it looks to me like there are several parameters in the KarplusRings engine that are not clamped to safe values and could easily blow the thing up.
in particular, the last parameter looks like “bpf_res”, which if “res” means “resonance” is a misnomer - it is mapped to the rq parameter of the BPF UGen, so controls the reciprocal of its Q/resonance.
in other words, the rq value specifies bandwidth in proportion to center frequency.
setting rq=0 makes an infinitely resonant biquad; noise is fed through this into the input of a Pluck karplus-strong resonator. this could certainly blow up the resonator.
RQ should be clamped to, i dunno, 0.00166666 or something. (that’s a bandwidth of 1 cent if my math is right)