[time-nuts] locking oscillators - an increase in power and/or stability ?

[Dr. David Kirkby (Kirkby Microwave Ltd)]

On 8 Oct 2014 20:18, "paul swed" <paulswedb at gmail.com> wrote:
>
> David the locking makes sense but the other numbers do not make sense.
> Combined best case would be +3 DBm also you say the current on the diodes
> goes down. More power for less power in is not adding up.

I went to a users meeting yesterday at Keysight (formally Agilent) on mm
wave / THz things. I discussed this Gunn diode issue with a couple of
people.

1) There was a guy from Virginia Diodes

http://vadiodes.com/index.php/en/

speaking. His company makes commercial extension units that get VNAs
working above 1 THz and sources to several THz, so really are at the
cutting edge.

2) A guy from a Scottish university (forget which) that makes Gunn diodes
at 100's GHz.

I couldn't resist asking both about this.

1) The academic making the Gunn diodes at 100's of GHz could not believe it.

2) The guy from Virginia Diodes was *not* at all surprised.  He said that
the magic T is now part of the cavity.  The cavity has been completely
changed,  so it is hardly surprising that the diodes performed differently
in a completely different cavity. He said that the properties of the cavity
far removed from the oscillation frequency will effect how the diode
performs

They usually use isolators,  but even different isolators change how
oscillators behave.

The fact that the oscillators were then coupled would again change the
conditions, so all the more reason to expect different behaviour.

Dave, G8WRB.