I found very much the same as Rick did, with a completely different test setup (SDR-IQ).
The details vary somewhat with frequency and supply voltage but the three main features reported by Rick remain the same: >the ten watt PA has worse IMD just before the big PA kicks in, in the >11-12 watt range. >Then it is very clean (approximately -40 db for third >order) at its best, and ok in the 30 watt range (where it would be set to run >my 8877 amp) > but then gets quite ugly above 65 watts. And I mean in the -20 db range at 100 watts. My figures confirm those three main features (assuming we're both reporting 3rd order IMD, "dB below either tone"). The other important feature is that the levels of higher-order IMD do not fall quickly when tuning away from the main signal. If we had reported "total occupied bandwidth at -60dBc", those results would not look pretty at all! Those measurements were first made about 7 years ago on my original low-serial K3, and today I made similar measurements on the new K3S. The 10/12W PA is different from the one in the K3 (two TO220 RF power transistors in the K3S, where the K3 had three). However, the 100W PA is the original module from my 7-year-old K3, so the IMD performance of that module may not be representative of current production. Today, anyone can make 2-tone IMD measurements using a modern SDR and a little care in setting up the input levels. The K3 itself provides the 2-tone test signal (CONFIG:2TONE On). If anyone can report comparable results for an all-new K3S/100, that would be very interesting to see... but don't expect any radical changes in those three main features reported by Rick. 73 from Ian GM3SEK >-----Original Message----- >From: Elecraft [mailto:elecraft-boun...@mailman.qth.net] On Behalf Of >Rick Stealey >Sent: 01 September 2015 16:18 >To: elecraft@mailman.qth.net >Subject: Re: [Elecraft] K3S IMD Questions > >I have been playing with a Flex 6300 here in my shack recently. I consider it >a great piece of test equipment for analyzing signals in the shack as well as >on the air. It does everything and more, as a spectrum analyzer, that my >old HP141 analog setup did, except it doesn't have a tracking generator. >My measurements are not completed and I am still experimenting, but >here is what I've done so far. Phase noise, cw transmitted bandwidth, and >2 tone IMD tests. My two rigs are a K3/100 and an FT100D (how good, and >how bad can you get!) >Using a Flex 6300 as a spectrum analyzer, I fed a greatly attenuated signal >into it, keeping the level just under 0 dbm, and assuming it is not >generating imd of its own, but I haven't quantified that yet. >Phase noise tests - I was interested in seeing if it could display the >RELATIVE benefit of the new synthesizer on my K3, looking at phase noise. >Here are PRELIMINARY RELATIVE results, without numbers, in order of >phase noise level at 2 khz separation:An 8640B signal generator (known to >have low phase noise)The K3 with original synthesizerThe FT100 After I >install the new synthesizer I hope to be able to compare it to the original >one. >I verified the FFT spectrum analyzer (Flex 6300) can give an instant view of >transmitted bandwith of the cw signal sending fast dots. MUCH easier and >better than the old analog HP spectrum analyzer which had to sweep >extremely slowly across the band with a narrow filter. With this setup you >hold the key closed, and hit the control-print screen buttons on the >keyboard and you have a beautiful picture ! Ain't technology great? >2-tone IMD measurements -The two tone test of the K3 showed something >interesting, since it has both 10 watt and a 100 watt PAs installed. Starting >with low power and cranking up the power control, you notice, of course, >that the ten watt PA has worse IMD just before the big PA kicks in, in the >11-12 watt range. Then it is very clean (approximately -40 db for third >order) at its best, and ok in the 30 watt range (where it would be set to run >my 8877 amp) but then gets quite ugly above 65 watts. And I mean in the - >20 db range at 100 watts. IF it was overloading the spectrum analyzer >these tests would be invalid, so I need to verify that. As I said these tests >are preliminary. I need a few more hours. > >In my humble opinion, running at full power, 100 watts barefoot, my K3 >would be transmitting an unacceptably sloppy signal. Keeping it in the 65 >watt range or under, or driving an amp would be better, provided your amp >is clean. My 8877 amp is VERY clean. I don't know about a solid state >amp. Some of the ugliest signals I have seen on the bands with the Flex >panadapter in very limited observations have come from Expert 1500 watt >amps (2 examples). The very best was an Anan 100 with a Ten Tec 1500 >watt tube amp. The waterfall was a brick wall - NOTHING whatsoever >outside the 3 khz bandwidth. >Rick K2XT > > > > > >______________________________________________________________ >Elecraft mailing list >Home: http://mailman.qth.net/mailman/listinfo/elecraft >Help: http://mailman.qth.net/mmfaq.htm >Post: mailto:Elecraft@mailman.qth.net > >This list hosted by: http://www.qsl.net >Please help support this email list: http://www.qsl.net/donate.html >Message delivered to gm3...@ifwtech.co.uk ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
