Hey Tyee. Very cool! You can start with my version, it sounds nice. John is on the road to Dallas at the moment. He has these speakers with him. I'll let you know what new crossover tricks he has up his sleeve when I get to Dallas. Give it about a week.
I think you'll really enjoy these. =)
I think you'll really enjoy these. =)
Servus and thanks a lot for sharing! I 've been hooked to my OB speaker too (Monacor DT-104 tweeter, Beyma 6,5" mid and 12" woofer).
Do you think your horn tweeter can play below 700Hz and still keep up with the woofer (highpass at about 55Hz)? I'm aiming for 100dB/1m. It would enable good integration with the 6,5" mid I have.
Last edited:
Hi All, I have been following this thread on and off for a while. I never had much interest in OB speakers, but to be fair, I never really listened to them. Well, this past weekend was the AK fest and I sat and listened to the Salk Audio OBs for a couple hours. Now I want to build a pair with stuff I have laying about. So, what are the critical specs I need to look for in the drivers? I have a lot of drivers and would rather go about this with some clarity of thought vs. just trying everything I have to see what works.
Edit: Here is what I have so far. The old but wonderful Philips tweeter. The back is enclosed so not really OB and I don't have specs on them.
The Parts Express PA130 5" midrange driver.... here is the spec sheet http://www.parts-express.com/pdf/295-010s.pdf
And this might work for the woof - CERWIN VEGA S-10.8 10" 8 OHM 400W STEALTH WOOFER
This is all I could find on them
• Woofer Size: 10".
• Frequency Response: 40Hz -500Hz.
• Power Handling: 200 watts RMS (400 watts Maximum).
• Sensitivity (2.83V @ 1 meter): 93dB.
• Impedance: 8 ohm.
• Mounting Diameter: 9.0" (22.86 cm.).
I assume that with OB you use normal crossover designs???
Cheers.
Edit: Here is what I have so far. The old but wonderful Philips tweeter. The back is enclosed so not really OB and I don't have specs on them.
The Parts Express PA130 5" midrange driver.... here is the spec sheet http://www.parts-express.com/pdf/295-010s.pdf
And this might work for the woof - CERWIN VEGA S-10.8 10" 8 OHM 400W STEALTH WOOFER
This is all I could find on them
• Woofer Size: 10".
• Frequency Response: 40Hz -500Hz.
• Power Handling: 200 watts RMS (400 watts Maximum).
• Sensitivity (2.83V @ 1 meter): 93dB.
• Impedance: 8 ohm.
• Mounting Diameter: 9.0" (22.86 cm.).
I assume that with OB you use normal crossover designs???
Cheers.
Last edited:
Open Baffle Crossovers
And a happy day to all you mothers out there.
OK, let's talk crossovers - those little circuits the fullrange guys love to hate.
But let's talk OB crossovers.
For the most part a crossover designed for a box speaker will not work well on an Open Baffle speaker. It took me almost 3 years to realize this, even with help and hints! So maybe this will get you there faster.
The main problem your going to deal with on Open Baffle is what I call RRoW "Rising Response of Woofers". It has to be dealt with. As most of you know, on an open baffle that sound from the back will wrap around the back - and vice versa. This becomes a problem at wavelengths longer than the baffle face. Why? Because these sound waves are out of phase and thus cancel each other out. So you don't get much bass. Midrange and high frequencies don't suffer this fate, thus they don't get attenuated.
So you end up with the woofers having a rising response as the frequency goes up because the bass is canceling out. The higher the frequency, the louder the woofer gets (up to a point). So the woofer ends up with a rising response toward the high end. Or a falling response toward the low end, if you look at it like that. Using a conventional crossover is going to leave you with far too much midrange from your the woofer and a thin, unhappy tonal balance. Rising Response of the Woofer
And a happy day to all you mothers out there.
OK, let's talk crossovers - those little circuits the fullrange guys love to hate.
But let's talk OB crossovers.
For the most part a crossover designed for a box speaker will not work well on an Open Baffle speaker. It took me almost 3 years to realize this, even with help and hints! So maybe this will get you there faster.
The main problem your going to deal with on Open Baffle is what I call RRoW "Rising Response of Woofers". It has to be dealt with. As most of you know, on an open baffle that sound from the back will wrap around the back - and vice versa. This becomes a problem at wavelengths longer than the baffle face. Why? Because these sound waves are out of phase and thus cancel each other out. So you don't get much bass. Midrange and high frequencies don't suffer this fate, thus they don't get attenuated.
So you end up with the woofers having a rising response as the frequency goes up because the bass is canceling out. The higher the frequency, the louder the woofer gets (up to a point). So the woofer ends up with a rising response toward the high end. Or a falling response toward the low end, if you look at it like that. Using a conventional crossover is going to leave you with far too much midrange from your the woofer and a thin, unhappy tonal balance. Rising Response of the Woofer
Open BAffle Crossovers - part Deux
How do we fix that? Easy - cross lower. Lower than you think you should. Much lower.
Let's have a look at the crossover in this project. Using the simple 1st order low pass on the woofer (1 coil) you'll notice that a value of approx 12mH is spec'd. That means a low pass of 100Hz. Huh? How can you crossover to a tweeter at 100Hz? Well - you don't. That's the electrical crossover point. But acoustically it's crossing over near 1100Hz. How is that possible? Remember the RRoW in the post above? The very low crossover point on the woofer fights that rising response and brings the woofer back to a flatter response. And acoustically the tweeter or mid can now come in at about 1100Hz.
Hey that was easy! Ah, but there are trade-offs my friends. As always. What is the trade-off in this approach? Efficiency. A lot of efficiency loss. With the wrap around bass canceling of the open baffle we have thrown away a lot of low end response resulting in the RRoW. To get the woofer back to a balanced response we now have to throw away a lot of midrange - about 10dB! Ouch. But that's the laws of physics. The low 100Hz electrical crossover does this, peels off all that extra midrange energy.
So this speaker becomes rather low efficiency in the trade. Now maybe scraping 81dB @ 1W.
You can see why we want to start out with an efficient woofer. And you can guess what the next step is. Yep, attenuate the tweeter. It needs to come down in sensitivity to match the woofer. Happily for us, it can use a fairly conventail high pass filter. In this case 2nd order HP circa 1100Hz with heavy attenuation.
And it works! Works well, in fact. Choosing the right drivers and getting the crossover right can make the speaker sound great. And this is a great sounding rig for a "Nail two drivers to a board and use 5 crossover parts."
How do we fix that? Easy - cross lower. Lower than you think you should. Much lower.
Let's have a look at the crossover in this project. Using the simple 1st order low pass on the woofer (1 coil) you'll notice that a value of approx 12mH is spec'd. That means a low pass of 100Hz. Huh? How can you crossover to a tweeter at 100Hz? Well - you don't. That's the electrical crossover point. But acoustically it's crossing over near 1100Hz. How is that possible? Remember the RRoW in the post above? The very low crossover point on the woofer fights that rising response and brings the woofer back to a flatter response. And acoustically the tweeter or mid can now come in at about 1100Hz.
Hey that was easy! Ah, but there are trade-offs my friends. As always. What is the trade-off in this approach? Efficiency. A lot of efficiency loss. With the wrap around bass canceling of the open baffle we have thrown away a lot of low end response resulting in the RRoW. To get the woofer back to a balanced response we now have to throw away a lot of midrange - about 10dB! Ouch. But that's the laws of physics. The low 100Hz electrical crossover does this, peels off all that extra midrange energy.
So this speaker becomes rather low efficiency in the trade. Now maybe scraping 81dB @ 1W.
You can see why we want to start out with an efficient woofer. And you can guess what the next step is. Yep, attenuate the tweeter. It needs to come down in sensitivity to match the woofer. Happily for us, it can use a fairly conventail high pass filter. In this case 2nd order HP circa 1100Hz with heavy attenuation.
And it works! Works well, in fact. Choosing the right drivers and getting the crossover right can make the speaker sound great. And this is a great sounding rig for a "Nail two drivers to a board and use 5 crossover parts."
Open Baffle Crossovers - summing up
The 2 posts above are the basics, of course there is more to it than that, and other ways it can be done. But that's the basic approach for this type of speaker. You have to throw away a lot of midrange energy from the woofer to get the tonal balance right. Many OB rigs I have heard obviously do NOT understand this. Easy to tell - no bass. Yeah, the bass is "clean" - but mostly because there just isn't any. Bass harmonics, sure, but the bass note fundamentals are not there.
Now you might be thinking - "I don't want to throw away all the efficiency, I'll go active, or use an active EQ." Bad news, it's the same thing. If you want to achieve the correct tonal balance by active means (shelving filter, EQ, etc.) you will simply end up throwing more power at the bass. Raise the line level of the bass and the amp is also going to output more power relative to the mids. End result? Same as the passive filter. Nothing wrong with that, but just keep it in mind. You can't beat the physics.
Or can you? That's were we get into the high Qts woofers. Woofers with a Q >1.0 on baffle. In a box or on an infinite baffle these woofers have a big low end peak. But put them on an open baffle and much of that peak gets cancled out. But since the woofer is starting with that peak, it still has more bass left after the wrap around cancellation. So on the right baffle the woofer can have a response closer to flat. But that's for another thread.
Remember, the remarks above are a general guide, not a detailed technical guide to open baffles or crossovers. There will be nits to pick, but understanding the above will take you a looong way toward building a good open baffle crossover. Remember the RRoW!
The 2 posts above are the basics, of course there is more to it than that, and other ways it can be done. But that's the basic approach for this type of speaker. You have to throw away a lot of midrange energy from the woofer to get the tonal balance right. Many OB rigs I have heard obviously do NOT understand this. Easy to tell - no bass. Yeah, the bass is "clean" - but mostly because there just isn't any. Bass harmonics, sure, but the bass note fundamentals are not there.
Now you might be thinking - "I don't want to throw away all the efficiency, I'll go active, or use an active EQ." Bad news, it's the same thing. If you want to achieve the correct tonal balance by active means (shelving filter, EQ, etc.) you will simply end up throwing more power at the bass. Raise the line level of the bass and the amp is also going to output more power relative to the mids. End result? Same as the passive filter. Nothing wrong with that, but just keep it in mind. You can't beat the physics.
Or can you? That's were we get into the high Qts woofers. Woofers with a Q >1.0 on baffle. In a box or on an infinite baffle these woofers have a big low end peak. But put them on an open baffle and much of that peak gets cancled out. But since the woofer is starting with that peak, it still has more bass left after the wrap around cancellation. So on the right baffle the woofer can have a response closer to flat. But that's for another thread.
Remember, the remarks above are a general guide, not a detailed technical guide to open baffles or crossovers. There will be nits to pick, but understanding the above will take you a looong way toward building a good open baffle crossover. Remember the RRoW!
Well, that makes a lot of sense. One thing you said makes me feel a little better and that is that the mids and tweets are not as effected by the OB. Is that right? So the crossover for the mids and tweets can be the normal stuff? It is just bringing up the lower woofer to level what the RROW took away?
Also, maybe I did not read this thread right and I don't want to highjack it, but is it ok to talk about other OB designs here or should I start another thread? Cheers.
Also, maybe I did not read this thread right and I don't want to highjack it, but is it ok to talk about other OB designs here or should I start another thread? Cheers.
I'd say start another tread if it's another project.
Correct on the mids and tweets. It's a wavelength vs baffle width thing. If the drivers are open back, you'll often get even more energy than you expect in the room because the back wave isn't thrown away like in a box speaker. Some of that back energy splashes around the room and you'll hear it.
That's where the "art" in crossover design starts to come in. Getting all that balanced
Correct on the mids and tweets. It's a wavelength vs baffle width thing. If the drivers are open back, you'll often get even more energy than you expect in the room because the back wave isn't thrown away like in a box speaker. Some of that back energy splashes around the room and you'll hear it.
That's where the "art" in crossover design starts to come in. Getting all that balanced
Nice introduction to passive OB's crossover Pano
An illustration of john's crossover with its transfer function.
-3db points are 100Hz for the woofer and 1k for the tweeter.
The tweeter seems to be cross as "normal stuff". But there is some others troubles in the midrange because of dipole peak and dips.
An illustration of john's crossover with its transfer function.
-3db points are 100Hz for the woofer and 1k for the tweeter.
The tweeter seems to be cross as "normal stuff". But there is some others troubles in the midrange because of dipole peak and dips.
Attachments
Merci Jerome! And thanks for the plot.
As you can see, it's a sort of LP shelving filter on the woofer and the tweeter gets a little boost at the top.
Below is John's version with the larger inductor and a trap on the tweeter. Taken from actual on-baffle impedance curves. Remember, this is the crossover's filter function, not the frequency response.
As you can see, it's a sort of LP shelving filter on the woofer and the tweeter gets a little boost at the top.
Below is John's version with the larger inductor and a trap on the tweeter. Taken from actual on-baffle impedance curves. Remember, this is the crossover's filter function, not the frequency response.
Attachments
I´ve just finnished reading this thread, and it seams the big problem with the low xo for the tweeter is due too the radiating area vs. xmax ,beaming and distortion right?
I think the d3806 is too exspencive for a project made too be simple and cheap.
Here is my idea; what about using 2 cheaper tweeters on each speaker?
Twice the cone area, higher senseitivity and as a result lower distortion.
Or, mabe Im very wrong....
I think the d3806 is too exspencive for a project made too be simple and cheap.
Here is my idea; what about using 2 cheaper tweeters on each speaker?
Twice the cone area, higher senseitivity and as a result lower distortion.
Or, mabe Im very wrong....
.........Or, mabe Im very wrong....
Yes.
Assuming the diameter of the tweeter face plate is 4 inches, The center to center distance cannot be less than 4 inches between the domes. That's about 1 wavelength apart at about 3.6 Khz. So you will get phase cancellation as the frequency goes up right where it ruins everything.
Better to use a single tweeter which can handle the signal. Maybe use a 4th order LR crossover and push it's crossover frequency a tad higher . But then it also depends on the behaviour of the bass/mid driver.
Yes, sadly i know there will be phase issues. the 3,4khz will appear at 90* off axis paralell too the tweeter arrangement. So, if they are vertically aligned the horisontal axis response should remain the same, while the vertical response will have different "out-phasing" trough the entire range down too 3,4khz at +-90*(6,8khz@45* and so on...).
Anyways, its probably better too go threeway than 2 tweeters...
What about making a waveguide thats "tuned" at the tweeter f.res? supressing the f.res mechanically instead of electric? Just throwing some ideas out there
Anyways, its probably better too go threeway than 2 tweeters...
What about making a waveguide thats "tuned" at the tweeter f.res? supressing the f.res mechanically instead of electric? Just throwing some ideas out there
I think a Dayton RS52 dome with one of the tiny Dayton tweeters could work better, and be cheaper, than the D38. Something similar to the M&T part of Zaph's ZDT3.5 perhaps?
Zaph|Audio - ZDT3.5
Zaph|Audio - ZDT3.5
Hello All…
When I originally designed the Manzanita, it was an experiment about determining how small and simple a musical OB could be. A small, (by OB standards) practical stand mounted OB that could really play loud enough to be usable into the low 40’s, assuming a typical residential room. I set the baffle size upper limit to that of the once popular large stand mounted two & three way boxes from the late sixties through the early 80’s. Think Large Advent, AR3 and JBL L-100. This limited the woofer size at no more than 12” and baffle width around 14”-15”. Height less than 28”. Depth less than 12”.
The Peerless woofer (830669) proved to be up to the task. But, for cost and simplicity sake, the TDFC ended up being the big compromise of the original design. At low to moderate levels in a small room it does surprisingly well. However, it was obvious that a tweeter that could play lower and louder was needed. Enter the D3800. Not that long ago the D3800 was available for something like $65 wholesale, which made it practical. Since then, as we all know, the price has sky rocketed.
I have tried to find a suitable single driver replacement for the D3800. One that retained the original concept goals of simplicity, performance and cost. Basically, a very low distortion midrange that really can provide some “air” up top. And, has the ability to play loud enough to keep up with the Peerless.
One candidate was the RS52 with a trap to tame the 13 KHz break up peak and some pad bypass to try to get just a little more extension up high. Actually, the RS 52 can work pretty well, better on the low end than the D3800, but not as well up top. And as it is acknowledged here, a dual tweeter set up just creates more problems and complexity than it is worth.
Most of my experiments have been with various small 2”-4” wide range and or full range cone drivers. So far, unfortunately, none have proven to equal or exceed the D3800. There are new drivers coming on the market all the time. Sooner or later one with a reasonable price with the required performance qualities will come along.
I have made several 3 way versions, retaining the great Peerless and using both dome and cone midranges and various tweeters. The better configurations are much improved over the D3800 based two way. But, you have a much more complicated speaker and higher costs of course.
Like many here, I will be watching to see if anyone finds the magic driver that for a few dollars can best the D3800. And if I find that driver, I will let you know. John
Select a Folder Deleted Items Drafts Sent Items
When I originally designed the Manzanita, it was an experiment about determining how small and simple a musical OB could be. A small, (by OB standards) practical stand mounted OB that could really play loud enough to be usable into the low 40’s, assuming a typical residential room. I set the baffle size upper limit to that of the once popular large stand mounted two & three way boxes from the late sixties through the early 80’s. Think Large Advent, AR3 and JBL L-100. This limited the woofer size at no more than 12” and baffle width around 14”-15”. Height less than 28”. Depth less than 12”.
The Peerless woofer (830669) proved to be up to the task. But, for cost and simplicity sake, the TDFC ended up being the big compromise of the original design. At low to moderate levels in a small room it does surprisingly well. However, it was obvious that a tweeter that could play lower and louder was needed. Enter the D3800. Not that long ago the D3800 was available for something like $65 wholesale, which made it practical. Since then, as we all know, the price has sky rocketed.
I have tried to find a suitable single driver replacement for the D3800. One that retained the original concept goals of simplicity, performance and cost. Basically, a very low distortion midrange that really can provide some “air” up top. And, has the ability to play loud enough to keep up with the Peerless.
One candidate was the RS52 with a trap to tame the 13 KHz break up peak and some pad bypass to try to get just a little more extension up high. Actually, the RS 52 can work pretty well, better on the low end than the D3800, but not as well up top. And as it is acknowledged here, a dual tweeter set up just creates more problems and complexity than it is worth.
Most of my experiments have been with various small 2”-4” wide range and or full range cone drivers. So far, unfortunately, none have proven to equal or exceed the D3800. There are new drivers coming on the market all the time. Sooner or later one with a reasonable price with the required performance qualities will come along.
I have made several 3 way versions, retaining the great Peerless and using both dome and cone midranges and various tweeters. The better configurations are much improved over the D3800 based two way. But, you have a much more complicated speaker and higher costs of course.
Like many here, I will be watching to see if anyone finds the magic driver that for a few dollars can best the D3800. And if I find that driver, I will let you know. John
Select a Folder Deleted Items Drafts Sent Items