Metallic screwdriver effect on trimmer (variable capacitor)

Im currently designing an RF PA (in the range of 750Mhz, 10W output)

The output matching circuit to the antenna consist of segments of transmission line and variable capacitors (trimmers).

Originally, the PA oscillates, so I try adjusting the trimmers. The thing is, I don't have a plastic screwdriver for adjusting them, so whenever I touch the trimmers with a metallic one, I modify the output matching circuit.

There's a trimmer configuration, where If I touch one of the trimmers with the metallic screwdriver the PA works great, but as soon as I let go it oscillates or gain drops (And I can't seem to get that state besides touching the trimmer).

My question is, whats the effect of a metallic screwdriver and a person holding it on a trimmer, so that I can simulate such 'circuit' with discrete components? (for instance, if touching it makes the capacitance higher, i may add a parallel capacitor so that I dont have to touch it to make it work; or maybe the effect is that of an open transmission linem, you get the idea)

Thanks for any feedback.

You probably change te value of the trimmer capacitor regarding to ground.So,input-output matching circuit has not carefully been designed taking the stability into account.Check the circuit and re-build it again..
Becuse there is no such model what you did to trimmer capacitor, you only create some amount of parasitic capacitor..

Jeez, the answer is obvious. Go out and buy a crate of screwdrivers, and ship those suckahs!

The problem isn't the metallic screwdriver (if anything, they help to pinpoint the design fault). Still, It's good advice :)
I'll get to redesign the PA, though there isn't much I can do other than trial and error; the data provided by manufacturer is only the series equivalent for optimun load. Other than matching that, i'm out of ideas.

I used to have a good tech. One day he said "I can get this unit in spec to ship it if I hold my finger right here". That was good temporarily, but I was only able to ship 10 units that way.

Hello,

Touching a trimmer will cause additional capacitance, but there is more. You also add a real part (as the induced current in your finger produces heat).

This real part may be the reason that your PA becomes stable, just adding some capacitance in your simulation does ignore the real part. When you want to model this, you need a series resistance or parallel resistance. What is best? I don't know. I never measured S11 for trimmer with my finger on it.

Just for my interest, when adjusted for good performance (with screwdriver on the trimmer), is the oscillating frequency in or out of band? If out of band you may add some kind of diplexer circuit to make sure the transistor sees some real part.

I ended up redisigning the output matching network, after breaking one of the PCB tracks after much soldering / desoldering (now it has wider tracks, and more vias all around). Now I't works like a charm, and I get to keep my fingers attached to my hands.

Hello,

Good thing! What type of active component did you use?

I'm using a freescale LDMOS PA module, don't have the exact model at hand, but it was something like MWxxxxxxx

maybe this advice could help: to minimize the tunner stick effect, try to solder the tunning pin of the trimmer to ground.

You have to use a plastic screwdriver (it can be buyed in every electronics shop) because the metallic one cannot be used for a variable capacitor.
Is not a problem of your project but of the metallic screwdriver
see also www.uwaverf.altervista.org/

and for absolutely no cost make one out of a bit of plastic shaft eg.. an old knitting needle trim up one end with a sharp knife :)

done in 5 minutes

Dave