I was looking forward to CQ WW DX. I now have 99 entities towards my DXCC, with 76 confirmed. I felt sure I would break the 100 mark. While I would not be ready to submit my DXCC application, because I need physical cards for all the contacts, it would feel good to know I had contacted 100 different entities around the world.
But it was not to be.
I worked countries 97, 98 and 99 just before the contest. 7X5QB, Algeria, was on 7135kHz on Thursday night (Friday morning at 0216z) and I worked him after breaking through the pileup. I'd seen D44AC, Cape Verde, on the cluster the previous day and managed to work him before the pileup became to great.
My antenna was not working as well as it should have been, so the amplifier would not stay online. The Reflected Power warning kept tripping. It was late after a long day at work, so rather than pull it all down in the dark, I switched in some attenuation so I did not drive the amplifier too hard. I used a step attenuator and clicked in just enough dB to stop it dropping out.
I worked V26/JH5GHM who was transmitting on 7062.92kHz and listening on 7170kHz. My second vfo and split function worked flawlessly and he was quickly in the log. But something hissed and crackled inside the amplifier.
On Friday night I noted many new entities I needed to work, including Israel, Kuwait, Afganistan and others. But for some reason my amplifier was only giving 50 Watts, rather than the usual 500 Watts. And then it was giving zero Watts. I pulled the cover off and could see one of the transistors was fractured and another was black. I bet all four are bad.
So despite complex protection circuits, I have managed to destroy the FETs again. It's lucky that I still have some spares from the cheap lot I bought from Hong Kong. It was foolish of me to run the amp right on the edge of the trip limit. Even though the protection circuit allowed it.
Enter Comments Here:
On: 11/21/12 6:06 Victor said: |
Hi! Have you measured your bulk of transistors to find the matched quad (or even two pairs)? It\'s important for IMD as well as for overheating the transistors. Second thing - the use of sequencer for input and output relays could solve your problem. |
On: 11/21/12 9:09 Chris AC2CZ said: |
I started with a matched set, but that is long gone. The design sets the bias current for each FET, so gate threshold voltage does not need to be matched. The gain needs to be matched within 20%, but I have not measured them. That\'s a good suggestion. Given I only have 5 transistors left, I could at least try to get them into the best matched pairs. I\'ll look into a sequencer. I don\'t have that at present. I agree that could be an issue. |
Copyright 2001-2021 Chris Thompson
Send me an email