[Mep-dev] Dayton Presentation Snapshot 2 (Michelle)

[Grant Hodgson]

Michelle

Ok, only comment is regarding 2nd paragraph, p32 - the sampling rate is 
not a function of the RF filter - it's a function of the IF filter.  If 
a good IF filter is used (say 9, 11 or more poles) then the sampling 
rate can be close to twice the IF filter bandwidth without too much 
aliasing.

I'm assuming that MEP will only want to sample any given 10MHz block of 
spectrum at any one time?  It's a comms. transceceiver, after all, not a 
spectrum analyser - unless I've mis-understood; the synthesiser will 
determine which 10MHz chunk of spectrum is downconverted to baseband.

So a really good 10MHz low-pass filter will allow sampling at (say) 25M 
samples/sec or similar without too many problems.  The RF bandwidth can 
be considerably wider, and would ideally be as wide as the whole amateur 
allocation- 3400-3475MHz/5650-5850MHz.

Regarding phase noise/stability - my wholehearted advice is just don't 
even mention it.  For one thing, it would seem to be at a much lower 
level of detail than the rest of the presentation.  For another, it's a 
very complicated subject in it's own right, and can quickly lead to 
mis-understanding and worse.  Third, it can get quite emotive, partly 
due to prejudice from some previously, poorly designed products from 
years gone past.  Fourthly.... I could go on.  The uWSDR synthesiser 
will be good enough to support both narrowband speech and high-speed 
data, and I'm more than happy to share design notes, phase noise plots, 
jitter calculations, residual FM measurements etc., but I don't think 
that will help to convey the message about what MEP is all about.

Nice photo on p13 BTW.  I could have sworn that you were a bit older 
than that, but I usually get these things wrong.  I'll buy you a candy 
bar next time we meet!

regards

Grant G8UBN