S4's crossover revamp

jujubean

Ok with most speaker manufacturer's the major drawback of most lower end designs is in the crossover. Looking at the schematics for the S4 this is where polk cut costs on these as well.

I want to do a 12db up and down at 3k with these. It looks like they crossed them at 3k and they use 4ohm drivers. Or crossed at 6k and used 8 ohm.

I want to do 3k to stop the breakup on the mid. I can definitely tell it is left to roll off naturally and it annoys the crap out of me.

When i crossover actively and listen to the mid by itself under 3k it sounds wonderful. So I just wanted to get some opinions on this. Since the drivers are dynamic balance, and klippel tested, after having a good set of crossovers on them they should sound like a whole new animal.

has anyone else done anything like this before? Were you satisfied afterwords?

It should cost around 150 bucks to get the supplies to do all 4 of my S4's. Let me know if you think it is worth it! I'll probably end up doing my cs100 as while i am at it. Although with the smaller woofers in it, it has less distortion in the upper midrange, but i might as well while i am doing the S4's.

Face

Not a good idea if you don't know what you're doing. FYI, online crossover calculators and generic crossovers are junk.

jujubean

I've built many crossovers before lol

I was just wondering really if you think the speakers would benefit a lot from it.

I will also be custom winding my own inductors. 10 18ga inductors for 20 bucks sounds like a great deal to me !!! This is what i usually do for inductors. Plus i get to pick my color or enamel coating lol

RuSsMaN

Other than throwing a Sonicap in place of stock (or whatever flavor you prefer), the crossovers are the least of your worries in the S series, the cabinets are the big 'sound killer' in that equation. The drivers are of decent quality, but the cabinets are thin and far from inert.

Just my .02, if you feel like hotrodding them and you aren't worried about cost, start with new cabinets, then look at the crossover.

Cheers,
Russ

jujubean

That's a great idea. If i could get my hands on the T/S params of the drivers i would remake the cabinets in a heartbeat. Anyone have a tester? I could send you the drivers and the money for return shipping if you do.

Or if there are published specs on them, that would be great.

zingo

Ambitious!

ShinAce

Since the woofers have been played for quite some time, they are ready to be tested for small signal specs, aka, Thiele/Small parameters.

See
http://sound.westhost.com/tsp.htm

Make note this method is only measuring the resistance of the speaker, not the impedance. It does not account for capacitive versus inductive loads.

You can get all the specs from a plot of the impedance. Then you need to load the driver to get compliance(Vas). You can add weight to the cone to add mass, or you can install the speaker in a well sealed box and measure the new resonance frequency. Make note that added weight does not need to 'stick' to the cone. The speaker is not supposed to move when you're doing this. Very very little power is to be applied to the speaker.

It seems pointless, since they're already in a box. Why not just build a heavy duty version of the box?

Also, I say proceed with the crossover mod. A lot of crossover mods here on CP consist of capacitor upgrades, but you plan on redesigning it to minimize cone breakup modes and dispersion issues. Good on you! Remember that adding a woofer section to the crossover basically invalidates the design of the tweeter section.

Face

jujubean wrote: »

I've built many crossovers before lol

I've built many too, but there's a large difference between building one and designing one.

jujubean wrote: »

I will also be custom winding my own inductors. 10 18ga inductors for 20 bucks sounds like a great deal to me !!!

How long would that take?

jujubean wrote: »

That's a great idea. If i could get my hands on the T/S params of the drivers i would remake the cabinets in a heartbeat. Anyone have a tester? I could send you the drivers and the money for return shipping if you do.

Using the T&S parameters along will only get you halfway there. You would also need the raw driver's measured response in the cabinet you're using to get something accurate.

If you were to build new cabinets, the cabinet volume, front panel dimensions(width is the most important), and driver placement must all be the same unless you have the proper measuring equipment and software to re-engineer the crossover.

I agree with the posters above and replacing the capacitors in the HF circuit with something better and building better cabinets would be your best bet.

skrol

Flame suit on. ... I say go for it. If you have the time and money and desire, what do you have to loose. You will certainly learn from the experience no matter if it turns out good or not. Worst case you can fall back to the stock XO and cabinet. In any event it sounds like some good fun.
Stan

jujubean

I have also "Designed" many crossovers with zobels and notch filters and whatnot, It's not as serious as you are making it out to be if you have general electrical knowledge lol.

But anyways I think i will go through with the cabinets and the crossovers. I'll post a build log once I get all the supplies together.

ShinAce

jujubean wrote: »

I have also "Designed" many crossovers with zobels and notch filters and whatnot, It's not as serious as you are making it out to be if you have general electrical knowledge lol.

But anyways I think i will go through with the cabinets and the crossovers. I'll post a build log once I get all the supplies together.

I want to make sure we're on the same page:
-Zobel = easy
-noth filter = intermediate
-low pass x-over = easy
-high pass x-over = easy
-good response in the crossover region = trial and error

You can use the EE method(what you're doing) by using a crossover frequency away from eash drivers' resonance, then correcting impedance with Zobels.

The responses you're getting are basically "Do it right the first time". In other words, build it, then tweak it. You won't find cheap variable caps. So you need to figure out if you buy a few values and play around. I like to do this. Get some cheap electrolytics, design, build, then tweak by measuring the whole speaker with a mic. Once you get the curves you want, then buy good quality components. This works well with the caps, but not so well with the inductors. Testing with a thin gauge air core and then finally using a heavy gauge steel core is not recommended. Use the type(air, iron) and gauge that will be finally installed. This won't be a problem for you, since you're comfy winding your own.

Again, I say go for it. But maximize your investment by not settling with what looks good on paper and sounds 'fine'.

skrol

Amen. There certainly is more to it than designing filters. The drivers themselves have reactive impedance characteristics which play with and against the crossover. It can be done however. There are some good designers of home brew speaker systems out there. Perhaps you are one of them.

What you are proposing is something that I wish I had the time and experience to take on. I had designed and built a few speakers back in my teens / early 20's that were nothing to write home about but it was fun. I know a bit more now and would like to do it again. Maybe I'll start with a decent kit speaker.

Face

http://sites.google.com/site/undefinition/simulated-measurements