So, I am upgrading the capacitors in the crossover of my speakers. Most of the swaps have been straight forward, but there is one on the midrange that I'd like some advice on. There are 3 components parallel to the woofer; (From Positive -----> Negative) a 10uf electrolytic, a small air core inductor and a 5w 0.27 ohm wirewound resistor. Normally a Zobel Network is just a capacitor and resistor right? Is this some sort of phase alignment network? Anyways, I've figured that I need approximately 0.3 ohms of added series resistance if I replace the electrolytic with a polypropylene film. My question is, does it matter where I add that resistance within the 3 components? For example, should I add a 0.3 ohm resistor before the film capacitor, after the film capacitor, or just increase the original 0.27 ohm resistor to a 0.56 ohm, or does it even really matter? Thanks!
Doesn't matter, since all the elements in a series circuit branch can be in any order or series combination.
The sum of their impedances will be the same, regardless.
The sum of their impedances will be the same, regardless.
Last edited:
No, it is a notch filter, taming some woofer cone break peak at higher frequencies.
Replacing an electrolytic with a film cap in a notch filter might narrow the notch considerably due to the much lower ESR of the film cap vs. the electrolytic. A series resistor might be necessary to emulate the ESR of the original cap.
notch filter terminology is sometimes confusing. The "series" notch type has 3 series element but is put in parallel with the speaker. The parallel type has 3 elements in parallel with each other but is placed in series with the speaker.
Does one need to look at the ESR in the frequency gap of the notch with a good LCR bridge handtool? So very different result between an impedance notch filter at 300 hz or at 7 000 hz for illustration ?
In pratical terms I would not bother in cheking that difference with caps. Just get the new caps and listen. Variations may occur in cables, components, drivers and they are insignificant. Also you have them (differences) also in matching pairs, components and drivers. The listening experience will be here more important.
ESR differences between old electrolytic and new, or electrolytic and film can be quite high, on the order of 0.5 Ohms or more, and in some shunt circuits can lead to large swings in the circuit impedance, as well as reduce the effect of a notch filter.
Some homework I do myself tomorow with XSim
I assume one has to know the DC resistance of the coil as well. And the exact capacitance & esr of the lytucs when new... if only 20 years old a simple measurement should suffice
ESR being related to the frequencies if variations are that way important in an impedance notch filter, one should to meassure the former cap. The higher the frequency the bigger the ESR with some caps as lytics rough foil or MKT/mylars (the DF increase with frequency) much more. Some says till the 80s the difference is limited in high capacitance. But with 10 uF you have the smooth lytics too that have at the opposit of the two others a relative low ESR at high frequency .
That's why I asked if in such impedance correction the impedance bump at a given frequency matters ? Many MKT have a bader DF at high frequency than a god 10 uF lytic smooth foil according some manufacterers. Can we hear it I don't know, haven't tied but in a simplier Zobel network , with a 10 uF MKP which has a better constant ESR in the audio band vs the former lytic cap I had a smoothness gain... but the lytic was old and measured 9uF instead 10 uF ! Now AllenB told to me no so long ago, the lytics have a constant enough esr on the audio band to not bother too much if I remember correctly.
XSim free crossover designer : so easy as you can setup the esr of the coils & caps and see the result on the curve !
I assume one has to know the DC resistance of the coil as well. And the exact capacitance & esr of the lytucs when new... if only 20 years old a simple measurement should suffice
ESR being related to the frequencies if variations are that way important in an impedance notch filter, one should to meassure the former cap. The higher the frequency the bigger the ESR with some caps as lytics rough foil or MKT/mylars (the DF increase with frequency) much more. Some says till the 80s the difference is limited in high capacitance. But with 10 uF you have the smooth lytics too that have at the opposit of the two others a relative low ESR at high frequency .
That's why I asked if in such impedance correction the impedance bump at a given frequency matters ? Many MKT have a bader DF at high frequency than a god 10 uF lytic smooth foil according some manufacterers. Can we hear it I don't know, haven't tied but in a simplier Zobel network , with a 10 uF MKP which has a better constant ESR in the audio band vs the former lytic cap I had a smoothness gain... but the lytic was old and measured 9uF instead 10 uF ! Now AllenB told to me no so long ago, the lytics have a constant enough esr on the audio band to not bother too much if I remember correctly.
XSim free crossover designer : so easy as you can setup the esr of the coils & caps and see the result on the curve !
a 10 uF electrolytic may have an ESR on the order of 10s of ohms. The ESR of a film cap will likely be on the order of milliohms. I would expect great differences in the characteristics of resonant circuits (read notch filters) built with these two caps.