[Elecraft] Power control on 50MHz

[Ian White GM3SEK]

Wayne wrote:
>
>We switch the KPA3 out at lower power levels to optimize the drive, 
>ensuring a high carrier-to-noise ratio at lower-level stages.
>
>In any case, IMD should be quite respectable on 6 m even with the LPA 
>by itself.

Understood; but the simple term "IMD" covers a wide range of issues. It 
isn't all about 3rd order IMD; the overall width of the signal is 
determined by how rapidly the higher orders die away.

Comparing the LPA near its maximum output and the HPA near its minimum 
(say 11W versus 15W) the LPA gives more 3rd-order IMD, but the 
higher-order IMD products taper away more quickly so the signal overall 
is narrower.

There is also the issue of cooling; for example, with intensive RTTY or 
WSJT modes, 10W output would be pushing the passively cooled LPA close 
to its thermal limits. In such a case, it would be better to switch in 
the HPA so that both amplifiers are operating well within their thermal 
limits.

The best solution to meet the wide range of possible user needs would be 
a configurable breakpoint between the LPA and the HPA. If the breakpoint 
could be adjusted between say 6W and 12W (configurable and stored per 
band), that would cover a wide range of practical user needs without 
affecting the SNR of the transmitted signal by more than 3dB.

Then there is W7GJ's point about the gap in available power output on 
50MHz:
>As solid state amps become more prevalent, more folks will be wanting 
>to be able to adjust the power in that range to make sure they do not 
>overdrive their amplifiers.

As well as the wide range of possible power amplifiers, there is also a 
wide range of legal power limits around the world. With my particular 
50MHz PA, the gap in drive level available from the K3 makes it 
impossible to set the PA output to our 400W legal limit - the level 
jumps from 290W to 470W.

It is truly bizarre that output levels between 8W and 13W are not 
available on 50MHz. This issue is as old as the K3 itself, and it really 
does need to be fixed!

However, a complete solution would also include the configurable 
breakpoint between the LPA and the HPA as identified above.



-- 

73 from Ian GM3SEK
http://www.ifwtech.co.uk/g3sek