Ideal diode model issue in rectifier

I’m implementing a complex resonance circuit, that is tune for 3MHz. The ac is rectified into the resistive load. I want to work with an ideal diode model for simplification. I used a suggested ideal model from the forum. The issue is during normal operation I get a random spike of current in one of the diodes (each time is a different diode) which exceeds 1MA. This does not occur when using a real diode model.
Here is the circuit:

and here are the waveforms:

The spike appears ID4. You can observe the other current waveforms, and see not spike in them.

I tried adding capacitance to the diode model, and tried different RON and Vfwd combination- but it still occurred.

Am is missing something in the model? Thank you

.model IdealDiode d Ron=20m Roff=1G Vfwd=0 epsilon=0.5 Cjo=5p
The maximum time step may need to be reduced.

Thank you for the reply. I tried the above, but it did not change the situation.

Any chance you can simplify your circuit to a level that problem can still replicate which allow us to replicate such circuit? I simulated a diode bridge but cannot see abnormal spike.

It would be better if you showed the whole diagram. Unless, of course, it’s not a secret. You can apply method=gear and trtol=1 or trtol=7. It is possible to load the scheme by adding txt extension. I do this on another forum for transferring schematics in LTspice format.