Hi Dave, Thank you for the pictures. Worth thousand words, etc. And there IS difference between the two. Can you detail the setup from end to end, particularly your RX setup and K3 TX voice settings?
I'm not sure what your audio waveform is from exactly in A and B, you'll need to tell us. But if that's human voice, it's a very base-dominated voice spectrum, unlikely to generate obvious splatter. The 0-400 or 0-500 band, whatever it is, is 18 db above the next strongest band in your "modulation quickview", and so the lowest band is completely controlling the power, not the attack of the silibants. This would not be the measured waveform for a contest optimized voice, with the treble accentuated for "punch", where splatter-producing TX/amp setups show their true colors. Set up for contest "punch". Do the comparison with the lower frequencies severely reduced in the K3 TX equalization, and the compression up around 20-30. Or get your wife to do the voice and have her holler at the mike like it was the kids (only a few assumptions there...). Turn OFF the averaging in the program and leave the peak hold on. The duration and hence the average power of silibants are quite less than the lower frequencies, and the distortion products off silibants even less. It is the distortion's instantaneous peak power that causes the most problem, not its peak power times duration (average power). The speech used for the test should be on a recording. Run the K3 barefoot off the recording, bypassing the amp, measure and make your pictures, and then include the amp and repeat from the recording without changing anything on the K3. This is probably a 25 watt A/B and a 100 watt A/B. The receiving device should be run with FIXED gain, NO RX AGC, any overdrive issues worked out in advance. Compare those spectrums. But even this is still approximate, because the RX device (unless it's a very good RX device like another K3 or K2) can create the highest distortion which covers up any true differences. The real test would be to run a very stable square rise and fall CW dit with a pulse width of about 2 or 3 milliseconds with a repetition rate of about one second, and repeating the test for longer pulse widths. This requires a signal generator followed by a high power lab linear amp instead of a K3 to drive the amp, but would isolate added spectrum that occurs on predictable frequencies at predictable times and can be precisely measured. The receiving device and presentation would have to be certified for the general circumstances. I don't own this stuff either, just get to drool on it during isolated opportunities. As to others listening for splatter, one WOULD think others could hear it. However, back in the day of controversy over what rigs had and didn't have key clicks, it was amazing how many people could NOT hear or otherwise detect the bad clicks on a MEASURED dirty rig, for whatever reason I really don't know. So I'm skeptical since I don't see much difference in hearing splatter and hearing key clicks. The problem I have, even if the SBE does manage to do amplification relatively cleanly, is if someone else reads this thread later on without any blowback on the claim, the readers extrapolate to other amps and there it goes... So the SBE method is not going without blowback. SBE may think they have created a marketing differentiation, but for the time being they've crossed themselves off my list on account of recommending smashmouth drive and ALC. What else that I don't know about, yet? 73, Guy. On Sat, May 15, 2010 at 6:32 PM, David M. Elliott <immondi...@earthlink.net> wrote: > > > These two screen shots were taken using a WaveNode WN2. The exciter was a K3 > and the amplifier was the Expert 1K FA. Both were SSB at 3.952 MHz. The > conditions for both of these screen shots were the same except that the first > one was using an exciter power of 25 watts and the second was using an > exciter power of 100 watts. The amp was using ALC control in both cases. > > I don't see any signs of splatter in either case, do you? > > This test seems to support what the Expert folks are saying. > > 73 de W6BK > > > On May 15, 2010, at 1:48 PM, Guy Olinger K2AV wrote: > >> Because if you are using something you can measure, like two tone, the >> gains are already set and steady. To measure leading edge distortions >> before an AGC loop or loops bite and settle takes some pretty >> expensive lab equipment, both before and after the transmitter/amp. >> What you got? >> >> It IS terrible advice. >> >> 73, Guy. >> >> On Sat, May 15, 2010 at 4:25 PM, David M. Elliott >> <immondi...@earthlink.net> wrote: >>> If increased distortion is guaranteed, why doesn't it show on the spectrum >>> scope? >>> >>> We are obviously missing something here. >>> >>> 73 de W6BK >>> >>> >>> On May 15, 2010, at 11:03 AM, Vic K2VCO wrote: >>> >>>> Sorry, with all due respect this is HORRIBLE advice. >>>> >>>> I won't argue with those that say that the ALC should be connected for its >>>> protective >>>> value. But it should not be depended on to control output level -- the >>>> exciter should be >>>> adjusted to produce approximately the drive required for full output. >>>> >>>> The SPE Expert may have an exclusive design, but it is not exclusive >>>> enough to include a >>>> time machine so that it can know how loud the operator is planning to >>>> talk, and it doesn't >>>> control the delay inherent in the exciter's ALC circuitry. >>>> >>>> The result is guaranteed to be increased distortion. >>>> >>>> On 5/15/2010 4:25 AM, Björn Mohr wrote: >>>>> 2010/4/22, LA8AW<odd...@gmail.com>: >>>>> >>>>> >>>>> Here is the statement from the developer of the SPE Expert 1K-FA - >>>>> regarding >>>>>> ALC, quoted from an early Expert Reflector:.... >>>>>> >>>>>> I hope this will clarify the SPE Expert vs K3 - and ALC !! >>>>>> >>>>>> We really suggest to always leave the transceiver's output power manual >>>>>> control to >>>>>> its maximum level because the EXPERT, thanks to its exclusive design, >>>>>> will provide to >>>>>> control that level as long as the required optimal value will be reached. >>>> >>>> >>>> -- >>>> Vic, K2VCO >>>> Fresno CA >>>> http://www.qsl.net/k2vco/ >>>> ______________________________________________________________ >>>> 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 >>> >>> ______________________________________________________________ >>> 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 >>> > > ______________________________________________________________ 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
