The AM Forum

Similar to threads here on AMFone.net there have been ongoing discussions in EHam.net concerning fairly bad IM products in modern transistor rigs, e.g.,
compared to the last generation of tube rigs and much more.

One of the latest threads, now quite long at five pages, was originated by
Peter, DL8OV and contains content from many hams right up to adaptive
pre-distortion and even includes queries to major manufacturers asking why their rigs have such high IM products.  

Since there's a lot of duplication, the last page or two will reveal the gist of the discussion.

http://www.eham.net/ehamforum/smf/index.php/topic,100634.0.htmlhttp://www.eham.net/ehamforum/smf/index.php/topic,100634.0.html (http://www.eham.net/ehamforum/smf/index.php/topic,100634.0.htmlhttp://www.eham.net/ehamforum/smf/index.php/topic,100634.0.html)

Wayne, N6KR of Elecraft very promptly responded and said engineers would seriously look into it.  High praise for his quick response. There was some comment back and forth.  The K3 had been pointed out as particularly high in IM.

I have a K3 and a QS1R and thought that some spectral scans at various
power outputs might be interesting and would show lesser IM at lesser power levels.  They did and especially in the 30 watt range, products were sufficiently down to cleanly drive a typical high gain linear amplifier such as an Alpha 89.  I also wanted to duplicate Sherwood's K3 100 watt spec. scan.

My K3 Power output settings have been verified with a LP-100A Dig. vector wattmeter. The wattmeter and all switching, extra coax, etc., is then taken out of line and the K3 fed directly with a short piece of new coax into a 50 ohm dummy load using its nominal power set points.  The QS1R is simply placed a few feet away with a 50 ohm termination plug in the BNC antenna input and with about four inches of clip-leaded pickup wire to the center conductor to pick up signal leakage.  When all the superfluous switching, T/R relays in other amps., etc. were in line at the K3's output, there were serious oscillations and degradation of the spectral scans.  These show up as wah wah in the waterfall in last 100 watt shot.  I wonder now about
others' displays and findings.

Software used with the QS1R is SDRMax 5, latest version, 5.0.0.1
Since there's no direct hook-up the scans are well down from zero dBm on the attached jpg files (obtained direct from computer screen with snipping tool) but are accurate over the air samples with IM products accurately produced relative to the two tones internally generated within the K3.  Eight  times averaging of scans is used in SDRMax to better show IM products with minimal noise or grass in base line.

I believe the two tone generator in the K3 matches the specs. of the stand-alone generator also sold by Elecraft.

It was found that the 10 watt driver amp in the K3 may be a contributor to higher IM products in the 100 watt KPA3 at its higher output levels. I'll let Elecraft make that determination of course.

Pix 1 , 6 watts base 10 watt amp
pix 2 , 10 watts base 10 watt amp
pix 3,  14 watts 100 Watt amp just turned on with internal k3 relay.

I have looked at my transmitters the same way.
Mine are all tube AM transmitters, and much better then the K3 at its rated output.
I think the screen modulator was not as clean, but I never fine tuned the adjustments, and it was only highish on the 1st even harmonic?

You can see some poor ssb signals on the bands, but what really stands out is the class E rigs being 25KHz wide.
It seems like most of them are 20 to 25 KHz wide for whatever reason.

 

That's more an operator problem than a technical one.  ;)

Johnny Reverb isn't dead he just moved.

Every one I have ever seen/heard looks like this:
Not a great shot, but looks at the waterfall. exactly 20Kc wide.

That's not 20 kHz wide in my book. All modulation products appear to be more than 40 dB down from the carrier.  From Part 97:

§97.3 Definitions.

(8) Bandwidth. The width of a frequency band outside of which the mean power of the transmitted signal is attenuated at least 26 dB below the mean power of the transmitted signal within the band.