Kenwood LSK-737 crossover

[abhi1981]

Hello everybody.
Recently got a pair of Kenwood LSK-737's . The original crossover is missing and has been replaced with a mixture of caps and inductors , connected in a way I don't understand  . Neither series nor parallel.
Please help me with the right crossover circuit. I was not able to find the exact circuit diagram, but found one for LSK-703. It should be similar to LSK-737.
Also, bipolar caps are not available here in India . Can I use film caps like shown in the picture?

[Gyro]

It looks to me as if the manufacturer has improved the crossover network on the higher model.

The LSK-703, as shown in the schematic, is hardly a crossover at all - just a couple of series capacitors (and level setting resistors) to roll off the upper and midrange drivers to protect them against low frequency energy. There is no high frequency filtering on the bass driver, other than its voice coil inductance and cone mass. Likewise on the midrange driver, that receives high frequency energy too.

The 737 crossover has a combination of L-C filter networks to provide steeper and more specific filtering between the drivers. Whether this sounds better than the original or not, I can't say, but it has probably improved efficiency and maybe power handling.

Such crossover construction (hand assembled, with components roughly glued to a piece of chipboard) is fairly common in a low-end speakers, I wouldn't immediately assume that it wasn't a later implementation by the manufacturer.

I don't know if you are wanting to replace the capacitors in the crossovers that you have, but it will be hard to make up values like 18uF using ordinary film capacitor.


P.S. It looks as if the midrange drivers have been replaced with a different size at some point.

[abhi1981]

Yes, I want to change the capacitors a they are showing higher readings on a capacitor tester. The midrange is also not giving clear output.
The midrange driver is original, there's a garnish ring around it which I have removed.
I have attached a rough circuit diagram (sorry about my bad drawing  ) of the crossover, which I strongly believe is not original because the person who sold it turned out to be a shady character. He must have switched the original with this locally made crossover. Believe me, they do such things here
Can film capacitors be connected in parallel to double the values??

[Gyro]

The crossover, as you have drawn it, makes sense. C-L filter high pass for the tweeters, Capacitive high pass on the mid, and L-C low pass on the woofer. Certainly more like what I would expect to see in a speaker crossover than the very minimal 703 one. The shadiest thing I see is the replaced midrange drivers, which are probably a repair.

Yes, you can parallel a number of film capacitors if you want. In parallel connection, their values simply add.


EDIT: Sorry, I didn't notice your comment on the midrange 'garnish ring'. That is good. If you are replacing capacitors, I would probably concentrate on the midrange one first.

[mansaxel]

Yes, I want to change the capacitors a they are showing higher readings on a capacitor tester. The midrange is also not giving clear output.
The midrange driver is original, there's a garnish ring around it which I have removed.
I have attached a rough circuit diagram (sorry about my bad drawing  ) of the crossover, which I strongly believe is not original because the person who sold it turned out to be a shady character. He must have switched the original with this locally made crossover. Believe me, they do such things here
Can film capacitors be connected in parallel to double the values??
(Attachment Link)

The woofer circuit is -- providing the component values are sane -- a 12dB/octave low-pass filter.

The midrange still is a 6dB/octave high-pass, and not a band-pass. As stated in the thread earlier, this trusts the natural roll-off of the midrange to keep it from interfering with the treble drivers.

The treble also is 12dB/octave, but high-pass.

All in all, there is a plan to the circuit. Is it the best circuit for the application? Perhaps not. I'd think twice about the lack of low-pass for midrange, OTOH, the first circuit from the manual also lacks this.

A lot of the character and clarity of a multi-element loudspeaker system lies in the interaction (or lack thereof) between drivers, and this is largely governed by whether they try to reproduce the same frequencies.  To achieve this most crossovers actually have some gap in the response curve so that there is a 3 dB drop in level at the crossover frequency when the outputs are summed. It does look non-intuitive on paper, but it works in reality.  A high-pass-only midrange meeting a high-pass treble circuit will not satisfy this -- there will need to be a low-pass filter on the midrange too. But, this was not deemed a problem in the original circuit, so..

The proper minimal-compromise solution of course is an active crossover and separate power amps per speaker, but that's seriously expensive.

Oh, and on capacitors, have a look here: https://courses.lumenlearning.com/physics/chapter/19-6-capacitors-in-series-and-parallel/

[abhi1981]

Thanks for the valuable comments.
How can I improve the crossover circuit? Please prvide a circuit diagram, if possible.
Some 3 way crossovers are available here, but I don't know about their efficiency.
https://www.amazon.in/Electronicspices-Speaker-Frequency-Crossover-Multicolour/dp/B07RJT2TGL/ref=sr_1_3?dchild=1&keywords=audio+crossover&qid=1633069957&sr=8-3
https://www.amazon.in/NEXT-GEEK-Crossover-Frequency-Distributor/dp/B092FX655L/ref=sr_1_5?dchild=1&keywords=audio+crossover&qid=1633069957&sr=8-5

[mansaxel]

Thanks for the valuable comments.
How can I improve the crossover circuit? Please prvide a circuit diagram, if possible.
Some 3 way crossovers are available here, but I don't know about their efficiency.
https://www.amazon.in/Electronicspices-Speaker-Frequency-Crossover-Multicolour/dp/B07RJT2TGL/ref=sr_1_3?dchild=1&keywords=audio+crossover&qid=1633069957&sr=8-3
https://www.amazon.in/NEXT-GEEK-Crossover-Frequency-Distributor/dp/B092FX655L/ref=sr_1_5?dchild=1&keywords=audio+crossover&qid=1633069957&sr=8-5

I think that question suffers from not enough information. Since the inductor values are unknown, we can't tell for sure what the crossover frequencies are. Some help can be had here where you can look up values for various crossovers. Try finding your capacitors there, and assuming that the inductor was chosen well, you can then derive its value. From that, you can then use other pages on that site to help calculate values for other components.

Keep in mind that the speaker impedances are parts of the filter designs (they're called RC-filters for a reason!) and therefore you must know the nominal impedance. DC resistance of the winding often gives some guidance, but is not precise. Most of the time the DC Ω value will be lower (by 20-30%, roughly) than the AC impedance, so a 4Ω woofer will have a DC resistance of perhaps 3,1Ω or so.

Good luck!

[abhi1981]

Replaced all the caps with polyester ones. Its working quite well.

[Gyro]

Thanks for reporting the outcome.

You may want to use some hot melt glue or similar, to prevent rattles, if you are satisfied with it.

[abhi1981]

Actually I did use hot glue, just didnt take pictures.