Behalf Of Alan
Sent: Saturday, August 30, 2008 2:05 PM
Subject: Re: [softrock40] Re: rxtx+xtall
Has anyone actually verified this?
If it does degrade performance what should be done?
If you need frequencies higher than CMOS what is the alternative?
What sort of output matching should be used for frequencies higher than the
FIN1002 can achieve?
The tests done recently showed the LVDS to have lower amounts of spurii
associated with the internal crystal frequency.
And the specified stability is better.
73 Alan G4ZFQ
If you want really low levels of noise and jitter then you are going to need
to be extremely careful in the design and implementation of the circuit. For
example if the SI570 chip shares power with other devices then they can put
noise into either the ground or power to the SI570 making it appear a lot
noisier than what the chip is capable of.
Separate regulator for the SI570
Buffering of the output signals
Really good bypassing of the power lines at multiple frequencies.
Really good grounding to keep the noise low
Careful layout of the traces of the PCB; make sure they are not near other
signals, such as CPU, and 12C lines etc.
None of this is rocket science but if you ignore these principles you will
have more noise. And just because it's a kit of some sort it doesn't mean
that careful attention was paid to minimize noise.
www.softrockradio.org www.smallcpu.org www.qrpradio.org
"Blessed are the cracked for they shall let the light in."