Is my filter envelope out of sync? (Filter-based 909 kick)

Hi, first time posting. I have decided to learn Max and the best way I know how to do it is to make a kick in an unorthodox (or rather the most orthodox) way. In the patch you can see that I'm trying to do the 909 kick with filters (as it was initially designed in the analogue world). The basic principle is you send an impulse into a resonant HP filter. The cutoff/centre frequency gives you the pitch and the Q amount gives the decay. To keep the sync and set the envelope I am using phasor~ into ramp~ into shape~ into scale~. This gives me a smooth curve of values that send signal to filtercoeff~ "frequency" and do their magic for the biquad~.

The problem is still somewhere around ramp/shape/scale (most likely the scale~) as I see that the peak values are different. When the sound is on you can hear the peak pitch go in a cycle (sin) rising several hits in a row and then lowering. What could be the reason for that?

It is unlikely to be biquad~ as this happens with all filters (svf, live, etc) although for sure I need to learn filter design in order to use one biquad~ instead of five as I'm doing right now. But that is for the next improvements.

Edit. After googling more about self oscillating filters I see that it is the filters' problem as they don't get enough energy, or there is a very sweet spot for Q I need to find to get it to oscillate on a consistent basis. Compression and overdrive should also help. Thanks for reading and hope you enjoyed this little panic moment I had.