Just whipped up a quick layout of my new speakers im going to build, the woofers were originally in a 80 litre box with 2 of em, so im giving them about 40litres. Got some d19's lying around so i thought id use them.
I just need some help on how to make a crossover for them, a simple one if possible.
Heres the pic
Thanks !
I just need some help on how to make a crossover for them, a simple one if possible.
Heres the pic
Thanks !
Let me start with an assumption: you're trying to take the trouble to make something really good. If that assumption is wrong, skip everything I have to say. I assume this because the cabinet design looks like something pretty ambitious.
You'll never get that driver combination to work really well. The problem is matching of dispersion characteristics of the two drivers at realistic crossover frequencies. The 8" will start having its dispersion narrowing at a frequency well below where you can reasonably cross over to the tweeter (unless you like distortion and fried tweeters). What will happen with any crossover that protects the tweeter is that the off axis frequecy response will show a big midrange dip up to the point where the tweeter cuts in, where the response will rise. So even if the direct sound is flat, the reverberant response won't be at all smooth. This has been the plague of all the 8" two-ways that I've ever listened to. Imaging and spectral balance suffer badly.
Consider using a smaller driver for the woofer (a 5" would match your tweeter well) or a tweeter that will go down much lower in frequency than a D19. By all means, use a high-slope crossover so that the woofer's break-up frequncy is well into the stop-band.
You'll never get that driver combination to work really well. The problem is matching of dispersion characteristics of the two drivers at realistic crossover frequencies. The 8" will start having its dispersion narrowing at a frequency well below where you can reasonably cross over to the tweeter (unless you like distortion and fried tweeters). What will happen with any crossover that protects the tweeter is that the off axis frequecy response will show a big midrange dip up to the point where the tweeter cuts in, where the response will rise. So even if the direct sound is flat, the reverberant response won't be at all smooth. This has been the plague of all the 8" two-ways that I've ever listened to. Imaging and spectral balance suffer badly.
Consider using a smaller driver for the woofer (a 5" would match your tweeter well) or a tweeter that will go down much lower in frequency than a D19. By all means, use a high-slope crossover so that the woofer's break-up frequncy is well into the stop-band.
Ok i get what your saying 
Could you recommend a smaller woofer that would suit the tweeter i have, that also wont cost me an arm and a leg ??
Also for these speakers i was hoping to get a descent midrange and highrange out of them. They dont need to be heavy in bass since ill be using a small subwoofer with them. Could i use a midrange or midwoofer instead of a woofer ??
Im pretty new to speaker designing so excuse me if i say some wrong stuff
Could you recommend a smaller woofer that would suit the tweeter i have, that also wont cost me an arm and a leg ??
Also for these speakers i was hoping to get a descent midrange and highrange out of them. They dont need to be heavy in bass since ill be using a small subwoofer with them. Could i use a midrange or midwoofer instead of a woofer ??
Im pretty new to speaker designing so excuse me if i say some wrong stuff
I don't know what's available in Oz, but I'd look for 130mm or 160-170mm drivers from Vifa, Audax, Seas, or Morel. You might browse around the Madisound website for some ideas about models and pricing.
I'm just the opposite of you- woodworking is something I can't really do well and I'm not equipped for. Too bad you don't live in California!
I'm just the opposite of you- woodworking is something I can't really do well and I'm not equipped for. Too bad you don't live in California!
Adding a midrange would be ok but it does add extra costs, and then it would involve making a more detailed crossover, i was hoping to just get a simple project going 
Im really new to crossovers and stuff, was hoping maybe just to use a cap or whatever.
Could someone just give me a simple way of linking the drivers together that i originally specified by using a cap, no crossover etc.
thanks in advance
Im really new to crossovers and stuff, was hoping maybe just to use a cap or whatever.
Could someone just give me a simple way of linking the drivers together that i originally specified by using a cap, no crossover etc.
thanks in advance
Using a capacitor is not a bad start; after that, you can move up to using a cap and inductor. Lots of "professional" speakers survive using just a cap, or just a cap and an inductor.
The problem is that an 8" and a 1" have a hard time working together most of the time. Cone breakup makes the frequency response very uneven past the point of cone breakup. If you want to send these frequencies to the tweeter, you have to have a tweeter that won't completely die if you do so. With speakers as large as 8", this usually means a 3-way system, though not always. My friend Merrilee owns a studio, and her monitors are 2-way with an 8" paper cone / foam surround and a 1" titanium tweeter, and they seem to work pretty well for her.
Also, if the 8" has a larger voice coil, this helps defeat cone breakup. That's why you don't find 12"+ speakers with 1" voice coils (except in guitar amps); the cone breakup gets to be too much.
So why not make the voice coil speaker size and have virtually no breakup? Low efficiency. So, like everything in the audiophile world, it's a trade between two desireable traits.
Anyway. Tangent over. Measure the frequency response of the two drivers, and pick a crossover point that doesn't cripple either driver.
The problem is that an 8" and a 1" have a hard time working together most of the time. Cone breakup makes the frequency response very uneven past the point of cone breakup. If you want to send these frequencies to the tweeter, you have to have a tweeter that won't completely die if you do so. With speakers as large as 8", this usually means a 3-way system, though not always. My friend Merrilee owns a studio, and her monitors are 2-way with an 8" paper cone / foam surround and a 1" titanium tweeter, and they seem to work pretty well for her.
Also, if the 8" has a larger voice coil, this helps defeat cone breakup. That's why you don't find 12"+ speakers with 1" voice coils (except in guitar amps); the cone breakup gets to be too much.
So why not make the voice coil speaker size and have virtually no breakup? Low efficiency. So, like everything in the audiophile world, it's a trade between two desireable traits.
Anyway. Tangent over. Measure the frequency response of the two drivers, and pick a crossover point that doesn't cripple either driver.
SY said:
What SY says is certainly true on the face of it... but there are quite a few counterexamples of 8" 2-ways with XOs in the same region that are quite good, and some that are 1st rate, so it is not an impossible situation. (althou they usually XO to a larger T.
What kind of woofer? What kind of cone material? (there are tricks you can do with a paper cone to improve its hi-frequency dispersion).
dave
Nappylady said:
Yep... it is common in full-ranges & extended range woofers to have a controlled decoupling of more and more of the outer part of the cone as the frequency increases.
To do this with something like puzzlecoat you create a set of ever-more heavily coated annular rings on the cone (ie thickest ring at the centre). To do this you define the rings. Then coat 1 only goes on the inner ring, a 2nd cost also covers the 2nd ring, so on until the entire cone has at least one coat.
dave
The cone is made of paper, surround is made of foam, and the dustcap is like some hard cloth.
Theyre out of some old radioshack speakers, i think i heard somewhere coral used to make alot of the drivers for radioshack speakers. Coral is vifa/scanspeak now right ?
I think id like to go with just using a cap for this project to just get me started. How would i go about choosing the right cap for the drivers ?
Also im interested about how you said to improve the paper speakers dispersion.
Theyre out of some old radioshack speakers, i think i heard somewhere coral used to make alot of the drivers for radioshack speakers. Coral is vifa/scanspeak now right ?
I think id like to go with just using a cap for this project to just get me started. How would i go about choosing the right cap for the drivers ?
Also im interested about how you said to improve the paper speakers dispersion.
The cap is chosen by value according to what crossover frequency you desire.
Measure the frequency response of the drivers. Only this will tell you where the best place to crossover is.
Even measuring by ear isn't so bad, with a trained ear, but microphones are cheap (you can get a mic element from Radio Shack for <$5) and give better-looking graphs.
I don't know the equation for which value capacitor you want, but I know that as the frequency you want increases, the capacitance must decrease.
Maybe what you could do is get the assorted caps from Radio Shack and play with different values to see what sound you like best.
Let us know, in any case, what you do, and if you get frequency reponse graphs, post 'em--I'm curious now.
Measure the frequency response of the drivers. Only this will tell you where the best place to crossover is.
Even measuring by ear isn't so bad, with a trained ear, but microphones are cheap (you can get a mic element from Radio Shack for <$5) and give better-looking graphs.
I don't know the equation for which value capacitor you want, but I know that as the frequency you want increases, the capacitance must decrease.
Maybe what you could do is get the assorted caps from Radio Shack and play with different values to see what sound you like best.
Let us know, in any case, what you do, and if you get frequency reponse graphs, post 'em--I'm curious now.
postman said:Planet10 would i have to do any measure ments before trying to coat the speaker with puzzlecoat?
No measures needed, the process is completly impirical. A rigourous experiment could probably analytically measure the results, but would require a quantity of the same driver and a series of treatments backed up by listening.
dave
postman said:
Puzzlecoat is actually a name from a 25 year old product, that like kleenex or Xerox, has become generic in my mind. Follow the link for modern sources. In a pinch, white wood glue will also work.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.