re-designing a speaker crossover
- Thread starter mantis
- Start date
TerryO
Super Member
Roger, I'm quite sure that Ken Kantor just read the OPs post incorrectly.
Or, what is even more probable, Ken's EVIL TWIN has taken on his identity.....Hmmmm
Best Regards,
TerryO
ken kantor
Fowl Humor
Indeed! That wasn't a typo on my part, it more like a "reado"... I thought the original poster had said he found an inductor and a cap on the tweeter.
-k
TerryO
Super Member
Ok, just to clear this up, are you yourself or are you actually your EVIL TWIN?
Best Regards,
TerryO
I have rebuilt many older (vintage) crossovers. The idea is to have any one driver do the work it's best at and let the other driver take over when things get beyond it. Take Winsloe Burhoes design. The EPI module allowed the woofer to roll off mechanically in the 1.8 to 2 K range. The tweeter started to come in around 1.6 and it was controlled by a single cap. There is no way to improve this design using these drivers, because the design uses the mechanical characteristics of the drivers. I think AR did that too?
You can improve the performance of the speakers with poly caps, better coils, hi-freq by-pass cap, etc. And, in the case of the AR by first dismantling the the tweeter pot and cleaning it, treating it to Fader Lube, etc. - then start on the upgrades. But wholesale replacement - to what end? To get a steeper slope on a two way system will leave you with a dip in frequency response at the crossover point?
You can improve the performance of the speakers with poly caps, better coils, hi-freq by-pass cap, etc. And, in the case of the AR by first dismantling the the tweeter pot and cleaning it, treating it to Fader Lube, etc. - then start on the upgrades. But wholesale replacement - to what end? To get a steeper slope on a two way system will leave you with a dip in frequency response at the crossover point?
the thread that would not die, apparently....lol
Face: thank you!
dnewma04: I am a humble noob here. I have much to learn about amps and electronics, read many threads, and am grateful to those generous enough to share knowledge. I will be doing some basic (to most of you) projects soon and will be here in DIY asking for your help.
My last post was, as you surmised, in response to one of ken kantor's assertions. But my purpose was merely to clarify for the benefit of subsequent readers. Nothing personal was intended (I did not "quote") and no disrespect was implied.
If or when I make any incorrect assumptions or assertions in any of my posts I hope to be corrected. For myself and for any subsequent AK readers. We are all here to learn.
BrocLuno: There would not be a dip in frequency response at the crossover point if a 2nd order (or 3rd, 4th order) crossover were designed properly.
However, there would be a dip in response if the existing 1st order crossover was "modified" to be 2nd order. If, for example, the coil currently in series with the woofer remained intact and a cap was inserted in parallel with the woofer it would cause the woofer's response to attenuate quicker and create a dip in overall response.
Another way to say this: You could have a properly designed 1st order crossover and a properly designed 2nd order crossover for any given pair of drivers. But the values of the series coil (woofer) and series cap (tweeter) would not be the same in the 1st order vs the 2nd order crossovers.
Face: thank you!
dnewma04: I am a humble noob here. I have much to learn about amps and electronics, read many threads, and am grateful to those generous enough to share knowledge. I will be doing some basic (to most of you) projects soon and will be here in DIY asking for your help.
My last post was, as you surmised, in response to one of ken kantor's assertions. But my purpose was merely to clarify for the benefit of subsequent readers. Nothing personal was intended (I did not "quote") and no disrespect was implied.
If or when I make any incorrect assumptions or assertions in any of my posts I hope to be corrected. For myself and for any subsequent AK readers. We are all here to learn.
BrocLuno: There would not be a dip in frequency response at the crossover point if a 2nd order (or 3rd, 4th order) crossover were designed properly.
However, there would be a dip in response if the existing 1st order crossover was "modified" to be 2nd order. If, for example, the coil currently in series with the woofer remained intact and a cap was inserted in parallel with the woofer it would cause the woofer's response to attenuate quicker and create a dip in overall response.
Another way to say this: You could have a properly designed 1st order crossover and a properly designed 2nd order crossover for any given pair of drivers. But the values of the series coil (woofer) and series cap (tweeter) would not be the same in the 1st order vs the 2nd order crossovers.
Last edited:
ken kantor
Fowl Humor
Up until midnight, its me. After midnight, it's my evil twin. Of course, your timezone may vary....
-k
I hesitate to add to anything Ken Kantor has written (but fools rush in, so here goes).
It seems to me that the original poster was under the impression that higher order crossovers were inherently better than first or second order crossovers. My understanding is that this is just wrong: There is the issue of efficiency, and then there is that of sound -- one poster did mention that 6db crossovers can pass a square wave whereas steeper slope ones can't -- I was under the impression that with steeper slopes you get ripple or even ringing, and that's why they have problems with square waves.
The advantage of steeper slopes is better protection for the tweeter and more supression of bad out of band behavior of the drivers (though unless the slope is infinite, that bad behavior may be audible anyway).
The art of speaker design comes not in adding complexity and cost (any bonehead can do that), but in simplifying, while still achieving the desired sound. That's where someone like Burhoe (who, if I recall correctly, was involved in the design of the AR 4 or 4x) shows his worth -- a simple cap, and sometimes a voice coil used as an inductor with fine results because the drivers were well designed to work synergistically with simple crossovers.
I think this case may be one where if it ain't broke, don't fix it. If the OP is determined to fix it until it is broke (and there's a lot of fun and learning to be had doing so), do one speaker at a time, so it can be compared to the original, and hang on to the original crossover just in case we are all right in our naysaying.
It seems to me that the original poster was under the impression that higher order crossovers were inherently better than first or second order crossovers. My understanding is that this is just wrong: There is the issue of efficiency, and then there is that of sound -- one poster did mention that 6db crossovers can pass a square wave whereas steeper slope ones can't -- I was under the impression that with steeper slopes you get ripple or even ringing, and that's why they have problems with square waves.
The advantage of steeper slopes is better protection for the tweeter and more supression of bad out of band behavior of the drivers (though unless the slope is infinite, that bad behavior may be audible anyway).
The art of speaker design comes not in adding complexity and cost (any bonehead can do that), but in simplifying, while still achieving the desired sound. That's where someone like Burhoe (who, if I recall correctly, was involved in the design of the AR 4 or 4x) shows his worth -- a simple cap, and sometimes a voice coil used as an inductor with fine results because the drivers were well designed to work synergistically with simple crossovers.
I think this case may be one where if it ain't broke, don't fix it. If the OP is determined to fix it until it is broke (and there's a lot of fun and learning to be had doing so), do one speaker at a time, so it can be compared to the original, and hang on to the original crossover just in case we are all right in our naysaying.
dc270
Now Off the Reservation
Design tools, design tools, design tools, that is the thing I learned. Once I finally bathed myself in design software then it became much easier to understand and design boxes & xovers to fit given drivers.
I bought several Eminence programs from Part Express. I really like their xover software as it worked well with other programs and seemed pretty complete with many filter designs from Butterworth to Linkwitz/Rilley.
Arm your self with information, get a foot hold of the fundamentals. Once you have a good working knowledge then you can build anything that might work.
As was pointed out crossovers can be the wild card variable in speaker design. You should understand losses, storage, complexity penalties, and a miread of other annomolies associated with crossovers. Trying to design under the guide of guess work is a leader sure to lose his way.
DC
I bought several Eminence programs from Part Express. I really like their xover software as it worked well with other programs and seemed pretty complete with many filter designs from Butterworth to Linkwitz/Rilley.
Arm your self with information, get a foot hold of the fundamentals. Once you have a good working knowledge then you can build anything that might work.
As was pointed out crossovers can be the wild card variable in speaker design. You should understand losses, storage, complexity penalties, and a miread of other annomolies associated with crossovers. Trying to design under the guide of guess work is a leader sure to lose his way.
DC
ken kantor
Fowl Humor
Hi... attached is some background on crossover theory. I'll try and dig up some info on practical procedures, too.
-k
-k
thank you for that post