Tried going OB because the sealed mid was not cutting it, sounded too forward.
Obviously this was a test so was quick, non pretty MDF with no driver rebates but that doesnt matter.
The change from sealed to OB was not as massive as I had thought it would be but the more you listen to it the more you realise how free OB sounds when compared to boxes!
The biggest difference that was easily noticable was human speech from the TV, it sounds much more natural.
Either way ill continue to soak in sonic bliss. Now all I want to do is try it MTM....... why those excels so expensive!
Bass is XLS10 tweet a scan97.
Obviously this was a test so was quick, non pretty MDF with no driver rebates but that doesnt matter.
The change from sealed to OB was not as massive as I had thought it would be but the more you listen to it the more you realise how free OB sounds when compared to boxes!
The biggest difference that was easily noticable was human speech from the TV, it sounds much more natural.
Either way ill continue to soak in sonic bliss. Now all I want to do is try it MTM....... why those excels so expensive!
Bass is XLS10 tweet a scan97.
Attachments
Wow, that looks a lot like what I'm planning on trying with my setup. I have a 12" woofer in a sealed enclosure, and will be adding a midrange and tweeter on top. That's encouraging to know that it sounds good 
What did you do for the crossovers (active/passive, frequencies, slopes)? How did you decide on the baffle dimensions and the driver placement?
What did you do for the crossovers (active/passive, frequencies, slopes)? How did you decide on the baffle dimensions and the driver placement?
I've commented on your post just now, if you read this first its not good news im affraid
I have mentioned excursion as a limiting factor. These W15CY's have 4mm one way linear and are the limiting factor quite considerably when it comes to SPL, this is not a power issue, the driver simply runs out of excursion.
Well the crossover is active which makes dipole EQ easy. I had used some much smaller sealed boxes for the mid/tweet before. How did I come up with baffle dimensions? Simple make it the same width as the bass box! and then make the height just right for tweeter axis on ears! Driver placement? just thought id try offset tweeters.
I then simply measured the whole lot to set EQ levels correctly and to alter the xover.
I have mentioned excursion as a limiting factor. These W15CY's have 4mm one way linear and are the limiting factor quite considerably when it comes to SPL, this is not a power issue, the driver simply runs out of excursion.
Well the crossover is active which makes dipole EQ easy. I had used some much smaller sealed boxes for the mid/tweet before. How did I come up with baffle dimensions? Simple make it the same width as the bass box! and then make the height just right for tweeter axis on ears! Driver placement? just thought id try offset tweeters.
I then simply measured the whole lot to set EQ levels correctly and to alter the xover.
Looks like we're posting to two threads in parallel Thanks for your advice, I'll see how things go. If OB doesn't work then I'll try a sealed box and place the horn tweeter on top.
I just made shorter speaker stands to accomodate this baffle-on-top idea, and now I'm thinking I might have to make them even shorter. Or put the tweeter on the bottom and the midrange on top.
Thanks for your advice, I'll see how things go. If OB doesn't work then I'll try a sealed box and place the horn tweeter on top.I just made shorter speaker stands to accomodate this baffle-on-top idea, and now I'm thinking I might have to make them even shorter. Or put the tweeter on the bottom and the midrange on top.
Could the dipole mid arrangement be rolling off your lower mid output around 150 Hz?
The Sd, (cone area) of the Seas W15CY is 75 sq cm. If the cone excursion is +/- 4 mm, that comes to about 1.9 cu inches.
According to Small's chart, 112 dB at 150 Hz requires about 3 cu inches. Since you will be crossing over at 150 Hz, the output should be 3 dB down or so.
All in all, it would seem that you have enough output down to 150 Hz for a system SPL of 110 dB or more. That ought to be loud.
Is it possible you are getting cancellation at the lowest frequencies of the dipole section?
http://www.diyaudio.com/forums/showthread.php?threadid=5668&highlight=
Hmmm interesting you should say that.
I made the dipole EQ following how linkwitz did it with the pheonix. And all I can say is that with music I run out of excursion. Thing is even though I might be able to achieve 110dB output at 150hz, there is considerable output below 150hz which causes the cone to obviously move more. I think perhaps that the increase in cone excursion at these frequencies is greater than the amount of attenuation a 2nd order electrical xover can provide. Also if I was getting any cancelation surely that would just affect the audible loudness and not the cone displacement?
When I say they cant go loud, I mean subjectively in comparison to a sealed cabient, they do still produce a fair bit of level.
I will look into that link you provided but now I need to sleep
Cheers Matt
P.S. if you are going to make a three way and use active xovers and and are using a mid/bass for mids DO OPEN BAFFLE, your ears will thankyou for it. You do get enough SPL out of them and they are not difficult to design.
Also when I said quite considerably I was also refering to how the woofer and tweeter were performing and that they could each go louder, the mid was the limit and not the other two.
I made the dipole EQ following how linkwitz did it with the pheonix. And all I can say is that with music I run out of excursion. Thing is even though I might be able to achieve 110dB output at 150hz, there is considerable output below 150hz which causes the cone to obviously move more. I think perhaps that the increase in cone excursion at these frequencies is greater than the amount of attenuation a 2nd order electrical xover can provide. Also if I was getting any cancelation surely that would just affect the audible loudness and not the cone displacement?
When I say they cant go loud, I mean subjectively in comparison to a sealed cabient, they do still produce a fair bit of level.
I will look into that link you provided but now I need to sleep
Cheers Matt
P.S. if you are going to make a three way and use active xovers and and are using a mid/bass for mids DO OPEN BAFFLE, your ears will thankyou for it. You do get enough SPL out of them and they are not difficult to design.
Also when I said quite considerably I was also refering to how the woofer and tweeter were performing and that they could each go louder, the mid was the limit and not the other two.
5th element said:
Well, I am no expert on dipoles. My reading on the subject is largely confined to HiFi Loudspeakers and Enclosures, written by Abraham Cohen in 1956. As I recall, he said that to prevent cancellation, the open baffle should be one half wavelength of the lowest frequency passed. Since the wavelength of 1 Hz is 13,500 inches, we have:
13,500 inches ÷ 150 = 90 in.
The wavelength of 150 Hz is 90 in. But we only need half a wavelength for our baffle. So: 90 ÷ 2 = 45 inches.
However, judging by your picture with the Peerless 10 incher, your baffle is only about 12 inches wide-about 1/4 what is needed to prevent partial cancellation.
I realize that Kuei has said that the baffle width is less than what people said way back when, but still-this is only one quarter the old standard.
This would still not cause excess cone excursion by itself. It would cause a drop in SPL. But if your system is EQ'd to compensate for this loss of SPL due to partial cancellation, then definitely cone excursion needs are increased-perhaps drastically. If the SPL dropoff is 6 dB and you equalize to compensate, then your cone has to travel twice as far as if there were no cancellation.
I think that might be happening here.
As long as this is not a finished product, how about it you just temporarily attach two pieces a foot wide two feet high or so and going back at 90º, sort of to follow the contour of the box sides? Just something to knock together quickly. See if that changes the situation at all. You might have to minimize the equalization to bring things back into balance-but that would be a good thing for the cone excursion.
just note that for the open baffle :
Above the units bass resonance 12dB / octave roll-off keeps
the excursion constant for the same input at any frequency.
So with the baffle boost before 150Hz quite a lot of unecessary
low frequency excursion will be going on. I'd say you need 3rd
order electrical roll-off to give a 4th order L/R to match the bass
unit crossover, lifting 150HZ a little higher would also help.
sreten.
Above the units bass resonance 12dB / octave roll-off keeps
the excursion constant for the same input at any frequency.
So with the baffle boost before 150Hz quite a lot of unecessary
low frequency excursion will be going on. I'd say you need 3rd
order electrical roll-off to give a 4th order L/R to match the bass
unit crossover, lifting 150HZ a little higher would also help.
sreten.
Member
Joined 2003
5th,
When I was experimenting with rough wood, I attached 6-12" baffle side panels with hinges to quickly adjust wing angle. Signal was wide-band pink noise through a P17...no EQ as I was interested only in "character". I found 90 degree panels introduced a distinct resonance. I assume the resonance could have been damped (just like a box) but decided not to mess with mother nature. (Angled wings may also help minimize diffraction effects.)
While you're experimenting, you might try hinges to see what you think.
Paul
4-way dipoles
When I was experimenting with rough wood, I attached 6-12" baffle side panels with hinges to quickly adjust wing angle. Signal was wide-band pink noise through a P17...no EQ as I was interested only in "character". I found 90 degree panels introduced a distinct resonance. I assume the resonance could have been damped (just like a box) but decided not to mess with mother nature. (Angled wings may also help minimize diffraction effects.)
While you're experimenting, you might try hinges to see what you think.
Paul
4-way dipoles
Sreten yes I realise that the mid is not rolled off at 24dB octave to match the bass as of yet because the active xover was built for the sealed application. 12dB roll of sealed cab + the 12dB roll off of the active filter gave me the 24dB slopes. Yes I would agree that making the active filter 18dB would be good and when I get round to it I will.
I am very happy with the way it sounds and max SPL is not that low anyway. If funds permit Ill MTM it anyway and use 18dB xover, the combined 6dB extra boost from MTM and the additional 6dB in the roll off should give me more then enough extra SPL.
No this is not a finished design! I intend to make the open baffle out of solid wood or MDF and birdseye maple/maple burr veneer. I have a surplus of both anyway.
I am very happy with the way it sounds and max SPL is not that low anyway. If funds permit Ill MTM it anyway and use 18dB xover, the combined 6dB extra boost from MTM and the additional 6dB in the roll off should give me more then enough extra SPL.
No this is not a finished design! I intend to make the open baffle out of solid wood or MDF and birdseye maple/maple burr veneer. I have a surplus of both anyway.
Sretan:
You don't think 5th Element has partial cancellation issues here? At 150 Hz, his wavelength is 90 inches, his baffle is 12 inches. Wouldn't the rear wave diffract around the edge and partially cancel the front wave?
If he wants to handle it by adding another mid, that is fine. I just think that might be the nature of the problem.
You don't think 5th Element has partial cancellation issues here? At 150 Hz, his wavelength is 90 inches, his baffle is 12 inches. Wouldn't the rear wave diffract around the edge and partially cancel the front wave?
If he wants to handle it by adding another mid, that is fine. I just think that might be the nature of the problem.
kelticwizard said:
He's got open baffle roll-off allright, if thats what you mean,
which he's compensated for by adding a progressive boost
down to 150Hz.
Cone excursion may be equivalent to 110dB SPL but actual
SPL certainly is nowhere near this due to the OB roll-off.
As the bass end is not dipole, and you don't have H or U
frame resonances to consider, which sets Linkwitz's low
crossover frequency going the MTM route does not seem
sensible.
A c/o frequency of 250 Hz, less open baffle compensation boost
and 18dB octave c/o slope would probably achieve similar results.
sreten.http://www.geocities.com/kreskovs/NaO.html
This may be of some interest, regarding the extra tweeter.
sreten.
This may be of some interest, regarding the extra tweeter.
sreten.
5th element said:
Simply that your not being forced to go MTM by circumstances,
and that it appears some redesign would get round the MTs
problems or at least decrease them substantially.
I find it hard to why 150Hz is the crossover frequency in this
case, I can't see the technical justification for it being so low.
But if you want dipole to as low a frequency as possible
then an MTM is obviously going to be better than an MT.
sreten.
I dont know the XLS resolving power of detail etc in comparison to the W15CY's. I had originally got the W15 crossed over at 300hz to a peerless 850146 lowering this to 150 sounded better because the lower range of the W15 was better then the 850's.
I would have thought the XLS would be worse then the 850 when it comes to detail etc so id prefer to cross over low. 150 sounds like it works well. When listening to the bass units on their own very little midrange is apprent. So the w15's are doing all the work where its required for the best sound quality. If I was using a W26F excel bass for the bass then I wouldnt hesistate to cross over higher, but to get the best sound from this I think 150 is about best.
I would have thought the XLS would be worse then the 850 when it comes to detail etc so id prefer to cross over low. 150 sounds like it works well. When listening to the bass units on their own very little midrange is apprent. So the w15's are doing all the work where its required for the best sound quality. If I was using a W26F excel bass for the bass then I wouldnt hesistate to cross over higher, but to get the best sound from this I think 150 is about best.
5th element:
It's up to you, of course. But since it takes only half the volume of air moved to generate a tone half an octave above at the same SPL, then changing the crossover point to 210 Hz instead of 150 Hz will result in an increase of max SPL for the midrange of 6 dB. That is with the same cone excursion.
Also, since the diffraction rolloff increases as you go lower, you might even get an additional decibel or two on top of that.
It's up to you, of course. But since it takes only half the volume of air moved to generate a tone half an octave above at the same SPL, then changing the crossover point to 210 Hz instead of 150 Hz will result in an increase of max SPL for the midrange of 6 dB. That is with the same cone excursion.
Also, since the diffraction rolloff increases as you go lower, you might even get an additional decibel or two on top of that.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.