What the heck are "zobel" circuits?


I recently bought a pair of Classe 15 Amps from a gentlemen in Toronto. He was a great guy, but a bit of a fanatic. He heavily modified these amps - one mod was to get rid of the 'zobel' circuits at the speaker binding posts. As a safety prcaution, he gave me a pair of 'outboard' circuits to be attached via Banana plugs in the back of my amps. He explained that they prevented runaway high impedance that could oscillate the amps. I noticed that Alpha-Core speakers are now offering what the call an RC circuit built into their speaker cables, that I believe effective obtains the same results. Can anybody further enlighten me?
dbamac

Showing 1 response by gs5556

Basically, it's an R/C circuit shunting the amp output to ground. One problem amplifiers have is the instability resulting from low frequency signals. The amps will oscillate at a high frequency when there is no inductive load at the speaker. The zobel network acts like an automobile shock absorber. The resistor is the spring, which absorbs the deformation energy and the capacitor is the shock piston which dampens the vibration. When an amp sees a capacitive load from high C cables or no high frequencies, it starts to oscillate, or the road's bumpy with the auto analogy and the zobel network matches the speaker load to a resistive load and dampens the isolation.

By removing the zobel network, you remove the matching network interface to the speaker that makes the speaker behave as a resistive load and high frequency inductive loads cannot be damped out. Yes, the amp can oscillate when there is a high inductive load from the speaker (high frequencies) but some crossovers have a zobel network which prevents a back-oscillation. If there is no ZN at the speaker, then no big deal - the ZN at the amp output will take care of it (unless the amp has an output inductor which could cause a problem, theoretically).

I think that the "outboard circuits" he provided you with are nothing more than a matching network consisting of an R1-R2-C1-(R1)C2 and an R3-L1-C3 circuit designed to make the crossover network appear resistive. If it works, more power to him, If it sounds better, then, hey, why not. Personally, I never seen a problem arising out of this and I think this is a non-issue as I trust amp designers to do the right thing when I plunks down my cash.