You have not mentioned how the resonance is achieved in your circuit. Specifically, how the capacitance C is related to the problem or if the resonant circuit is, for instance, a LC-tank in series or in parallel.

If you are simply using a coil and no additional discrete component, the system will still resonate at its "natural resonance frequency" which considers the effective parasitic capacitance Cparas of the coil + wiring. If that's the case, any minimum change of the wires will change Cparas and you will have the scenario that you mentioned.

However, if you select a real capacitor component with capacitance C to be added to the circuit and C is at least one order of magnitude bigger than Cparas (the latter is typically less than few pF for low-frequency circuits), then you have a way to better control the resonance frequency because you have relatively "fixed" values for L and C. If the inductance L is the same (i.e., the coil was maintained when you added the capacitor C), it also means that the resonance frequency will drastically change to other region. If you don't want this, you may need to redesign the coil, thus changing L and maintaining the resonance at 300kHz of your example.

To get more details about this problem, see these "old" but extremely helpful references:

F. Terman, Radio Engineers’ Handbook. McGraw-Hill, 1943.

F. W. Grover, Inductance Calculations. Dover, 2009.

@Agnelo R Silva Thank you for this valuable info and references. I will look into it. To answer your question: Resonance is achieved using an L-Network (an inductor in series with a power capacitor bank). The coil I'm using is parallel to this network.

Do you have estimated values for L and C? Have you measured the Q (or equivalent bandwidth)? Do you know how to estimate such value?

My L is 78.8 mH and my C is 41.9 nF. I have not measured the Q. I do however have the equation: Q=sqrt(RL/RG - 1). It gives me a complex number as my RL is 0.09 ohms and RG is 50 ohm.