[time-nuts] low noise multiplication to 100 MHz

Garry Thorp GThorp at pascall.co.uk
Tue Jan 26 13:12:50 EST 2016


3) Input and output level (?????????)
the oscillator is a HCMOS output, so figure swinging about 3.5V

Any multiplier configuration will produce lots of different harmonics, and will need fairly serious filtering after it if you want a clean 100MHz.

If you have a 20MHz oscillator with CMOS output, it should produce a healthy component at 100MHz provided its output is reasonably symmetrical, say ~55:45 or better. The easiest (and probably lowest noise) approach might be to extract the 100MHz component directly. A common-base circuit such as the attached sketch would probably work OK.

Rin is needed to avoid overloading the CMOS output. The input tuned circuit Q could be made quite high, as its ESR can be incorporated into Rin. As this would take very little signal current from the oscillator at 20MHz or 60MHz, most of the CMOS output current capability would be available at 100MHz.

A common-emitter amplifier with a tuned emitter load would take less current from the oscillator, but probably isn't practical if you only have 5V regulated supply available.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: CMOS oscillator x5.pdf
Type: application/pdf
Size: 84038 bytes
Desc: CMOS oscillator x5.pdf
URL: <http://lists.febo.com/pipermail/time-nuts_lists.febo.com/attachments/20160126/ae3b4be8/attachment.pdf>

More information about the time-nuts mailing list