was trying and testing again. I knew it had worked with the inverted L on both bands in CQ WW. And I knew the dipole got matched on 160m once although last week it was a ‘no go’ again. I figured out that I was having two different problems with the MFJ-998 QRO AutoTuner. It took a while to get this so I was walking around in circles.

  • Problem 1: No match on 160m. It occurred on the 80m dipole but it happened at first on the vertical too. It got solved by placing an emergency order on So I concluded that still there were stray currents on the shield of the coax, messing the tuner’s circuitry.
  • Problem 2: the amp giving an ‘arc fault’ but only on 80m from about 30W of drive (300W out). On 160m I could get a full kW out. This one drove me crazy. If you know what the problem is, you can fix it. But this could be anywhere. In the antenna? The radials? The feedpoint assembly? The coax? The MFJ itself (the horror!). I tried everything over and over for the last four weeks except the latter.

So it was time for affirmative action. Problem 1: get more ‘choking’ right at the feedpoint. I had 5 big ferrite toroids (outer diameter 60mm) in my stock so I glued these together and wound a 1:1 current transformer on it with 2x 1.5mm² wire. I put this right at the feedpoint which I had to take apart again. The cores were just hanging there in the open, not in an enclosure, for temporary testing purposes only.

Remember my soaked feedpoint (link)? I did not want to have this again although it never happened before so I packed this plug very watertight. After removing the self amalgamating tape there I discovered the female PL quite damaged by sparks. The inner insulator was black and showed some cracks! This must be the cause or result of problem 2 (amp tripping with arc faults). The reason was pretty obvious.  I put some Penetrox in the receptacle where the male PL’s pin goes in. For some reason (too much?) the Penetrox got out and flowed between the center pin and the outer thread. As a result, the insulating distance between shield (GND) and center tip became too small and when the voltage between both becomes too high: sparks. OK, found a solution for problem 2 while fixing problem 1. Great! I made a quick temporary assembly to jumper the cracked female chassis PL. Testing time!

On 160m, the MFJ found a stable match right away. I could pump out 1kW on 160m in the inverted L. Now I was anxious to test on 80m. With the Penetrox and cracked PL removed, there should be no arcing anymore. Well cut off my legs and call me shorty! 30W made the amp trip. I told ON4BHQ, who followed my recent antenna woes closely, that I was about to smash the dreaded MFJ with a hammer and then run it over with a bulldozer. After all: I just changed everything for a third time. I spent all my free time during the last 5 weeks on this @#{@|{|@{#@ project. What else could it be than something inside the MFJ-998? The good thing was the temporary current balun did the job to isolate the feedpoint. I went shopping for an enclosure the next day and finished the balun.

balun1 balun2


I planned LZ DX 4 weeks ago but spent that weekend trying to get the antenna working. Fortunately CQ WW worked out ok. Last week for the UBA Winter contest I changed my last minute plan because of no 160m. Now it’s 9A CW and once again… I pulled myself together and installed the finished balun. That left me with the arc fault on 80m. There were two things I relied on and didn’t test. Being a) the coax from the antenna tuner to the box in the garden and b) the jumper from this box to the actual antenna. The big coax (a) worked fine with the previous antenna on 80m. The jumper (b) was a new one after the soaked feedpoint issue. And both had worked on 80m during CQ WW. So why suspect this?

Stupid me. You shouldn’t trust anything. Nothing is too obvious not to check it. Yesterday I put a loose coax from MFJ to the antenna effectively bypassing cables (a) and (b). A long shot but there you go: full QRO on 80m without arcing. Yeah, don’t tell – I know. I cut the sausage in half and used the big RG-217 coax from MFJ to outside box, and replaced the new jumper with the loose cable. Once again success: full QRO and no arcing. So the arcing ghost I chased for over 4 weeks hides in a newly made 4m long jumper. I need to scrutinize it to see what exactly is wrong with it. I never had a faulty cable in my ham career. So I need to know why this jumper makes my amp trip. I just taped all plugs so I could play in the 9A CW contest this weekend. More on that later.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.