[skite30]

I have a pair of 2-way speakers I am replacing the caps in. It is a 3rd order crossover. The manufactures manual says the crossover is supposed to be 2.800. This would call for a 4.7uf cap. I measured the caps that are in there and they measure 4.43 and 4.56 , this would make the crossover closer to 3.000. Will this make a significant difference or is it within tolerances for the caps.

[mgalusha]

If the existing caps are electrolytic, then that is a remarkably good match. If they are film it's still close to 5%, which is not bad for unmatched caps. Some vendors will match them for you for a small fee. Having them the same is probably more important than the exact value but if you're going to pay to have them matched they can likely get darn close to the specified value. Some of the Sonicaps I've received from GR Research came with the measured values written on the caps and I didn't ask for them, nice touch.

[skite30]

The original caps are electrolytic. I am replacing them with some obbligato's that tested at 4.71 and 4.68. I didn't know if the larger value would make much of a difference or if I should try to match the original values closer.

[mgalusha]

I imagine the value of the Obbligato's will be just fine, certainly about as close as you're likely to find in regards to the design values, at least without buying/sorting a bunch. Personally I've had very good results with the Obbligato caps in speakers.

[skite30]

I like what I hear from the obbligat's a lot so far. I would like to try some russian pio's if I can work out the right values.

[andyr]

I am replacing the original caps with some obbligatos that tested at 4.71 and 4.68. I didn't know if the larger value would make much of a difference or if I should try to match the original values closer.

Dear God - the original caps were nominally 4.7uF and your replacements are 4.71uF & 4.68uF! 

The most difference (4.68uF) is 0.02uF away from theoretical.  That's 0.4% away!  Spectacular matching! 

Of course, that's only if your meter is actually reading correctly! 

Regards,

Andy

[skite30]

I guess I haven't done a very good job of asking this question. The original caps measure right at 4.4. This puts the crossover at about 3,000. The owners manual says the crossover point for these speakers should be 2.8 which would be a cap value of 4.7. I installed caps with a value of 4.7 or very close to it. I was asking if I should try to more closely match the 4.4 values that were in there or if the difference between the 4.7 and 4.4 was insignificant and within normal tolerances. I agree the 4.7 obbligato's are very well matched.

[Kevin Haskins]

It is hard to know without the actual model of the network and the impedance of the actual transducer.   In some spots... you can have gross errors in crossover values and little impact on the response.  In other networks... the values are very critical.   This is the case in notch filters or any filter that has a very high Q value.   If it is acting over a narrow band the values tend to be more critical.   On something like a 1st order electrical... you can have very large differences with almost no difference in the final measured response.   

I'd match the design value as within 3% and call it good.   Most MKP caps even the 5% variety come out < +/-3% and some tighter than that.   Get the 4.7uF part and call it good.

[skite30]

The driver is 8 ohm, the crossover circuit is 3rd order. It has 3 6uf caps going to the positive post, 2 are wired together in parallel and the other is wired across them. I hope I described that correctly. It also has 2-6uf caps that come before the inductor which I am not planning on changing. Thanks I believe this answers my question.

[face]

Where are you getting these calculations from?

Replace your original caps with 4.4uf's.  Don't replace them with larger 4.7's because you will lower the crossover frequency and possibly introduce distortion, phase problems, a lump in response, etc..

[Speedskater]

You are making too many assumptions:
a) The designer may have voiced the speakers and then drawn the schematic.
b) Then it may be that all the capacitors in their stock have different actual value than printed value.
c) The driver very well may not be 8.0 Ohms at crossover frequency.
d) There is more than one way to design a cross-over.
e) Skilled designers consider factors that don't show up on the schematic. (like imperfect capacitors and wire resistance in inductors)

[markC]

Or the electrolytics are getting weak and their value has decreased.

[skite30]

I am trying to get the crossover values back to where they originally were. I bought the speakers used from a man who sent the crossovers off to be upgraded. He included the original crossovers for the tweeters but not the woofers. The speakers sounded like crap when I got them, sibilant, congested and distorted. I pulled the crossovers out and started measuring values. The tweeter cap measured 6.2 instead of 4.4. My question is should I replace the cap with a 4.7 to get the value the schematic calls for or the 4.4 measured value of the original caps in the circuit. I am using the original inductors.

[face]

Is the schematic from the manufacturer or are you using an online calculator?  If the schematic is from the manufacturer, I would use those values and replace them with decent film caps.  Don't use online calculators, they're far from accurate most of the time.

[skite30]

I am using the owners manual for the crossover point. the 4.7 cap value was derived by a speaker tech who works for a local high end audio shop. I am using the actual measured values of 4.4 - 4.5  from the original crossover. I guess when I look at it from that angle I should go with the 4.4  measured value and be done with it. The speakers are about 10 years old.

[markC]

If the cap values are being evaluated by the crossover point alone, I would certainly go with the measured value.