... i can post later if a diagram would help. ...
That would be helpful.
What you're seeing are what would be called "spurious" harmonics: those not related to the center frequency of the xtal controlled driver. Their source(s) are difficult to pinpoint but are generally related to circuit design and the purity of the driver sig.
As you may know (or at least suspect),
any amplifier is designed to operate best within a specific frequency bandwidth, the tighter the better. This is both for the purity of the final signal as it is for the efficiency (power produced) of the amplifier.
Your device was designed for 140MHz. You are asking this same circuit to operate at 105.8MHz. It would be an exceptional (
very,
very rare) circuit that could do that cleanly, let alone efficiently.
Since the harmonic(s) are most probably being generated within the final amplifier's design and component arrangement, the only option for elimination is a "post process" filtering scenario. This solution would not, of course, eliminate the basic problem of the over heating in the final amplifier stage.
So, in essence, You're asking the device to perform
way outside it's design envelope. The only realistic solution is to replace the current tank components (and other stuff as well) with appropriately selected components for the frequency you want them to work with.
Finally, if you can live with the heating issue, you could ( and should, for legal reasons) filter out the spurious elements of the signal with filters placed
after the final amp. This will, of course, reduce the final signal strength, not be an easy task and may introduce signal purity issues themselves.