Monday, March 30, 2009

This is how bad it is

This morning I was listening to VK9LA calling CQ on 3523. He was very weak about 5 dB out of the noise but Q5 most of the time. On 3522 there was a QSO. Here is a shot of one of the stations down the band. This guy was not very strong, only about S-7 but just look at his bandwidth on skimmer! My listening freq is where the little green arrow is on skimmer. You can see the tails of the key clix extend well past my listening freq. If you look you can just see the VK9 start to transmit near the arrow. The VK could be copied but so could the clix and no amount of filtering on my end will solve this problem. Note skimmer uses intensity as the way it displays signal strength. If you look closely you can see the key clix are just about the same intensity as the VK9. Also I can tell you from this picture this guy is transmitting through a 2.7khz crystal filter because that is his bandwidth, and the offest of his CW carrier is about 1000hz. Just by switching in his CW filter on transmit this guy could dramatically improve his click problem. See how easy the fix could be even for a legacy transmitter. In fact there are some manufacturers that do just that. You can tell because they have a crappy signal contained within a 500hz bandwidth. You can pick up this kind of signal from the panadapter alone. When your signal has clicks your signal looks like the Eiffel tower with little bumps going up the sides. The higher the bumps the worse the clicks

For comparison here is a shot of the guy to whom he was talking. This fellow is about the same signal strength. I won't publish the first guy's call but he is a W1 with a fancy schamncy Extra call. I'm sure if I sent him these pics he wouldn't have the first clue how to fix the problem because it probably has been manufactured into his radio. If this signal was on SSB guys would be checking in raising hell about the buckshot in this guys signal, but because its on CW this guy probably will never know how bad his signal is. If this was an SDR, first of all you would never see such a distorted and broad signal, unless you were flat topping a linear. Second if you did have such a signal you would merely adjust the code (as in software, not as in Morse) and clean things up, and in one fell swoop you could cure all the SDR's that you have sold. You can tell this is not a linear flat topping situation since the clicks are finely delimited to 2.7khz. Linear flat topping would extend even farther up and down the band and would tend to be symmetric.

A picture is worth a thousand words. Its amazing how much information you can glean from this system. Why if I was an OO I could run amok writing people up!!!