[hpsdr] KXPA100 Predistortion

[dx at hb9emp.ch]

Dear all,

I recently got a Elecraft KXPA100 PA. The Predistortion tests were not 
very nice at all. Max -40dBc at IM3. As a comparison, with a small 16W 
amplifier (minikits.au, also 13.8V supply) I reach more than -60dBc. 
With a SPE 2k I easily reach -50dBc. This makes me think why the KXPA100 
is performing so badly. Is somebody experienced with this PA?

Any modification ideas? The TRX switching is done by PIN diodes, did 
this cause any problems in relation to predistortion to somebody? I can 
imagine that the decoupling between input and output is not very high 
(probably somewhere about 30 dB?) I could bypass the full TRX switching 
path since I use it in TX only.

Any other ideas what can cause degradation of IM? I know about saturated 
ferrites. Since I also have this bad performance on low output power I 
don't think this is the reason.
DC Supply is well buffered with large capacitors at the input. Even a DC 
ripple reduction by factor 5 did not have any impact.

Generally spoken did somebody improve an IM in an amplifier? If so, what 
was the main degradation reason?

Thanks for your comments.

73
Mischa HB9EMP