Hi friends, i need some suggestions. I have built a pair of speakers with 2 piece of peerless 830860 in paralell. seas 27tdfc as tweeter. I have put together a crossover that sounds okay, but on some recordings it is a kind of harshness to the upper midrange, and the clarity isnt as good as i want. I have thought of doing a 2 1/2 way, but then i will lose sensivity?
Thanks
Thanks
Attachments
No you don't loose any sensitivity by going with a 2.5-way.
Maybe you just need to refine your xo but a 2.5-way may be a better solution since the center to center spacing between the tweeter and the bottom woofer is far too large to be desirable.
Can you post the following?
1 - baffle dimensions and driver positioning
2 - net internal volume and port dimensions (if ported)
3 - resistances of your 2 inductors
Maybe you just need to refine your xo but a 2.5-way may be a better solution since the center to center spacing between the tweeter and the bottom woofer is far too large to be desirable.
Can you post the following?
1 - baffle dimensions and driver positioning
2 - net internal volume and port dimensions (if ported)
3 - resistances of your 2 inductors
vvindir, if you are asking if you take your current two way and remake the crossover into a 2.5way, yes you'll loose some effeciency, about 3dB in the midrange to the crossover point on the 860's. Reducing the crossover point on the lower woofer to say something like 400Hz or so allowing it to compensate for baffle step loss in the low midrange bass region and down.
As you have it configured now there is bad diffraction issues due to both woofers placed at increasingly further distance with the same signal driven into will muck up your mids for sure. Changing the crossover to a 2.5way will correct most of this problem. You will also be required to reduce the tweeter level by roughly 3dB.
As you have it configured now there is bad diffraction issues due to both woofers placed at increasingly further distance with the same signal driven into will muck up your mids for sure. Changing the crossover to a 2.5way will correct most of this problem. You will also be required to reduce the tweeter level by roughly 3dB.
I seem to be in "It won't work..." mode lately, but this design looks badly thought out. Shame because your box is superb! 
Why should it be wired in negative polarity? It makes no sense. It also looks like there is too much overlap between the drivers at crossover. Phase and frequency response must be all over the place. The box looks oversized too. Vas of about 10.6L on Qts of 0.41 per woofer.
Maybe you can find a suitable adaptable design for 130mm bass/mids at Troels Gravesen.
This one using your woofer looks like a start:
Peerless HDS PPB 830860
Here's how you MTM D'Appolito some similar drivers:
Vifa PL14WJ-
How you adapt all that to 2.5 way, or D'Appolito is something you can spend more time than I will. But Troels' designs might help.
Why should it be wired in negative polarity? It makes no sense. It also looks like there is too much overlap between the drivers at crossover. Phase and frequency response must be all over the place. The box looks oversized too. Vas of about 10.6L on Qts of 0.41 per woofer.
Maybe you can find a suitable adaptable design for 130mm bass/mids at Troels Gravesen.
This one using your woofer looks like a start:
Peerless HDS PPB 830860
Here's how you MTM D'Appolito some similar drivers:
Vifa PL14WJ-
How you adapt all that to 2.5 way, or D'Appolito is something you can spend more time than I will. But Troels' designs might help.
Last edited:
I don't see the box as too big. You can always add bricks, chunks of wood scrap etc, anything non compressible with a known volume to reduce the enclosures internal volume..
Tweeter can be crossed down to 2.2k (looks like) which makes a very good match to the 860 and is as close as you could get without cutting away it's bezel like Troels does
Basically the way I see it change the crossover, keep the nice enclosures
Tweeter can be crossed down to 2.2k (looks like) which makes a very good match to the 860 and is as close as you could get without cutting away it's bezel like Troels does
Basically the way I see it change the crossover, keep the nice enclosures
...and since you're more woodworker
I tried something new on the baffle called plastidip, the result was horrible, i have to sand everything down and apply the good old bengalakk.
I tried something new on the baffle called plastidip, the result was horrible, i have to sand everything down and apply the good old bengalakk.
OMG
know the stuff too well
Hopefully, this will give you an idea of what's going on.
First graph below is a simulation of your speakers with your current xo's using Peerless's specs for the woofers and Zaph's measurement for the tweeter. Baffle diffraction and loss for your baffle dimensions and driver positioning are included. Tweeter polarity is reversed.
Black line is the summed response.
Green is the power response.
Grey is the response with the polarity reversed.
And the thinner lines are the drivers' phase response.
Crossover is at about 4000Hz. Clearly the polarity shouldn't be reversed and yet that deep null may not be that bad if you also take the power response in the xo region into account, which is still fairly high in SPL and probably averages things out a little bit. It may in fact sound better than with the tweeter connected with normal polarity because that results in more peaking around 2000 - 3000Hz which might sound worse. Phase alignment though is nowhere near to being close.
In terms of hearing some harshness in the upper midrange, I suspect it is coming from the woofers in the 2000 - 4000Hz range but I'm just guessing. Looking at the raw FR of those drivers, between 2000 and about 6000Hz appears to be where some driver resonances are going on. The Seas Fs is way down at 500Hz and is recommended for a xo down to 1500Hz so I don't think it's coming from there.
Second graph is your same speaker but this time with a new 2.5-way xo but with the xo point moved down this time to 2000Hz. The woofers contribution between 2000 and 4000Hz is much less this way. Phase is almost perfectly aligned right at the xo point which is excellent. Xo's are both still 2nd order but with an additional inductor on the 2nd, .5 woofer after the 1st woofer. Woofer slope is 3rd order acoustic and the tweeter slope is almost 4th order acoustic. Impedance minima are at about 3ohm.
Here are the values I used:
Tweeter
C = 13uF
L = .16mH (.5ohm)
R (after the xo) = .75ohm
Top and Bottom Woofer (before both of them)
L = 1mH (.38ohm)
C = 33uF (+ .5ohm shunt resistor)
Bottom Woofer (after the 1st one)
L = 1.75mH (.6ohm)
Let me know if you want me to post a schematic. The resistance of the woofer inductors can probably be changed a little without much effect.
Cheers
First graph below is a simulation of your speakers with your current xo's using Peerless's specs for the woofers and Zaph's measurement for the tweeter. Baffle diffraction and loss for your baffle dimensions and driver positioning are included. Tweeter polarity is reversed.
Black line is the summed response.
Green is the power response.
Grey is the response with the polarity reversed.
And the thinner lines are the drivers' phase response.
Crossover is at about 4000Hz. Clearly the polarity shouldn't be reversed and yet that deep null may not be that bad if you also take the power response in the xo region into account, which is still fairly high in SPL and probably averages things out a little bit. It may in fact sound better than with the tweeter connected with normal polarity because that results in more peaking around 2000 - 3000Hz which might sound worse. Phase alignment though is nowhere near to being close.
In terms of hearing some harshness in the upper midrange, I suspect it is coming from the woofers in the 2000 - 4000Hz range but I'm just guessing. Looking at the raw FR of those drivers, between 2000 and about 6000Hz appears to be where some driver resonances are going on. The Seas Fs is way down at 500Hz and is recommended for a xo down to 1500Hz so I don't think it's coming from there.
Second graph is your same speaker but this time with a new 2.5-way xo but with the xo point moved down this time to 2000Hz. The woofers contribution between 2000 and 4000Hz is much less this way. Phase is almost perfectly aligned right at the xo point which is excellent. Xo's are both still 2nd order but with an additional inductor on the 2nd, .5 woofer after the 1st woofer. Woofer slope is 3rd order acoustic and the tweeter slope is almost 4th order acoustic. Impedance minima are at about 3ohm.
Here are the values I used:
Tweeter
C = 13uF
L = .16mH (.5ohm)
R (after the xo) = .75ohm
Top and Bottom Woofer (before both of them)
L = 1mH (.38ohm)
C = 33uF (+ .5ohm shunt resistor)
Bottom Woofer (after the 1st one)
L = 1.75mH (.6ohm)
Let me know if you want me to post a schematic. The resistance of the woofer inductors can probably be changed a little without much effect.
Cheers
Attachments
Your choice of drivers is very good. Sadly a 2 way system like this has a crossover point right in the area of frequency where the ear is most sensitive (1kHZ - 6kHZ). I happen to be using those same 6 inch Peerless drivers as woofers in a 3 way system I built, and I am amazed at how well they perform. Very few 1 inch dome tweeters perform well below about 4kHZ with a one pole xover filter, 3kHZ with a 2 pole, etc.. I'm using the Seas Millenium tweeter in one system all the way down to 1.4kHZ, but with a 4 pole active crossover (similar to what Linkwitz is doing in his Orion system). My research showed that there's only a handfull of 1 inch drivers that can do that well. Above about 2kHZ, most 6 inch drivers not only get too directional, but also have resonance issues right where the ear is most sensitive (3-6kHZ). The harder the cone material, the more severe the resonance. You want those to be rather attenuated, since being a resonance, they smear the sound in time (ring), which makes them even more perceivable psycho-acoustically.
I'd go 3 way and biamp the lower crossover point with 4th order (24dB/oct cutoff rate). X would be about 500HZ 4 pole active, I'd also use a closed box so I could use active EQ to force the bass to be acoustically flat down to 32HZ.. I'd use the Vifa 3 inch TC9FD18-08 driver in a sub-enclosure for the midrange. That's actually virtually what I'm doing in a speaker system I'm building right now. I have an 8 inch going up to 500HZ, 4th order active X to the 3 inch Vifa which goes up to 7kHZ, where a 1/2 inch dome takes over with a 1 pole Xover. When the X is above 6kHZ, I recommend ribbon tweeters, such as the Fountek 1.5 inch.
By the way, a speaker that measures flat in an anechoic chamber will often sound harsh in a real world room, hence the invention of tone controls. Probably largely because of the abrupt dispersion change around 2-4kHZ.
I'd go 3 way and biamp the lower crossover point with 4th order (24dB/oct cutoff rate). X would be about 500HZ 4 pole active, I'd also use a closed box so I could use active EQ to force the bass to be acoustically flat down to 32HZ.. I'd use the Vifa 3 inch TC9FD18-08 driver in a sub-enclosure for the midrange. That's actually virtually what I'm doing in a speaker system I'm building right now. I have an 8 inch going up to 500HZ, 4th order active X to the 3 inch Vifa which goes up to 7kHZ, where a 1/2 inch dome takes over with a 1 pole Xover. When the X is above 6kHZ, I recommend ribbon tweeters, such as the Fountek 1.5 inch.
By the way, a speaker that measures flat in an anechoic chamber will often sound harsh in a real world room, hence the invention of tone controls. Probably largely because of the abrupt dispersion change around 2-4kHZ.
Last edited:
Sorry, but those values give (at least in my simulation) a peak of 2.5 dB in the filter transfer function and a very low impedance of 2.1 ohm at 4 kHz. Typo or intention?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.