I know for sure that no crossover simulator I've ever used is designed for OB. It can help you with some things, but it will be wrong about a lot. Once you do understand how different OB crossovers are, the more you can judge what the sims are telling you and how they apply.
The problem is that not a lot of people have experience with OB that they can pass on to others. If we were all building open baffle speakers, then we might have trouble designing box speakers. There are just other issue to deal with.
The best is to ask advice from experienced builders. It's not all trial and error, it's knowing the what works for OB.
The problem is that not a lot of people have experience with OB that they can pass on to others. If we were all building open baffle speakers, then we might have trouble designing box speakers. There are just other issue to deal with.
The best is to ask advice from experienced builders. It's not all trial and error, it's knowing the what works for OB.
Here's the basic approach that I know for OB. You need to figure out how the woofers are going to react on the baffle you've chosen. On an open baffle the bass cancellation will cause a rising response of the woofer. The sloping difference between the LF and mids can often be 10dB, You don't get that 10dB cancellation loss back, you have to attenuate the woofer's rising response to get a flat tonal balance. If you don't do that, you just end up with a lot of midrange. I have heard (and built) too many like that.
The size of the baffle and wings along with the type of woofer chosen will determine the shape of the woofer's rising response. Martin king choose the Alpha for its high Qts which means a flatter OB response at the bottom end. There is relatively less midrange to trade away.
Once you've started to tame the rising response of your woofers, you'll need to figure out how to match them to your midrange or fullrange. This is a little more like normal speaker work, except that the back wave off the mid-full will be bouncing off the walls and coming back to you - and that matters. When using a small fullrange driver you might end up crossing at 700hz to blend with the roll-off you've given to the woofers. With a driver as large as the Beta 12 LTA, you will probably crossover lower. You have to take that into account when shaping the response of your woofers.
Those are the basics. Other things like traps, EQ, Zobles, L-Pads, etc are similar enough to box speakers that you should have not particular problems.
The size of the baffle and wings along with the type of woofer chosen will determine the shape of the woofer's rising response. Martin king choose the Alpha for its high Qts which means a flatter OB response at the bottom end. There is relatively less midrange to trade away.
Once you've started to tame the rising response of your woofers, you'll need to figure out how to match them to your midrange or fullrange. This is a little more like normal speaker work, except that the back wave off the mid-full will be bouncing off the walls and coming back to you - and that matters. When using a small fullrange driver you might end up crossing at 700hz to blend with the roll-off you've given to the woofers. With a driver as large as the Beta 12 LTA, you will probably crossover lower. You have to take that into account when shaping the response of your woofers.
Those are the basics. Other things like traps, EQ, Zobles, L-Pads, etc are similar enough to box speakers that you should have not particular problems.
Thanks, Pano. Now I’m starting to get the picture from you and John. And I guess it really is obvious once you think about it. Most crossover simulators expect a box around the woofer, either sealed or ported, and provide a fairly predictable response accordingly.
But with no box, and a dipole sound radiation pattern, none of the standard formulas are applicable any more. So it’s almost impossible to design a crossover with the current programs. They don’t even address this issue. It’s a whole new ball game. Particularly with the rising response from the woofer that you describe.
However, this brings me back to the idea of a DSP. At least to get into the ballpark. It would seem that with a DSP, a microphone, and REW, you could at least get close to a good crossover design. Otherwise, since simulators are ruled out, how do you know even where to start.
But with no box, and a dipole sound radiation pattern, none of the standard formulas are applicable any more. So it’s almost impossible to design a crossover with the current programs. They don’t even address this issue. It’s a whole new ball game. Particularly with the rising response from the woofer that you describe.
However, this brings me back to the idea of a DSP. At least to get into the ballpark. It would seem that with a DSP, a microphone, and REW, you could at least get close to a good crossover design. Otherwise, since simulators are ruled out, how do you know even where to start.
Last edited:
I would recommend active crossovers and multiple amps, whether you use DSP or analogue crossovers. It will give you so much flexibility in matching driver outputs.
Last edited:
Yes, it's a different ball game. It's like you grew up playing baseball and understand the game, then you try to play cricket. Same thing, right? A ball is thrown, hit with a bat, some running. Same physics apply, but it's a different game and easy to get lost.
Yes, please do try DSP. It's very fast and flexible. Just keep in mind that "normal" approaches might not be the best approach.
Yes, please do try DSP. It's very fast and flexible. Just keep in mind that "normal" approaches might not be the best approach.
Thanks. That's actually part of the plan eventually. Have been looking at the 7 channel Emotiva amp just for that purpose.
One of the issues with DSP appears to be the quality of DAC that follows it. I have a miniDSP 2x4 HD and based on reports I don't expect the built-in DAC performance to be that good. There are solutions with external DACs, but it gets messy and expensive when you need 4 or 6 channels for a 2-way or 3-way speaker.
If you don't mind saying, I'm curious how you dealt with the DAC issue in your testing of DSP.
For what it's worth, I have built many biamped OB's with Eminence or
similar woofers crossed over to full rangers.
OB's give a different phase and gain curve than enclosures, so you
don't want to depend too much on textbook crossover settings.
The vast majority ended up doing their best with 12 dB/oct crossovers
ranging from about 60 to 200 Hz - usually the woofer is crossed over
approximately an octave below the FR driver, and most of the time the
two drivers are operated in-phase.
Interestingly, SL applied a very similar approach on the LX-minis, and
in the store you will find the ACN crossover that works pretty well for
this sort of thing.
similar woofers crossed over to full rangers.
OB's give a different phase and gain curve than enclosures, so you
don't want to depend too much on textbook crossover settings.
The vast majority ended up doing their best with 12 dB/oct crossovers
ranging from about 60 to 200 Hz - usually the woofer is crossed over
approximately an octave below the FR driver, and most of the time the
two drivers are operated in-phase.
Interestingly, SL applied a very similar approach on the LX-minis, and
in the store you will find the ACN crossover that works pretty well for
this sort of thing.
Yep. Crossing the woofer about an octave lower than the mid will mean that the woofer's acoustic output rolls off where the mid starts coming in. You'll end up with woofer low pass filters much lower than you think they should be, but the rising woofer response actually pulls the crossover point back up.
I've heard at least one pair of Nelson's OB speakers with fullrange mids, they were very nice.
The crossover Nelson mentioned is here:
Analog Crossover Network – diyAudio Store
I have not used one, but have probably heard one.
I've heard at least one pair of Nelson's OB speakers with fullrange mids, they were very nice.
The crossover Nelson mentioned is here:
Analog Crossover Network – diyAudio Store
I have not used one, but have probably heard one.
Part of the mix is there will be a narrow F peak with a typical medium to large OB baffle any where between 275-400 Hz. A typical first or second order passive crossover will do little to dampen this extra 2 or 3 db of narrow output. So you cross over a bit lower than calculations indicate you should and gee wiz, you end up with a surprisingly flat response in the crossover region. It is all part of the game.
can the frequency of that peak be derived in a straightforward manner from baffle dimensions? "It's more mysterious than that" or some such is a perfectly acceptable answer, and thanks to anyone who has the time.
Should be able to calculate it. In fact I think it shows up using Edge and Basta software if you put them in OB mode.
Wow, thanks for all your replies. It does show to me that the simple answer I was looking for isn't that simple..... Since I can get the MiniDSP at a discount I ordered one... Tomorrow it will be delivered. This doesn't mean I loose the possible passive solution out of sight.
But this does mean I can get building and experiment..... All these drivers are lying here way to long to be heard....
I will keep you posted, R.
But this does mean I can get building and experiment..... All these drivers are lying here way to long to be heard....
I will keep you posted, R.
Please keep us up to date on your progress.Nelson, Pano, John,
Thank you all for the posts explaining the issues with designing crossovers for open baffle. Those may be some of the most valuable posts of all for OB novices.
I agree with Xerxes75 - Wow. It shows that the simple answer some of us have been looking for isn’t that simple. And like him, I think I’ll try the miniDSP route to start with and at least get in the ball park. From there maybe Nelson’s ACN or a passive XO would be the next step.
And this brings me to another related issue. I tried several months ago to get the administrators of this board to set up a separated category under Loudspeakers for Open Baffle. They rejected my suggestion as not being necessary, particularly since a new atomic powered, super duper, forum search feature was imminent and would preclude the need for a new and separate OB category.
I haven’t seen this new feature, yet, and still remain convinced that a separated category for OB is very desirable. And I think this tread really proves my point. It will probably be lost to most future OB fans, since it appears under the Multi-Way heading. But would be found much easier if it was under an Open Baffle section heading. In fact, I think it is so valuable that it would be a Sticky under that section if such existed.
So finally, if any of you three, or anyone else, agrees with me about the need for a separated OB section and have any clout with the administrators you may want to see what you can do to have them set up a separate OB topic under Loudspeakers.
Thank all three of you again for your very valuable contributions.
Thank you all for the posts explaining the issues with designing crossovers for open baffle. Those may be some of the most valuable posts of all for OB novices.
I agree with Xerxes75 - Wow. It shows that the simple answer some of us have been looking for isn’t that simple. And like him, I think I’ll try the miniDSP route to start with and at least get in the ball park. From there maybe Nelson’s ACN or a passive XO would be the next step.
And this brings me to another related issue. I tried several months ago to get the administrators of this board to set up a separated category under Loudspeakers for Open Baffle. They rejected my suggestion as not being necessary, particularly since a new atomic powered, super duper, forum search feature was imminent and would preclude the need for a new and separate OB category.
I haven’t seen this new feature, yet, and still remain convinced that a separated category for OB is very desirable. And I think this tread really proves my point. It will probably be lost to most future OB fans, since it appears under the Multi-Way heading. But would be found much easier if it was under an Open Baffle section heading. In fact, I think it is so valuable that it would be a Sticky under that section if such existed.
So finally, if any of you three, or anyone else, agrees with me about the need for a separated OB section and have any clout with the administrators you may want to see what you can do to have them set up a separate OB topic under Loudspeakers.
Thank all three of you again for your very valuable contributions.
I think OB's have gained quite a bit of ground since 1973 when Audio Magazine
turned down an article I wrote about them, saying that they didn't see interest
in it. Of course perhaps they were simply being polite....
turned down an article I wrote about them, saying that they didn't see interest
in it. Of course perhaps they were simply being polite....
There is good news and some bad news. The good news is that my MiniDSP arrived en I managed to get some large holes in pair of old door and fitted the drivers.
I got the MiniDSP up an running... In the first test however I noticed that one of my Alpha 15A was stuck.
I removed it to investigate... It does move but very difficult.
So I wil disconnect one of the Alpha's on the other speaker so I can at least listen to it tomorrow....
I got the MiniDSP up an running... In the first test however I noticed that one of my Alpha 15A was stuck.
I removed it to investigate... It does move but very difficult.
So I wil disconnect one of the Alpha's on the other speaker so I can at least listen to it tomorrow....
Oh no, that is bad news. 🙁 Well, as least they aren't expensive drivers of you want to get another.
A little tardy to the party but I'd just like to say that I'm having a ball with 2 MiniDSP 2x4's running a pair of 4 way OB's. Only thing I didn't like was the lack of a power supply. Seriously?
Otherwise, the price to performance ratio is unbeatable. I have spent more on passive X/O parts for a three way pair than I paid for one of these and a nice second hand amp. And that was just the first round of design. After ordering more and more parts after several sessions of tweaking the passive X/O I could have bought another MiniDSP and amp. Just to be clear: I'm talking about a passive design I was working on before I started on the OB actives.
If you like building speakers or just want to be one and done, a multi-amp DSP setup isn't a purchase, it's an investment that will save you time and money in the long run IMHO (if you know what you're doing).
Even if you don't know what you're doing the chances of coming away with something listenable are much better.
Otherwise, the price to performance ratio is unbeatable. I have spent more on passive X/O parts for a three way pair than I paid for one of these and a nice second hand amp. And that was just the first round of design. After ordering more and more parts after several sessions of tweaking the passive X/O I could have bought another MiniDSP and amp. Just to be clear: I'm talking about a passive design I was working on before I started on the OB actives.
If you like building speakers or just want to be one and done, a multi-amp DSP setup isn't a purchase, it's an investment that will save you time and money in the long run IMHO (if you know what you're doing).
Even if you don't know what you're doing the chances of coming away with something listenable are much better.
The replacement driver is on its way... how was my day with "just" one pair of Alpha 15's? Is was pleasantly surprised about the amount of basic. But it is loud! And kind of cold.... I have no experience or knowledge about minidsp and put in place the simplest of simplest settings, so I think there is a lot to gain.
Oh and they are large! I have seen a lot of pictures with this setup, but now standing in my modest living room they are huge!...haha
Oh and they are large! I have seen a lot of pictures with this setup, but now standing in my modest living room they are huge!...haha
