I use close-fitting lids to reduce possible waveguide modes and fit microwave-absorbent rubbery plastic stuff near the input of amplifiers as Geoff suggests.  Sometimes I have to glue a few bits to the PCB, sometimes just in the lid. I have milled a few covers like the ones on the ex-cellphone kit, with separate pockets per stage, but that's a lot of faff. 

I have some Eccosorb from RF-Microwave that works well.  Wurth do some that is powered ferrite in rubber with self-adhesive backing, but that's really only for lower frequencies https://www.we-online.de/katalog/datasheet/34401.pdf

This stuff from Farnell is rated to 10GHz and is very thin (0.3mm) and self-adhesive. https://uk.farnell.com/kemet/efa-03-240x80t0800/noise-suppression-sheet-240x80x0/dp/2949250

Neil G4DBN

I've tried a 10dB pad on the output of the txtr to the PA, it doesnt help (except in reducing the power out of the PA) ... in between the two stages of the amp section is a SAW filter and the oscillation is always in band. This leads me to believe it really is RF feedback through the input stripline going round the PA loop ...  I'm going to invert the board as a first step, I am pretty convinced it is not the devices going into self oscillation in themselves as without the PA attached it is as clean as a whistle on the spectrum analyser, and it is not the match the tvtr is seeing, as the 10db SMA pad would have sorted it.

On Sat, 26 Sep 2020 at 11:03, geoffrey pike via groups.io <gi0gdp=yahoo.co.uk@groups.io> wrote:

Hi Robin,

Have you tried some microwave damping foam ?, you will be surprised in it effectiveness.

My 13cm transverter that i built needed a wee bit here and there to tame it.

I use microwave plate warmers that you get from Lakeland to do the job but if you can get

the proper stuff then all the better. nothing worse than oscillating PAs. Even after this a handful of

sma 3dB attenuators are useful in stabilising things

Geoff

GI0GDP

On Saturday, 26 September 2020, 01:35:31 BST, Robin Szemeti - G1YFG <robin@...> wrote:

Hi,

Has anyone any experience of this kit oscillating on transmit?

Even with no LO and no IF signal applied, the output pair of devices bursts into oscillation.   When you transmit an FM or reasonably strong SSB signal, it is enough to "swamp" the oscillation and make it appear to work correctly, but on quiet SSB sections it oscillates again.

If I remove the input coupling capacitor from the first amp, disconnecting it from the stripline feed that runs down the middle of the board, it becomes stable, but with the stripline connected, even if I remove the bias from the PIN diode T/R switch on the mixer it burst into oscillation still.

I'm beginning to think that having a 40dB gain amplifier with such a long length of exposed stripline across the board may not be the worlds greatest plan.

Any thoughts?  I'm tempted to construct a tin-plate box around the board, but I am not convinced it will help.

--

-- 
Neil
http://g4dbn.uk