I've seen a lot of Cobra29/Uniden PC-66/68/76/78-type radios with a particular odd habit.
When they remove the jumper wire that feeds modulated power to the driver transistor, a problem arises. One, single disc capacitor to ground is used to filter stray RF from leaking "upstream" from out of BOTH the driver's output circuit AND the final's output circuit. One disc cap to ground (on the frontwards jumper-wire hole) filters (decouples, bypasses) BOTH final and driver stages.
Removing the jumper wire, to insert either a fixed resistor/electrolytic OR a transistor for "variable", now UNHOOKS the bypass capacitor from the driver's output circuit.
Folks who patch a 2-Watt control across the jumper wire's holes in the board, will turn it to get the desired carrier, unhook it and read the pot with an ohmmeter. Then, a fixed resistor close to that value SHOULD give you the same carrier power, right?
The answer is, only sometimes. The "missing" bypass disc cap makes the driver stage unstable. The fix is to add a .02 uf (or so) disc to ground from the rear-most pad where the jumper wire was pulled. Now, the driver has its own bypass cap, and carrier settings will be a lot more stable.
Been waiting for one of the online descriptions of "Put Variable In Your Cobra 29" to include this little detail. Still waiting.
73