Polk csi40 center ch speaker crossover caps replacement

[budwich]

Did a "gross" impedance measurement of the inputs on my polk csi40 center channel speaker. It showed about 4ohms. Seemed kind of low. There are two midrange drivers and one twitter ... 2 way system.
All measurements at the drivers seem ok in the 4 - 6 ohm range (the two mid range are in series).

Anyway, in the crossover on the tweeter side, there are two caps in parallel basically leaving an inductor .... "nice simple setup", especially when no schematics around. The question is finding a replacement. One is a TI .47uf met 100vdc while other is a 9.1uf 50v electrolytic NP. I can find the .47 but not so much with a 9.1. I guess the question is what can the 9.1 be replaced with... a higher capacitance or lower... or can I just combine the values and find something with one capacitor that is close ... again higher or lower?

Thanks with any guidance on this, as for what I have read cross over design is almost a "black art"... :-)

[lti]

Re: Polk csi40 center ch speaker crossover caps replacement

Lower capacitance would be safer, but you might create an audible dip in the output volume near the crossover frequency. You can make your own parallel combination to make 9.57uF (or as close as you can get). The tweeter section of the crossover is basically a high-pass filter, but there's usually some other stuff added (at least an attenuator because tweeters tend to be louder than midrange/woofer drivers).

Is there a reason for changing the caps besides the low impedance you measured? Some speakers have some pretty low dips in the impedance.

[budwich]

Re: Polk csi40 center ch speaker crossover caps replacement

thanks for the guidance. One cap wouldn't measure (.47) although I might have damaged it trying to isolate it from the board to get a measurement. As for reason, I compared the impedance tests with the other speakers in the system (9) and it seemed significantly different from the other ones which seemed closer to stated values... off by an ohm or two. I know that a "dc impedance" as provided by a DMM isn't that great but figured that it should at least be closer as opposed to "half" the value. The speaker is pretty old so I figured caps are the most like source of a possible problem. Another thing is the speaker's sound seemed somewhat "harsh" at times (not the best hearing) as the highs seem distorted but not all time time.

As for the circuit, there is only a low value resistor (rx5w0r5j) in the high frequency side of things. No adjustable attentuator. The resistor measures very low... about .3 ohm but if understand the part number, its less than an ohm.

[budwich]

Re: Polk csi40 center ch speaker crossover caps replacement

I replaced the .47 but only got one that is 600v rated which makes it larger (fit on the board took a bit of lead bending). The speaker appears to sound better although it could be just a "psychological high" so to speak. :-) I am waiting on the other cap replacement as it was 20% off its marked value.