[600MRG] After Market VFO s comparisons (Frank Lotito)

[Hans Summers]

Hi Frank

Does anyone have subjective comments on the U3S vs DDS kits' performance at
various frequencies?  Has anyone performed "quantitative  comparative"
measurements on these kits?  My specific interest is for LF / MF / HF
transmitter application.


Since I designed the QRP Labs products I should warn you in advance that I
could be seen as biased ;-)

First off - there may be a bit of confusion. The QRP Labs Ultimate3S lit
http://qrp-labs.com/ultimate3/u3s kit is a weak signal modes autonomous
transmitter (WSPR, JT65, CW, JT9, etc etc). But we also have a VFO/SigGen
kit http://qrp-labs.com/vfo which has mostly the same hardware but includes
a rotary encoder for tuning, and has different firmware with features for a
VFO.

The Si5351A Synth chip is a digital phase locked loop. It is quite
different from Direct Digital Synthesis and they do have different
performance characteristics.

One key difference is that DRS chips have a sinewave output (and squarewave
is possible using the onchip comparator as a zero crossing detector),
whereas the Si5351A has only a squarewave output and would need to be
followed by a low pass filter to turn it into a sinewave. In many
applications however, a squarewave is preferable in any case.

Generally speaking a DDS has close-in unfilterable spurs that a PLL does
not; and a PLL synthesizer has phase noise that a DDS does not. These
performance characteristics are inherent to different synthesis method
implemented. This is another key performance difference. In practice both
the spurious content of a DDS and the phase noise content of a Si5351A are
both at such a very low level that for the vast majority of applications it
will be unimportant. This is even more so at LF/MF where the reference
frequency is divided by such a large ratio to get to the output.

I like the Si5351A chip because it allows much greater output frequency
range than a low to mid-cost DDS, the output is cleaner, it has three
separate outputs, and importantly the chip is very low cost compared to the
DRS chips!

In terms of the implementation, the QRP Labs kits have the nice GPS
discipline feature (add 1pps signal from a GPS). I'm not sure that the DDS
kit you mentioned provides that.

I realise this doesn't exactly answer your question about qualitative
quantitative measurements but since I didn't see any other answers I
thought I'd comment.

In summary I think that for an LF/MF transmitter application either a DDS
based or Si5351A based solution will work nicely and there aren't any
performance-related preferences that will really matter in that
application.

73 Hans G0UPL
http://qrp-labs.com