Jahuzi, what would be really interesting is to combine the Neo3 (placed vetically like in your aino gradient) with the Audax HM100Z0. Crossover probably at 2,5khz. You shuold ge a much more better vertical pattern and horizontally it should be similar. Did your have tried this in the past?
I wonder how to estimate the vertical response of the neo 8 or 10....
If ust the woofer tweeter combination in xdir i get a very narrow vertical pattern. But the neo 8,10 is more like a small array, isn't it? So how will the vertcial pattern be for a combination with neo 3?
Hi CharlieLaub,
AinoGradient is crossed @2,5-4khz. If i could cross @1,5khz or something like this, i could probably use your recommended driver. Do you have an idea hot it compares to Neo8?
If i understood you right, then the resonance @800hz is not seen by the sb17cac....
First conclusion!
The best tweeter solution in my eyes are:
- Peerless OC25SC
*perfect vertical and horizontal
*special baffle shape (https://www.diyaudio.com/forums/multi-way/161299-directivity-dipole-tweeters-28.html#post6508139)
*crossover at 3-4khz, so a 2" to 3" inch is needed, if we want a good backside radiation pattern (https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270)
* If the high mid driver are round we cam put 4 of them compact together around the OC25SC.
BG Neo3 (os similar...)
*perfect horizontal (possibly with felt pads)
*Crossover at 2-3 khz, so we could cross to some well known 4"(https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270)
*probably easier for passive crossovers
* If 3 Way probably easier
I will make some simulation tests with virtuixcad with drivers from this thread:
https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270
1.
OC25SC with 4 FRS5X.
(Unfortunatly, i have no backsider radiation patter for 3"drivers that could be superior)
2.Neo3 with 4 inch or Neo8
(probably i will look again on the 6,5"of CharlieLaub)
Using Virtuixcad i will take a look at these points:
Use of Duelund Filter
Dispersion
Distortion
Max Spl
Any other suggestion?
Best regards, Tomas
The best tweeter solution in my eyes are:
- Peerless OC25SC
*perfect vertical and horizontal
*special baffle shape (https://www.diyaudio.com/forums/multi-way/161299-directivity-dipole-tweeters-28.html#post6508139)
*crossover at 3-4khz, so a 2" to 3" inch is needed, if we want a good backside radiation pattern (https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270)
* If the high mid driver are round we cam put 4 of them compact together around the OC25SC.
BG Neo3 (os similar...)
*perfect horizontal (possibly with felt pads)
*Crossover at 2-3 khz, so we could cross to some well known 4"(https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270)
*probably easier for passive crossovers
* If 3 Way probably easier
I will make some simulation tests with virtuixcad with drivers from this thread:
https://www.diyaudio.com/forums/multi-way/366758-midwoofer-dipol-configuration.html#post6505270
1.
OC25SC with 4 FRS5X.
(Unfortunatly, i have no backsider radiation patter for 3"drivers that could be superior)
2.Neo3 with 4 inch or Neo8
(probably i will look again on the 6,5"of CharlieLaub)
Using Virtuixcad i will take a look at these points:
Use of Duelund Filter
Dispersion
Distortion
Max Spl
Any other suggestion?
Best regards, Tomas
So, came to this conclusion:
OC25SC witfh FRS5x (or similar) would be a very nice combination, but not possible as 3-Way. At least i found no way to cross at 600Hz with a duelund filter. We would need 4 drivers to get a good filternetwork with good dispersion.
Any other suggestion about the above combination?
That means, for a good Tweeter in a 3 way i only see the Neo3 (or similar) as a possibility.
I now need a suitable middriver solution between the Neo3 (or GRS3) and my SPP-250 combination.
The conditions:
Best frequency range from 600Hz-1600Hz. If i could use it from 300HZ it would be better. It should be widerange because i need acoustically LR2 for a duelund filter. Interestingly the best drivers i have found are used in some Open Baffle-Designs:
4" Peerless HDS PPB 4-25/08 (was used from rudolph with neo 3 and two 10" as sub)
as 4' alternativ i have found the Dayton rs 125.
6,5" SB17CAC (used from CharlieLaub)
BG NEO 10 (used in several other projects)
The Neo 8 seems not to be suitable with acoustically LR2 at 600Hz.
Of course, if the driver is smaller, then the dispersion will be better. The goal is to find the sweet spot 🙂. Any suggestion?
OC25SC witfh FRS5x (or similar) would be a very nice combination, but not possible as 3-Way. At least i found no way to cross at 600Hz with a duelund filter. We would need 4 drivers to get a good filternetwork with good dispersion.
Any other suggestion about the above combination?
That means, for a good Tweeter in a 3 way i only see the Neo3 (or similar) as a possibility.
I now need a suitable middriver solution between the Neo3 (or GRS3) and my SPP-250 combination.
The conditions:
Best frequency range from 600Hz-1600Hz. If i could use it from 300HZ it would be better. It should be widerange because i need acoustically LR2 for a duelund filter. Interestingly the best drivers i have found are used in some Open Baffle-Designs:
4" Peerless HDS PPB 4-25/08 (was used from rudolph with neo 3 and two 10" as sub)
as 4' alternativ i have found the Dayton rs 125.
6,5" SB17CAC (used from CharlieLaub)
BG NEO 10 (used in several other projects)
The Neo 8 seems not to be suitable with acoustically LR2 at 600Hz.
Of course, if the driver is smaller, then the dispersion will be better. The goal is to find the sweet spot 🙂. Any suggestion?
After testing a bunch of drivers for OB use in the last few years, I have come to the conclusion that if you want to use a dipolar tweeter (and you should) then you will be limited to a crossover point of not less than about 2kHz, or possibly even slightly higher. This has to do with the fact that distortion is rising above 1 percent at modest playback levels for many dipolar tweeters starting around 2kHz and going lower in frequency. It's the distortion performance, and not the supported frequency response, that really limit how low in frequency you can use a tweeter. It's the same whether you are building a dipole or monopole speaker.
Given that requirement/limitation, you need to identify a driver for the next lower band that will operate as a nice dipolar source up to 2kHz or higher, so that it can be crossed over in that vicinity to the tweeter, without any response hiccups to the front, or rear.
For a given crossover point, do an analysis on the driver's capability to operate that low as a dipole, especially if it is a planar type (e.g. B&G NeoX or new GRS clones), and also how much power is needed to EQ up the dipole losses compared to the power rating of the driver, its Xmax capabilities, and how its distortion will increase with power input. The higher the voltage sensitivity, the less power you need to achieve a given SPL, so this is definitely something to consider for the critical midrange band. Below that things get easier in terms of drivers that will work as dipoles, since wavelengths get longer. The number of bands will depend on how loud you need/want the system to operate and how much baffle loading you desire to use.
For the mid and tweeter minimal to no baffle is often the way to go IMO but the size of the baffle around the driver can steadily increase as you go lower in frequency. But since you generally want to keep drivers near each other on the baffle, it is not easy to transition from one "size" baffle to another when drivers sit right next to each other. This is one reason, in the end, I just throw up my arms and try to use no baffle for all drivers, and design the system and choose drivers with that goal in mind.
Given that requirement/limitation, you need to identify a driver for the next lower band that will operate as a nice dipolar source up to 2kHz or higher, so that it can be crossed over in that vicinity to the tweeter, without any response hiccups to the front, or rear.
For a given crossover point, do an analysis on the driver's capability to operate that low as a dipole, especially if it is a planar type (e.g. B&G NeoX or new GRS clones), and also how much power is needed to EQ up the dipole losses compared to the power rating of the driver, its Xmax capabilities, and how its distortion will increase with power input. The higher the voltage sensitivity, the less power you need to achieve a given SPL, so this is definitely something to consider for the critical midrange band. Below that things get easier in terms of drivers that will work as dipoles, since wavelengths get longer. The number of bands will depend on how loud you need/want the system to operate and how much baffle loading you desire to use.
For the mid and tweeter minimal to no baffle is often the way to go IMO but the size of the baffle around the driver can steadily increase as you go lower in frequency. But since you generally want to keep drivers near each other on the baffle, it is not easy to transition from one "size" baffle to another when drivers sit right next to each other. This is one reason, in the end, I just throw up my arms and try to use no baffle for all drivers, and design the system and choose drivers with that goal in mind.
Just letting you know there is a BG "Neo8" and also an improved "Neo8S" which has a stronger magnet and better LF performance.
Also, this paper is an excellent overview of passive open-baffle design:
http://www.dipolplus.de/downloads/Open baffle 1.pdf
IMO if you want to make a true dipole system with passive XO you are hobbled out of the gate. The ability to use low-end roll-off compensation EQ, steep XO slopes, and time delay with DSP is going to help immensely to extract the best performance from each driver inside its passband. The best dipole designs operate on very slim margins of excursion, distortion and directivity.
That being said, if you follow the design guidelines many have offered in this thread, you can make a good sounding passive open baffle speaker. It probably won't be a "dipole" however.
Also, this paper is an excellent overview of passive open-baffle design:
http://www.dipolplus.de/downloads/Open baffle 1.pdf
IMO if you want to make a true dipole system with passive XO you are hobbled out of the gate. The ability to use low-end roll-off compensation EQ, steep XO slopes, and time delay with DSP is going to help immensely to extract the best performance from each driver inside its passband. The best dipole designs operate on very slim margins of excursion, distortion and directivity.
That being said, if you follow the design guidelines many have offered in this thread, you can make a good sounding passive open baffle speaker. It probably won't be a "dipole" however.
There are people building very successful open baffle speakers these days using a single full range driver. And that driver is obviously covering a lot more than just 2 to 3 octaves.
Here are links to a few examples:
DECWARE Caintuck Audio Open Baffle Loudspeaker
DECWARE Zen Master Series Loudspeakers
DECWARE Zen Master Series Loudspeakera
Not sure that you really need all those multiple drivers with the crossovers and other complexity issues that come with them to get very good OB performance. It might be a lot simpler than that now.
Last edited:
And don't let the prices shown in the links above concern you. Those are for completely finished speakers from Decware.
You can build the same thing yourself for a fraction of the price by buying the drivers directly from Lii Audio. Very simple build as well.
You can build the same thing yourself for a fraction of the price by buying the drivers directly from Lii Audio. Very simple build as well.
Sorry that train passed long time ago! Audaxes are used in another project!
Avalanche AS1 modernization
Vertical response is ok in +/- 5 deg window
Shared album - Juha Sirkka - Google Photos
Regarding those small Audaxes, I have used the HM130C0 (carbon fiber cone) in a past project, and this does work well in an OB/nude application up to 2.5kHz. Give it a look.
Sure, but what you call "successful" I call "full of built-in flaws". To each their own. To be completely honest, IMO a fullrange driver used in an OB application is a bad idea. What's happening off axis or to the rear, I wonder? They don't want to know or don't care. There are lots of speakers that can be designed in this way and will sound "good" to someone, who will then write about it. Welcome to our subjective audio hobby, folks.
IMHO...
Some say if you cross to a different driver anywhere above 1K or below 8k, you're building in a flaw also.
I've been listening to speakers designed this way my whole life, because it's a prevalent as a vehicle with 4 wheels... It's simply what I've heard, 99% of the time
The Altec Valencia's were the only 2-way speakers I've ever owned that didnt do that. I always liked them even with the squawky horns folks would laugh about "how bad they sound" in early forums, but no, went to the polk SDA-1 next with the dome tweeters crossed you know where.
I once owned Acoustats that didnt do that. I wish I threw the anchor down with those, but got rid of them too - think toddler with hardly protected large caps charged to 4kV set behind each one - 20 years ago. Back to another 'round of high cross multi-ways, never liking any of them "for some reason", completely lost in the speaker sauce again.
Clearly, this single aspect had dogged me my whole audio life. For my remaining time that I have ears that can still hear unaided, I'll remember and pay attention to that one.
Some say if you cross to a different driver anywhere above 1K or below 8k, you're building in a flaw also.
I've been listening to speakers designed this way my whole life, because it's a prevalent as a vehicle with 4 wheels... It's simply what I've heard, 99% of the time
The Altec Valencia's were the only 2-way speakers I've ever owned that didnt do that. I always liked them even with the squawky horns folks would laugh about "how bad they sound" in early forums, but no, went to the polk SDA-1 next with the dome tweeters crossed you know where.
I once owned Acoustats that didnt do that. I wish I threw the anchor down with those, but got rid of them too - think toddler with hardly protected large caps charged to 4kV set behind each one - 20 years ago. Back to another 'round of high cross multi-ways, never liking any of them "for some reason", completely lost in the speaker sauce again.
Clearly, this single aspect had dogged me my whole audio life. For my remaining time that I have ears that can still hear unaided, I'll remember and pay attention to that one.
Last edited:
The vaunted "keep out" crossover avoidance frequency range that I recall is 300-3kHz. This is where the ear is most sensitive. Funny that loudspeakers often cross over around 2kHz or 700Hz, and it doesn't seem to make people scream bloody murder. I think in the past when crossover design was more of an "art" and less was understood about loudspeaker performance in general, it was a more useful rule of thumb. But a well designed and implemented system can cross over anywhere IMHO without any sonic detriment. It's all in the implementation.
I really have no idea what loudspeakers you have listened to, but I don't find that any of my own experiences back up your claims.
I really have no idea what loudspeakers you have listened to, but I don't find that any of my own experiences back up your claims.
OBs have a lot of the advantages of dipoles. Dipoles have the advantage of constant directivity, anything else?
So here is the way I look at it. If someone wants to try open baffle they can go one of two significantly different ways.
They can attempt a complex 3-way design and go through lots of decision choices for drivers, crossovers, etc, not really knowing which is best. It can be very time consuming with still no perfect answers. Just look at all of the posts in this thread so far going back and forth on what might work and what might not in order to see what I’m talking about.
Or they can take a sheet of plywood, cut a singe large hole in it, and mount a full range driver specifically designed for OB with the right set of T/S parameters including a high Qts value. For about $400 they can quickly and easily have what is likely to be a very good OB speaker.
Now how will it sound compared to a 3-way? I don’t know. That depends a lot on how good the person is at designing both crossovers needed for the 3-way. Without a lot of prior experience that crossover design could become the weak link of the whole enterprise.
Either way you really won’t know what the performance is until you build it and try it in your listening room. But one approach certainly is a lot easier and less expensive than the other. And might be just as good. Or maybe even better. Who knows?
They can attempt a complex 3-way design and go through lots of decision choices for drivers, crossovers, etc, not really knowing which is best. It can be very time consuming with still no perfect answers. Just look at all of the posts in this thread so far going back and forth on what might work and what might not in order to see what I’m talking about.
Or they can take a sheet of plywood, cut a singe large hole in it, and mount a full range driver specifically designed for OB with the right set of T/S parameters including a high Qts value. For about $400 they can quickly and easily have what is likely to be a very good OB speaker.
Now how will it sound compared to a 3-way? I don’t know. That depends a lot on how good the person is at designing both crossovers needed for the 3-way. Without a lot of prior experience that crossover design could become the weak link of the whole enterprise.
Either way you really won’t know what the performance is until you build it and try it in your listening room. But one approach certainly is a lot easier and less expensive than the other. And might be just as good. Or maybe even better. Who knows?
It would be interesting to see independent measurements of a Decware speaker, including off-axis and distortion! The driver is special and of work of art as fullrange. The fullrange OB's that I've heard missed low bass and had wacky and higly directive treble. But being point source without xo has benefits yes.
We are at diyaudio.com, and people around here are interested in this hobby and mostly ready to give time and resources needed for multi-way speakers. Most have different kind of speakers in different setups, like me. In home theater room I have 2-way coaxials as LCR and they sound good too, but totally different in that totally different room.
We are at diyaudio.com, and people around here are interested in this hobby and mostly ready to give time and resources needed for multi-way speakers. Most have different kind of speakers in different setups, like me. In home theater room I have 2-way coaxials as LCR and they sound good too, but totally different in that totally different room.
I agree with most of what you are saying, but have a different perspective on the idea that everyone in this hobby has a preference for multi-way speakers.
I think the goal for most people here is to obtain the very highest sound quality possible for the their room and music listening tastes. And to do so within each person's available budget.
The answer might be a multi-way solution or might not. There is certainly a great deal of interest in full range as evidenced by that other nearby forum.
Really the only reason I introduced the single driver full range idea here was due to the initial post where it was suggested that a single drive could not cover more than 2 to 3 octaves. This was to show that, in fact, a single drive can cover considerably more range. Actually as much as many multi-way designs.
Last edited:
We hear the direct sound first then the sound from the rear reflected, like an ambience speaker. How important is the spectral content in that case?