1st order crossovers are much better with time aligned drivers so I'm looking for a solution so I can still have a flat and not a stepped or sloping (10-15°) baffle.
I saw this idea for the tweeter (see pic) but I'm not sure whether you'd get some horn loading and also diffraction problems with this. I'd also have to have the woofer protruding from the baffle by about 16mm (a round spacer the same diameter as the woofer). Is this a problem as currently it is flush?
Or is this going to cause more problems than it cures?
I saw this idea for the tweeter (see pic) but I'm not sure whether you'd get some horn loading and also diffraction problems with this. I'd also have to have the woofer protruding from the baffle by about 16mm (a round spacer the same diameter as the woofer). Is this a problem as currently it is flush?
Or is this going to cause more problems than it cures?
Attachments
The truth is that many of these factors are just not audible in practise. While is is commendable to aim for perfection, it can sometiimes be a source of needless worry and wasted time 😉
I had a speaker design with the tweeter mounted in it a separate housing which could be moved back and forward. I played around with it for ages but never did hear any difference.
I had a speaker design with the tweeter mounted in it a separate housing which could be moved back and forward. I played around with it for ages but never did hear any difference.
With all my apologies, Nuuk, I presently prototype a 2 way with a mobile tweeter enclosure and it makes much difference to move it forth or back by 1/2 inch. Hint: you can easyly percieve the lobing from the listening position. The effect for a given move length of the driver (in this case, the tweeter) increase with the crossover frequency.
Rabbitz, given the crossover frequency is in the 2k range, I don't think any horn loading is to appear in your case.
Rabbitz, given the crossover frequency is in the 2k range, I don't think any horn loading is to appear in your case.
With the woofers coil moving forth and back all the time is it really worth wasting some time to align the drivers to perfection???
Guss said:
Now that's a REALLY good question.
It seems that if nothing else, the excursion of the woofer will be responsible for "modulated lobing" and "modulated on axis frequency response" in the area of the crossover frequency.
Another good reason for a 2 way system actively crossed to woofers somewhere around 120Hz.
I was gonna say the same thing about doing a three way. But I have my tweeters in a separate enclosure and moving these back and forth doesnt seem to do much either to the sound. WHen simulating it makes a HUGE difference in the reverse null and a bit of difference in the non reverse.
I am all for time aligning but!, doesnt the amount you have to offset things differ with different crossover freqeuncies. I seem to remember linkwitz saying something about this, calculating time delay for the 4th order at the 1.5khz or so at with the orien.
Also how do you find out where the exact acoustic centres of drivers are? This need to be VERY accurate for LSPCAD pro altering things by half millimeters makes quite big differences.
I am all for time aligning but!, doesnt the amount you have to offset things differ with different crossover freqeuncies. I seem to remember linkwitz saying something about this, calculating time delay for the 4th order at the 1.5khz or so at with the orien.
Also how do you find out where the exact acoustic centres of drivers are? This need to be VERY accurate for LSPCAD pro altering things by half millimeters makes quite big differences.
According to Richard Heyser in one of his JAES papers, the acoustic center of a driver varies with frequency. This makes the time alignment solution much more difficult.
rabbitz said:
I would think that moving the woofer forward slightly is more likely to benefit than have a negative effect in regards to time alignment.Many "time aligned" designs place the tweeter slightly to the rear - moving the tweeter _forward_ is the way you wouldn't want to go.
this is not my area of expertise, however - I remember other posters making comments to the effect that the higher in hz you go the less important it is -
Again, if I remember correctly, this is only pertinent to within 1/2 of a wavelength (1/4 wavelength + or - a paticular wavelength)
Time wise, my guess is that you're OK.
I'm more concerned about protruding 5/8 of an inch - and don't really know, but I bet that's a little bit of a problem - whether it's audible or not, I just don't know, but that is something that probably needs to be looked at -
Regards
Ken L
I bow to your superior hearing 😉 I guess it will also depend on factors like tweeter type/quality and crossover frequency.
Once you add some baffle step compensation to to the bass/mid
in a two way physically vertically aligning the drivers becomes a
question of appearing to be time aligned rather than achieving
any time alignment.
🙂 sreten.
in a two way physically vertically aligning the drivers becomes a
question of appearing to be time aligned rather than achieving
any time alignment.
🙂 sreten.
I'm changing an existing project over to a SEAS 27TFFC tweeter so while I'm at it I thought I try to align the drivers to suit a series 1st order xo.
As far as the acoustic centres go there's a para in the LDCookbook: ".....for crossover design purposes, the only important factor is the relative amount of offset and not the driver acoustic centres. A loudspeaker's acoustic centre varies with frequency and by definition is a function of the natural phase response of the driver........for most purposes..........the radiating centre may be assumed to be the centre of the voice coil........using the voice coil centres will do a reasonable job."
Got to get it closer than it is now as with 1st order the variations in the crossover area are large.
12mm offset, up to 2.5dB
25mm offset, up to 5dB
50mm offset, up to 10dB
Apparently higher orders are very tolerant of offsets which is probably why you don't hear a difference when you move the tweeter back and forth.
Might be as simple as reversing the driver position with woofer on top with the tweeter below, then the driver geometry gives +15° tilt, 1st order xo gives -15° tilt which yields a 0° zero delay plane. I might be totally misunderstanding this, but this could be a solution.
As far as the acoustic centres go there's a para in the LDCookbook: ".....for crossover design purposes, the only important factor is the relative amount of offset and not the driver acoustic centres. A loudspeaker's acoustic centre varies with frequency and by definition is a function of the natural phase response of the driver........for most purposes..........the radiating centre may be assumed to be the centre of the voice coil........using the voice coil centres will do a reasonable job."
Got to get it closer than it is now as with 1st order the variations in the crossover area are large.
12mm offset, up to 2.5dB
25mm offset, up to 5dB
50mm offset, up to 10dB
Apparently higher orders are very tolerant of offsets which is probably why you don't hear a difference when you move the tweeter back and forth.
Might be as simple as reversing the driver position with woofer on top with the tweeter below, then the driver geometry gives +15° tilt, 1st order xo gives -15° tilt which yields a 0° zero delay plane. I might be totally misunderstanding this, but this could be a solution.
This is an existing project which has been finished for quite a while and I don't want to alter the baffle and apparantly a sloped baffle is not a full equivalent. Besides, I'd have to have a 15° angle which would look odd.Bricolo said:
Nuuk, I like your sense of humor, but your hearing is probably not so diffenrent from mine... As Bill Fitz. proposal, try with noise and you sould be fixed... And your hearing too. 😉
Nuuk said:
FYI for Nuuk, NOT trying to bash you, rather simply relating to your comments.
I have found, that in general, one doesn't understand something UNTIL that person, has been made aware of a problem, BY example.
When I say "BY Example" what I mean is that something can sound good, taste good, look good etc. (what about that car that "looked" real good, until it was found that there was 300kg of bodyputty on it ??) Anyhow, to Nuuk, I would have to assume that you have only listened to "better than average" speakers, hence you can't pinpoint the tweeter phase thing on the ones you are talking about, OR, the second fact, might be that when listening you got up and moved them in the midst of a song playing (very hard to hear since one gets very close to the tweeter, AND hears all the phasing going on, as one gets closer & farther in a few seconds, which cancels out the ears abillity to "preceive a difference" in placement for the next 10 min. or so).
I'm NOT being a "basher" when I'm saying this, as IT took me MANY tries to be able to hear 'tweeter phase" in the first place, but help is about to be spoken.... (well, typed, anyhow)...
Pink Noise ! TRUE pink noise (NOT the stuff between channels on the FM radio) get a pair of speakers, and place far enough apart, and away from you, that you can hear the "crossover effect" when moving your head from side to side, so to speak, "in and out of the sweet spot" between the pair, when listening to true pink noise.
The example of this, is the high "wwiisssshhhhh-ooooowwwooooo--hhhhsssssiiiiwww" (don't say it, blow it through your lips) this is "crossover effect" ( can also be called hetrodine effect), our ears are at different locations on our head for a purpose, that reason is so we can disern "loaction" of a sound ( time propagation thing, it takes different times for a sound to reach each ear, hence we can disern direction from that time delay).
BUT.... Now once you can hear the wishie thing from side to side.... Now learn the harder one... transitional effect (tiz also a form of crossover effect, however since there are 2 different drivers (woofer/tweeter, etc) doing the same job at the same time in one array (one cabinet) HERE'S the trick, hold your head sideways, and as you are doing that, squat down and stand up (basicly the same as standing side to side but in the vertical direction) try and find that same "sweet spot" between the 2 drivers IN the same cabinet, IT does take some trial and effort to find the effect, BUT once one has it.. you will never forget it !
In a nut shell, "wwwwooooohhhhhh-wwweeewwww-hhhhhhooooowwww" (again blow it, don't try to speak it)
It will generally be at a lower frequency that the crossover effect, simply because the sound will be from both drivers @ the cross frequency of the speakers crossover.
Oh, By the Way, I am using the industry standard "noted effects" terms NOT that effective when talking "layman" as they are confusing to people, so here is the list:
crossover effect :
The sound heard when one's ears "crossover" from one driver in an array, to another, of the same frequency spectrum.
Transition effect:
The sound heard when one's ears "tranistion" from a driver of one spectrum of sound, to another driver of a different spectrum of sound.
Oh, and one more
Interlacing Effect:
anyone want to try this one ?
These terms are in reference to the listener, and NOT the speaker.
Hey there ford I totally get what you mean with your wshhhhooooooshhhhhh decriptions that was not all in vain!
I again understand what you say about the vertical up and dwn thingy and I think I will try it.
I assume that the idea is to get the vertical sweet spot to correspond with the hight that you listen to the speakers at. If its lower then you move the tweeter back a bit and try again.
I again understand what you say about the vertical up and dwn thingy and I think I will try it.
I assume that the idea is to get the vertical sweet spot to correspond with the hight that you listen to the speakers at. If its lower then you move the tweeter back a bit and try again.
Right on the money 5th..
BUT remember one other thing..
Everything "Out of the sweet spot" is construde as "Off axis" and since your "image" is set the same both V & H, the off axis should follow sute, and simply mellow out, the farther you get off axis from the sweet spot, with out any harsh areas.
BUT remember one other thing..
Everything "Out of the sweet spot" is construde as "Off axis" and since your "image" is set the same both V & H, the off axis should follow sute, and simply mellow out, the farther you get off axis from the sweet spot, with out any harsh areas.
I just fired up cool edit and made sum pink stuff, its is a most intriguing sensation but completely expected, however now the trick is to find out what the sweet spot actually is! Its also fairly hard to do this due to the logistics of a human body
Out of interest these are ss97 tweet and Xcel w15cy mids crossed at 2k, im gonna try 1.5k too.
I suppose another main question is that if the acoustics centres are not entirely important then what data should you give to a LSPcad? If both drivers are mounted on the front baffle do you just guess as to where they are? I know there is a program out there that can calculate where they are.
I would imagine this is done with respect to the front panel, you input the distance of th mic to the baffle and then a chirrup is produced knowing the speed of sound, the program can calculate the time taken for the chirrup to be produced by the loudspkr and recieved by the mic, then takes into account the mic distance and bobs you mothers brother, it tells you where your acoustic centre is.
Out of interest these are ss97 tweet and Xcel w15cy mids crossed at 2k, im gonna try 1.5k too.
I suppose another main question is that if the acoustics centres are not entirely important then what data should you give to a LSPcad? If both drivers are mounted on the front baffle do you just guess as to where they are? I know there is a program out there that can calculate where they are.
I would imagine this is done with respect to the front panel, you input the distance of th mic to the baffle and then a chirrup is produced knowing the speed of sound, the program can calculate the time taken for the chirrup to be produced by the loudspkr and recieved by the mic, then takes into account the mic distance and bobs you mothers brother, it tells you where your acoustic centre is.
Done......
Thanks for your replies and comments.
After much reading and research ended up going with woofer on top with tweeter below, IDT (Inverted Driver Technology - name depends on whose marketing hype you read). I've made the assumption that a 1st order series has similar -15° polar tilt as a 1st order butterworth with normal polarity on the drivers when Zeta=1.
With the 1st order series xo giving a -15° polar tilt and the drivers giving a +20° radiation tilt, I am a lot closer than before with a ZDP of +5°. Tried to have a +15° tilt on the drivers, but the driver centres were too large. Wanted them less than the wavelength of the xo....... had to compromise as I felt the driver centres were more important.
The sound has a better image and depth than before so it has worked.
Here's the modified baffle and if anyone ever asks what's the black circle around the tweeter, I'll just tell them it's a new fandangled reflection absorber.
Thanks for your replies and comments.
After much reading and research ended up going with woofer on top with tweeter below, IDT (Inverted Driver Technology - name depends on whose marketing hype you read). I've made the assumption that a 1st order series has similar -15° polar tilt as a 1st order butterworth with normal polarity on the drivers when Zeta=1.
With the 1st order series xo giving a -15° polar tilt and the drivers giving a +20° radiation tilt, I am a lot closer than before with a ZDP of +5°. Tried to have a +15° tilt on the drivers, but the driver centres were too large. Wanted them less than the wavelength of the xo....... had to compromise as I felt the driver centres were more important.
The sound has a better image and depth than before so it has worked.
Here's the modified baffle and if anyone ever asks what's the black circle around the tweeter, I'll just tell them it's a new fandangled reflection absorber.
Attachments
Swapping driver positions......
To do this modification, I had to close the existing woofer hole and machine new ones.
Done by installing new centres for the router jig (home made Jasper style), then route the hole in the old woofer postion for the blanking piece which has a flange for location and strength. Once this was done, just a matter of routing the new holes and rebate for the tweeter, and adding stiffeners under the baffle. Since the woofer overlaps the tweeter to get a smaller centre distance, it was not rebated.
Here's a pic.
To do this modification, I had to close the existing woofer hole and machine new ones.
Done by installing new centres for the router jig (home made Jasper style), then route the hole in the old woofer postion for the blanking piece which has a flange for location and strength. Once this was done, just a matter of routing the new holes and rebate for the tweeter, and adding stiffeners under the baffle. Since the woofer overlaps the tweeter to get a smaller centre distance, it was not rebated.
Here's a pic.
Attachments
- Status
- Not open for further replies.