[Elecraft] KPA100 power on 20m/30m, update

[Ted Turk]

Geoff, Don,

As an ex sufferer of this disease, my 2p worth.

My investigations, admittedly only with the limited kit available to me 
including 20MHz scope and 2 or 3 different wattmeters, I think it was, 
showed that the output across all the bands up to 21MHz was remarkably 
constant for a given Q1,2 base voltage (and attempt at current 
calculation).  However, there was a very marked voltage (estimated 
current) loss across the base drive circuits on the affected bands 
(C80,81/R19,20).  Readings prior to this in the RF line merely 
confirmed the high drive requirement as did the Hi Cur on 20/30m.  With 
much help and advice from guys at my club, we too considered the 
possibility of other spurious resonant circuits but could find no 
evidence for these in my kit at a level that affected performance.

73,
Ted G7BQM K2#4732 and first year RF engineering student!

On 24 Jun 2005, at 02:31, Geoffrey Mackenzie-Kennedy wrote:

> Hello Don,
>
> Yes I understand the purpose of the RC compensation circuit in the 
> drive
> path to each transistor to offset their decreasing gain as the working
> frequency increases, and which cannot be fully dealt with by the RCL 
> network
> in the primary circuit of T1. As you say say they are not intentional 
> traps,
> which is why I use the term "spurious resonant circuits".  The 
> possibility
> of unwanted resonances in the feedback networks eg R39, C59, R35 is 
> very
> real. If for example R39 and C59 were to be a series resonant circuit 
> at X
> MHz,  the in-circuit gain of Q2 would decrease at X MHz.
>
> RF design is an unforgiving mistress!!
>
> 73,
> Geoff
> GM4ESD
>
>
>> Geoff,
>>
>> These are not intentional traps, but are a parallel RC circuit to 
>> adjust
> the
>> base drive with frequency (at low frequencies, the resistors are the 
>> main
>> path, at higher frequencies, the capacitors pass more RF).  This is 
>> done
> to
>> compensate for the gain vs. frequency characteristics of the 
>> transistors.
>> The inductance of the resistors is just an unfortunate by-product (but
> must
>> be considered just the same in the design).
>>
>> 73,
>> Don W3FPR
>>