You don't say whether the noise you have measured is current, or voltage, so I'm going to assume voltage. It is perfectly natural for the motor to draw harmonic current at levels that decrease roughly with increasing frequency. This is in the kHz range; the MHz stuff is likely arcing due to brush noise, more on that later.

If the harmonic currents generate excessive ripple voltage, that means the source impedance of your three phase power is too high. On thing you might do to check it (very roughly), is to replace the motor with a resistive load that draws the same rms current, and look at the open and loaded output potential, and compute the source resistance of your supply, from this simple equation:

R_out = R_load * (V_oc - V_L)/V_L,

where, R_out is the desired output resistance value, R_load is the resistance that draws the desired rms current, V_oc is the open circuit potential of your power source, and V_L is the output potential of your power source loaded by R_L.

The impedance should be measured where you connect to the motor; that will include the effect of the wiring, if any.

It is possible that the source resistance at 500 Hz is suitable, but the source inductance is too high. You might have to look at your wiring interconnection in that case, and also possibly your source of 500 Hz power.

Finally, noise approaching a megahertz is likely brush noise - arcing - unless this is an induction motor, in which case there should be no noise other than that induced at harmonics of the power frequency. Brush noise may be filtered via small caps placed line-to-line and/or line-to-ground at the motor's power connector. These caps can be 1 uF or less, and thus should have a negligible effect on load power factor.