2-way active - Looking for peer critique

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,

I opened this thread for anyone to have a look at my design choices and help me in the process in designing my first speaker. Any tips, critique and support is welcome as I am new to this and learning every day.

The monitor should have the following design parameters:

  • Usable for near-field monitoring (1-2 meters)
  • Constant directivity
  • Active XO + bi-amping
  • Low XO point (~1,5 khz)
  • Small woofer-to-tweeter distance
  • Sealed enclosure, for easy integration with subs later on and great transient respons
  • F3 around 60-70 hz
  • Enclosure made out of 25 mm MDF or HDF
  • Budget total max €1000 for the pair
To achieve these goals I have come to the conclusion I want to use these parts:

  • Tweeter: SB Acoustics SB26ADC + 5" elliptical 3D-printed WG made by Augerpro. (can be found here Open source Waveguides for CNC & 3D printing!)
  • Woofer: still unsure so advice is welcome. Right now choice is between Dayton RS180-4 (great for sealed but high 3rd harmonic distortion) or the SB Acoustics SB17NBAC-4 (not great for sealed and f3 goal, but great other parameters.)
  • Hypex FA122 plate amp for digital XO and amp.
  • Enclosure volume ~16L (could change depending on woofer choice).
  • For measuring: a MiniDSP UMIK-1.
For simulating the box enclosure I use UniBox and BoxSim 2.0.

So right now my progress is slowed down by not being able to find the right woofer for my design goals.

After this I would need to know how to properly measure my speaker and set the perfect XO-point. I know there are multiple software-programmes available for this: REW, RePhase and ARTA. I am not familiair with this so I will have to dive deeper into this. Any practical guides or tutorials are welcome.

Hopefully some of you are willing to help. Thanks in advance :).

I will keep this thread updated when progress is being made.
 
Last edited:
The SB17NBAC-4 does not start beaming from the 180-degree baffle-controlled polar directivity until above your 1.5kHz crossover, so a tweeter waveguide would create an abrupt change in polar directivity around the crossover.
====
With DSP time alignment, you could mount the 1.5kHz crossover tweeter for the closest center-to-center spacing on a flat baffle and get baffle-controlled smooth constant 180-degree directivity over the crossover, with just the normal SB26ADC tweeter beaming starting aroung 4kHz. Smooth controlled directivity. To reduce baffle edge distortion, construct a proven low distortion baffle with bevels or rounds. I think this is the best DSP supported design for a 1.5kHz crossover
=============
====
TYPICAL: Controlled Directivity Waveguide tweeter
1) find a midrange with a smooth high frequency, and modest cone break-up artifacts
2) set crossover frequency to where the midrange beaming matches the waveguide directivity
3) design a crossover steep enough to attenuate midrange cone breakup

For the SB17NBAC-4 midrange, if you moved the crossover to 2.0 - 2.2kHz where some beaming narrows the polar pattern to match the polar pattern of the tweeter waveguide, you could maintain smooth controlled directivity. BUT you would require a steep crossover to attenuate SB17NBAC-4 cone breakup.
====

Buchardt Audio uses the SB17NBAC with a round waveguide with a 2,000Hz Crossover on model BA S400. The depth of the waveguide is set to match the tweeter-midrange physical offset with a passive xover.
s400 detailed description — Buchardt Audio

Audioengine builds monitors similar to your goals
Speakers — AudioengineAudioengine
 

Attachments

  • BA_S400_1.jpg
    BA_S400_1.jpg
    92.3 KB · Views: 553
  • AudioEngine 5.jpg
    AudioEngine 5.jpg
    88.7 KB · Views: 556
  • example baffles.jpg
    example baffles.jpg
    149 KB · Views: 553
  • Compas Bevels.jpg
    Compas Bevels.jpg
    114.4 KB · Views: 548
Thanks for your reply Linesource.



If I understand correctly, the crossover point should be where both the woofer and the tweeter + WG drop at the same frequency, under the same angle. Should it be -6 dB at 60 degrees?
So of I were to crossover at 1,5 khz both the woofer and tweeter + WG should have their - 6 at this frequency. If so, this would be hard to achieve with a 6-7 inch woofer (haven't found one that suits these criteria). I would have to push the XO-frequency higher, or I would have to choose a bigger woofer, probably an 8 inch.
Please correct me if I am wrong.
 
Member
Joined 2005
Paid Member
What is the reason you want constant directivity / waveguide for a near-field monitor? With a listening distance of 1-2 m, you will mainly hear the direct sound from the drivers, and not so much the sound echoing back from the room. I'd be surprised if the benefits of the waveguide / constant-directivity approach would work out well in this scenario.
 
Excuses me, trying to search the right words.
I mean the rollof in dB under horizontal and vertical angles. As I learned when the directivity isn't controlled, resulting in peaking frequencys under angles, will make the sound different than when the roll-of is smooth. I learned this from the studys executed by Prof. Toole.
 
Hi all,

  • Usable for near-field monitoring (1-2 meters)
  • Constant directivity
You meant controlled directivity, not constant. If constant, specify value and frequency range.
  • Active XO + bi-amping
  • Low XO point (~1,5 khz)
  • Small woofer-to-tweeter distance
  • Sealed enclosure, for easy integration with subs later on and great transient respons
  • F3 around 60-70 hz
  • Enclosure made out of 25 mm MDF or HDF
No. 16 mm with dense reinforcements is better from gross volume and weight consideration. You want the thing to be compact. Thinner MDF tends to be stronger per slice area as well.
  • Budget total max €1000 for the pair
Budget is too low.
  • Tweeter: SB Acoustics SB26ADC + 5" elliptical 3D-printed WG made by Augerpro. (can be found here Open source Waveguides for CNC & 3D printing!)
  • Woofer: still unsure so advice is welcome. Right now choice is between Dayton RS180-4 (great for sealed but high 3rd harmonic distortion) or the SB Acoustics SB17NBAC-4 (not great for sealed and f3 goal, but great other parameters.)
Woofer's choice is wrong from low bass efficiency in small enclosure. After LT linearization to required extension it will run ot of rated power or driving voltage range. Search for woofers with decent power rating and with Bl^2/Re value higher than 12, don't affraid elevated Mms.
  • Hypex FA122 plate amp for digital XO and amp.
  • Enclosure volume ~16L (could change depending on woofer choice).
  • For measuring: a MiniDSP UMIK-1

  • For simulating the box enclosure I use UniBox and BoxSim 2.0.
Better use VituixCAD 2, it is a real design-devil for free.
 
You meant controlled directivity, not constant. If constant, specify value and frequency range.
No. 16 mm with dense reinforcements is better from gross volume and weight consideration. You want the thing to be compact. Thinner MDF tends to be stronger per slice area as well.
Budget is too low.
Woofer's choice is wrong from low bass efficiency in small enclosure. After LT linearization to required extension it will run ot of rated power or driving voltage range. Search for woofers with decent power rating and with Bl^2/Re value higher than 12, don't affraid elevated Mms.
Better use VituixCAD 2, it is a real design-devil for free.
Thanks for the feedback.

What are the specific specifications I should prioritse?
I know EBP should be favorable <80, VAS and FS should be relatively low (<25L and <40hz)
But where do I go from there?

I (think) I'd rather have a aluminium woofer because of the CSD is better than most paper woofers and helps against the back wave, and also because I want to use an aluminium tweeter.
 
Becouse of using DSP your powers unleashed. You don't need to stick to '70s T-S theory anymore. Your main points of consideration are:


- large Bl^2/Re
- large Sd * Xmas
- not too low Mms
- high long-term power rating
- stiff diaphragm material

- clever DSP work
- low and linear Le*

First parameter defines ability of the woofer to work efficiently in small (relative to Sd * Xmax) enclosure, second gives you objective information about bass capability of the woofer, third is necessary for bass efficiency maximization of combined enclosure-woofer-amplifier-DSP electromechanical system, fourth is necessary for low thermal compression and reliable operation in small enclosure after bass equalization, fifth is for durability and pistonic motion in as broad frequency range as possible, the last is a necessary binding factor for all the above aspects. The most important is Linkwitz-transformation a-ka bass equalization.

Your target of using metal drivers is right, sound of modern transducers under DSP equalization is counterintuitive to material's physical properties and other beliefs. You can focus on aesthetic and functional sides of the project and choose cabinet volume in a broad range of available alignments without much bothering (in acceptable range of woofer rated power and amplifier driving voltage).

*for midrange clarity.
In a last word - bass qualities from DSP sealed boxes (that's why I started from this) combined with midrange qualities (diaphragm, Le properities) leads to unusually low IMD content in the acoustic signal and to great quality of sound reproduction.
 
Last edited:
The SB17NBAC-4 does not start beaming from the 180-degree baffle-controlled polar directivity until above your 1.5kHz crossover, so a tweeter waveguide would create an abrupt change in polar directivity around the crossover.

With a 5" waveguide though it shouldn't really be affecting directivity down at 1.5kHz and have similarly wide dispersion to the sb17.

The nice thing about the sb17 is that although it breaks up like a classic metal cone it doesn't have any of the usually associated distortion products. A 2.5khz cover to a 5" waveguide would work fine. As should an xover at 1.5kHz.
 
Consider a coaxial driver?

If you will be using subs and/or you don't need much below 100 Hz or so, why not consider a smaller (5") coax driver? These can be of high quality (and price!) like BMS, or ....? You get slightly better alignment (2 dimensions) instead of just one dimension (height) with a "normal" baffle. Time alignment should be easier with a coax. The downside (there always are trade-offs) is .... ?
 
Aha, that's very helpfull.
I see to recquire a BL^2/Re of 12 I'd need a 8 inch woofer, from a quick look around.
I wonder though, because I want to cross them over to subs at 80-90hz, if a 6-7" would be sufficient anyways.
I worry about driver to driver distance and also the size of the speaker would be getting pretty wide. I acknowledge that maybe I'm asking too much from a small speaker, but wonder where the practical minimal requirements are for a f3 of 70 Hz from a relatively small sealed speaker.

Otherwise I might have to think about dropping the idea of a "full-size" WG and go for the Seas DXT tweeter to maintain the smallest driver to driver distance.
 
If you will be using subs and/or you don't need much below 100 Hz or so, why not consider a smaller (5") coax driver? These can be of high quality (and price!) like BMS, or ....? You get slightly better alignment (2 dimensions) instead of just one dimension (height) with a "normal" baffle. Time alignment should be easier with a coax. The downside (there always are trade-offs) is .... ?
Ah just read your reply after posting my reply above.
As I said, f3 of 70 Hz to cross to subs at 80-90 hz is my plan.

A coax would be a nice idea too, especially for near-field. But I'm even less knowledgeable about good coax drivers. I know Seas has a great one, but is too expensive for me. Thought about using a fullranger as well, but worry about high frequency response.
 
With a 5" waveguide though it shouldn't really be affecting directivity down at 1.5kHz and have similarly wide dispersion to the sb17.

The nice thing about the sb17 is that although it breaks up like a classic metal cone it doesn't have any of the usually associated distortion products. A 2.5khz cover to a 5" waveguide would work fine. As should an xover at 1.5kHz.
The SB driver looks great to me because of the low distortion yes.
How to match directivity correctly, still is confusing to me. Could you maybe respons to my reply at Linesource on page 1?
I thought on average a 6,5" driver would start beaming at around 1,5 khz, making a crossover at this frequency preferable.
 
Progress is being made.


This is the concept design of the box. Will let the front be CNC'd and the rest be done by a friend who is woodworker.


Dimensions (HxWxD): 34,5x19x33,5 cm. using 18 mm as sides and 36 mm as front.
An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.



Images: Imgur: The magic of the Internet
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.