[Elecraft] K3 - Revolutionary Noise Reduction

[Tony Fegan VE3QF]

         K3 - Revolutionary Noise Reduction

     It was announced today, by Professor Morsig at the Institute of 
Ljubljana, that the K3 has been chosen to test a revolutionary new 
approach to combating receiver noise and improving HF reception.
     The basic principal behind this approach is the shifting in 
frequency of noise by a few mHz (that's milli-hertz) away from the tuned 
frequency. This in effect produces a hole in the noise spectrum where 
weaker signals may be detected. The residual noise is directly 
proportional to the bandwidth of the hole. Therefore, for wider signals, 
such as SSB, a modified approach is required. The hole is moved 
progressively over the required signals bandwidth and the signal is 
sampled at each step and integrated to reproduce the original wideband 
signal.
     A second feature associated with this algorithm deals with 
variations in propagation. We all know that propagation is largely 
dependent on the solar flux. Propagation is changed due to the effects 
of the solar wind. A little known fact is that the remnants of the solar 
wind that reach the receive antenna modulates the noise. When the noise 
is detected this modulation can be interpreted to give information about 
the condition of the D, E and F layers. This can be used to produce a 
real-time table of MUF and critical frequencies for each layer. At night 
some kind of remote sensing with suitable location corrections has to be 
used.
     The third and more revolutionary feature, which depends heavily on 
the previous features, involves the generation and use of virtual 
transceivers or individual ionosphere layer repeaters. As an example we 
will assume a distant CW signal is transmitted on 21.110Mhz and the MUF 
is around 15.000Mhz. The signal will not be refracted to our receiving 
station and most of it is lost in space. If we transmit a heterodyning 
carrier pulse on 14.110MHz, which is below the MUF, then it will mix 
with the transmitted signal and produce a weak signal pulse at 7.000MHz 
which is reflected back to the receiving station. There is a time delay 
between the sending of the pulse and receiving the return pulse. The 
transmitted pulse length is made less than this delay so that the 
7.000MHz return pulse may be received between transmit pulses. This 
process is repeated and the received signals integrated to reconstitute 
the original signal sent.
     There are some operating problems to be worked out with the 
transmission of multiple heterodyning carriers. Thus only a few K3's are 
expected be loaded with the new firmware until these problems are 
resolved (spread spectrum has been suggested).
     New menu items may be added to include these features in a future 
K3 DSP firmware update. MUF OFF/AUTO and 1-10 will vary the effect of 
received signals using simulated MUF values. This can improve signals on 
all bands. Unfortunately this enhancement is not reciprocal and only 
works on receive. Thus K3 to K3 contacts should have a distinct 
advantage. The algorithms used are not open source as these features are 
expected to have a huge affect on the whole radio communications 
industry. It is still in the early beta-test stages.

     The above is a rough summary of the original unwinese language 
document submitted by Professor G. E. T. Morsig who is the 1st Associate 
Professor of Radio at the Institute of Ljubljana. He will be very 
interested in your assessment of this system and may be contacted via 
F(zero)OL (qrz.com).

     My apologies if I have made a minor slip or two in the translation. 
Unwinese is not an easy language to translate.

     Please do not contact Elecraft for this firmware until there is an 
official statement!

73 and Enjoy
     Tony Fegan VE3QF (anxiously awaiting further developments!)