It is my understanding (but I could be wrong on this) that baffle step correction corrects for the woofer radiating omni-directional whereas the tweeter does radiate half space due to the wave lengths being short with regards to the baffle width which leads to different effective efficiencies from the factory-given ones.
Edge diffraction is a different phenomenon ie a sharp edge results in a ghost tweeter that appears at the edge a few dB down from the actual one.
Two different problems with to different solutions. The first leads to the speakers sounding bass-light when free standing, the second to smeared imaging.
That said there is one solution which fixes both issues: Soffit mounting the speaker ie fit it flush into the wall. But that is not really feasible for most of us and so you fix the first by adding bsc to the xover and the second by rounding over the edges.
If you place the speakers against a wall I don't think bsc is necessary but you will get some minor comb filtering.
(If I got this completely wrong I hope somebody will correct me!)
Edge diffraction is a different phenomenon ie a sharp edge results in a ghost tweeter that appears at the edge a few dB down from the actual one.
Two different problems with to different solutions. The first leads to the speakers sounding bass-light when free standing, the second to smeared imaging.
That said there is one solution which fixes both issues: Soffit mounting the speaker ie fit it flush into the wall. But that is not really feasible for most of us and so you fix the first by adding bsc to the xover and the second by rounding over the edges.
If you place the speakers against a wall I don't think bsc is necessary but you will get some minor comb filtering.
(If I got this completely wrong I hope somebody will correct me!)
thanks for the reply. There are many variables that leads to many differences in FR, and your comments are precious.
My concerns regards the possibility of mantain the calculated FR introducing the roundover solution (yes, I don't want the "ghost tweeter").
So, Zaph has developed a crossover with a standard cheap cab and the resulting, smooth FR is probably calculated on this baffle. Rounding the edges, I fear that the response at cross frequency will be a little modified. I don't know precisely how, and how much!
BSC is a minor concern: probably a couple centimeters will be strongly audible, and I could sperimenting with different positioning or different woofer's coil values. Ok, this is more manageable for me. The real challenge is to mantain a good integration between tweeter and woofer, and maybe a strong roundover of the edges can modify it consistently. What do you think about?
My concerns regards the possibility of mantain the calculated FR introducing the roundover solution (yes, I don't want the "ghost tweeter").
So, Zaph has developed a crossover with a standard cheap cab and the resulting, smooth FR is probably calculated on this baffle. Rounding the edges, I fear that the response at cross frequency will be a little modified. I don't know precisely how, and how much!
BSC is a minor concern: probably a couple centimeters will be strongly audible, and I could sperimenting with different positioning or different woofer's coil values. Ok, this is more manageable for me. The real challenge is to mantain a good integration between tweeter and woofer, and maybe a strong roundover of the edges can modify it consistently. What do you think about?
I wish I could remember where I read it to provide you a link, but there was a write up on the web from a speaker builder who when he first started building speakers went to go to great pains to round all the edges of his enclosures over in order to supposedly minimize diffraction, and one day he got to thinking that diffraction due to square edges might not be the problem that everyone thinks it is, so he built two identical sets of speakers except one set had rounded edges and one set did not, and then he proceeded to conduct listening tests with the two pairs of speakers with a wide variety of listeners, and it turns out everyone thought the rounded/non-rounded speakers sounded identical, and after that he didn't bother taking the time to round the edges of his speakers over anymore.
How much difference round overs make depends on the directivity of the tweeter.
For a 90deg horn for instance it shouldn't make any or very, very little audible difference.
On the other hand it should be audible (more precise imaging, less comb filtering) for small domes which stand proud of their face plate.
Guess most tweeters would fall somewhere in between.
Personally I like round overs also (or even mostly) for aesthetic reasons. It makes the box look less monolithic.
I don't really see how it would make a great deal of difference to the crossover or FR. The total energy radiated remains the same, it's just that diffractions will be more diffuse which is a good thing IMO.
For a 90deg horn for instance it shouldn't make any or very, very little audible difference.
On the other hand it should be audible (more precise imaging, less comb filtering) for small domes which stand proud of their face plate.
Guess most tweeters would fall somewhere in between.
Personally I like round overs also (or even mostly) for aesthetic reasons. It makes the box look less monolithic.
I don't really see how it would make a great deal of difference to the crossover or FR. The total energy radiated remains the same, it's just that diffractions will be more diffuse which is a good thing IMO.
Doh! My beliefs start rattling.
My tweeter should be the scan speak 6600 aircirc. I suspect little directivity..
I'm not sure to understand: Also you all say that an avalon-like, faceted-edged baffle could sound approssimatively identical to a totally flat one (given a adequate total width)? Or that the measured FR of the tweeter don't vary with different edge's shape?
I have read anywhere that baffle dimensions MUST be mantained for a given crossover, (not only for BSC, expecially diffraction-related issues)... it seems intuitive that a larger totally flat surface plus sharp edge should be different to a smaller totally flat surface plus extended smooth&round edges. Maybe is not so, but the measured tweeter response should be different. If so, the FR of the whole speaker should be altered, maybe a little..maybe audibly.
Am I totally wrong?
My tweeter should be the scan speak 6600 aircirc. I suspect little directivity..
I'm not sure to understand: Also you all say that an avalon-like, faceted-edged baffle could sound approssimatively identical to a totally flat one (given a adequate total width)? Or that the measured FR of the tweeter don't vary with different edge's shape?
I have read anywhere that baffle dimensions MUST be mantained for a given crossover, (not only for BSC, expecially diffraction-related issues)... it seems intuitive that a larger totally flat surface plus sharp edge should be different to a smaller totally flat surface plus extended smooth&round edges. Maybe is not so, but the measured tweeter response should be different. If so, the FR of the whole speaker should be altered, maybe a little..maybe audibly.
Am I totally wrong?
Last edited:
Sorry, didn't see this earlier. I strongly suggest not using ferrite; smaller ones overload very easily, make a distinctive "crack" sound when they saturate. My last project needed a low R woofer series inductor, and the Erse SuperQ inductor worked out very well. Usually going a bit lower on Rdc is OK; there is always some variation from speaker cables, amplifier output impedance, etc. anyway, so the speaker cannot be too sensitive to small resistance changes.
For best diffraction control, a 4" (100mm) radius or larger is best. There was an excellent article, with measurements some years ago in Speaker Builder (back issues available from Old Colony.)
A wool felt covering on the front panel will help with diffraction. The felt should be high wool content, and relatively soft. F11 or F13 in our standard; 70-80% wool, inexpensive. It should be about 1/4" (6mm) thick, and the edges around the drivers should be tapered to prevent a step around the driver, especially the tweeter.
Without opening that ol' can of worms very far, some people do find that better crossover components, with good drivers and associated equipment, do sound better. Try for your self, trust your own ears. A lot of people respect Zaph, for a reason.
For best diffraction control, a 4" (100mm) radius or larger is best. There was an excellent article, with measurements some years ago in Speaker Builder (back issues available from Old Colony.)
A wool felt covering on the front panel will help with diffraction. The felt should be high wool content, and relatively soft. F11 or F13 in our standard; 70-80% wool, inexpensive. It should be about 1/4" (6mm) thick, and the edges around the drivers should be tapered to prevent a step around the driver, especially the tweeter.
Without opening that ol' can of worms very far, some people do find that better crossover components, with good drivers and associated equipment, do sound better. Try for your self, trust your own ears. A lot of people respect Zaph, for a reason.
Last edited:
I have all the basic stuff. (I'm waiting for a pair resistances, but they are on the way).
I'll begin with a cheap ferrite coil, and try the right amount of BSC. Then I'll break my porcelain pig and go for mundorf air cores.
I don't like felted baffle.. I'll go for a reasonable roundover (the woofer must fit! )
Build begins!
I'll begin with a cheap ferrite coil, and try the right amount of BSC. Then I'll break my porcelain pig and go for mundorf air cores.
I don't like felted baffle.. I'll go for a reasonable roundover (the woofer must fit!
)Build begins!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.