Hi,
as a followup to my othe thread (New diy open baffle project: 3-way with AMT) starting my open baffle project I am a bit further now in decision making:
Tweeter: AMT23D6.1-R (88dB, 8Omhs) is already ordered
Mid woofer: Audiotechnology C-Quenze 18 H 52 06 13 SDKA – last perhaps in a version with the J-magnet instead 90dB
I am now looking for a matching subwoofer, xo is planned around 200 Hz. I am thinking of one of the following:
I would prefer the 12" drivers as I awant to build a rather slim baffle but with side wings (U-frame). I tried to chosse drivers with higher sensitivity to compensate for the baffle loss for low frequencies. Are there reasosn one of them would not match well with the above mid/high drivers?
Another question:
Is there a way to simulate U-frame baffles? I wand to make them triangular shaped, is there a way to simulate this as well? Can I use Basta! for this?
Is the response the same when I "flip" the wings to the front as Basta! can only simulate flat baffles as far as I know....
as a followup to my othe thread (New diy open baffle project: 3-way with AMT) starting my open baffle project I am a bit further now in decision making:
Tweeter: AMT23D6.1-R (88dB, 8Omhs) is already ordered
Mid woofer: Audiotechnology C-Quenze 18 H 52 06 13 SDKA – last perhaps in a version with the J-magnet instead 90dB
I am now looking for a matching subwoofer, xo is planned around 200 Hz. I am thinking of one of the following:
- Acoustic Elegance Dipole12 or Dipole15
- Supravox 285 GMF
- SB Audience Bianco 15OB350 Open Baffle (or the 12OB150)
I would prefer the 12" drivers as I awant to build a rather slim baffle but with side wings (U-frame). I tried to chosse drivers with higher sensitivity to compensate for the baffle loss for low frequencies. Are there reasosn one of them would not match well with the above mid/high drivers?
Another question:
Is there a way to simulate U-frame baffles? I wand to make them triangular shaped, is there a way to simulate this as well? Can I use Basta! for this?
Is the response the same when I "flip" the wings to the front as Basta! can only simulate flat baffles as far as I know....
Hi,
so I have tried to model now the "lower" part of the system with basta!. For the AMT I did not find any S/T parameters.
I guess due to the AMT principle these might not apply anyhow. Any ideas how to model the AMT?
So here is my model: Trapezoid shape is supposed to simulate a U-frame with triangular sides, I hope one can do it this way.
Mid-woofer is the AT C-Quenze 18H52 driver, subwoofer: AE Dipole12. Any commenst or suggesntions?
so I have tried to model now the "lower" part of the system with basta!. For the AMT I did not find any S/T parameters.
I guess due to the AMT principle these might not apply anyhow. Any ideas how to model the AMT?
So here is my model: Trapezoid shape is supposed to simulate a U-frame with triangular sides, I hope one can do it this way.
Mid-woofer is the AT C-Quenze 18H52 driver, subwoofer: AE Dipole12. Any commenst or suggesntions?
So I learnt a bit more on how to do the modeling with basta. Here ist now my last design that makes me confident to go for the AE Dipole12. Crossovers are now all 1st order as I read somewhere that these are the "purest" in terms of phase delays.
My question: Is it "valid" to crossover between subwoofer and midwoofer at different frequencies? Or does this cause problems?
Here I use a xo frequency of 250 Hz for the subwoofer, but 900 Hz for the midwoofer to "cover up" the baffle resonance between 200 and 800. I tried to maintain a sensitivity of about 88dB as this would match with the AMT above 2400 Hz which is not included in this graph.
My question: Is it "valid" to crossover between subwoofer and midwoofer at different frequencies? Or does this cause problems?
Here I use a xo frequency of 250 Hz for the subwoofer, but 900 Hz for the midwoofer to "cover up" the baffle resonance between 200 and 800. I tried to maintain a sensitivity of about 88dB as this would match with the AMT above 2400 Hz which is not included in this graph.
Correct, each order increases by 90 deg, so 4th = 360 deg = a full 1 Hz out of phase. Yes, it's called 'gapping' to add time delay/phase rotation; very common when matching up a point source driver with its 90 deg of delay to a compression horn with its ~ zero delay.
That said, I leave it to the dipole experts to educate us WRT a full two octave gap.
That said, I leave it to the dipole experts to educate us WRT a full two octave gap.
BTW: This is where I found the reasoning why 1st order xo is "better":
https://www.3squareaudio.com/copy-of-1st-order-crossovers
Not sure if I would be able to hear the difference but I kind of like the idea......
https://www.3squareaudio.com/copy-of-1st-order-crossovers
Not sure if I would be able to hear the difference but I kind of like the idea......
One question regarding U-frame: I am wondering if I need to simulate that differently. When I "flip" the sings over, do I deen to make them twise as wide because the travel distance is not d*2 (d being the wing depth but only d (assuming the listener position is far enough away?
Right, I just didn't 'spell' it all out and neither did they. Regardless, We all hear the same, yet not so much, so depends on one's hearing acuity and where in our hearing BW the XO BW is plus how loud it is relative to ambient noise:
Instrument chart
Hearing sensitivity chart
Thanks! Wow! Sure has changed since my using it way back in its early days. Guess I need to learn it if I'm going to respond to its usage, so going to fall back to leaving it to the dipole experts to educate us.
I am now one step further and made definitve decisions on drivers. In addition to the Mundorf AMT I will have:
Audiotechnology 18 J 52 06 13 SDKAF as mid woofer and
Acoustic Elegance Dipole 12 D16 as bass woofer.
Now the question comes with what strategy do I continues:
How do I make definitve decisions on xo parts so that I do not end up with too many expensive "spare"parts?
I understand tha xo design is an iterative process that needs fine-tuning, but how does one do that with a minimum set of parts?
And what do I need to consider so that I do not end up causing damge to drivers while testing/tuning?
Is there a recommended parts list for tuning? Can I use cheaper equipment before buing the final "versions"?
As tools I will have an impedance/frequency response measurement box with built-in mono amp, parts are on the way to set this up...
Audiotechnology 18 J 52 06 13 SDKAF as mid woofer and
Acoustic Elegance Dipole 12 D16 as bass woofer.
Now the question comes with what strategy do I continues:
How do I make definitve decisions on xo parts so that I do not end up with too many expensive "spare"parts?
I understand tha xo design is an iterative process that needs fine-tuning, but how does one do that with a minimum set of parts?
And what do I need to consider so that I do not end up causing damge to drivers while testing/tuning?
Is there a recommended parts list for tuning? Can I use cheaper equipment before buing the final "versions"?
As tools I will have an impedance/frequency response measurement box with built-in mono amp, parts are on the way to set this up...
2 dimensional simulation like Edge or Basta had very low precision in OB simulation. OB by sound radiation meaning is 3D structure and you need some other type of simulation, somthing like AKABAK what use BEM (Boundary Element Method) and is able to simulate full OB 3D srturcture and it radiation.
Have you read through Linkwitz's site, there is a plethora of useful information.
He goes into crossover slopes for the LX521 project and how it effects the polar response of the speaker.
He goes into crossover slopes for the LX521 project and how it effects the polar response of the speaker.
Your mixed use of terms might cause confusion. Are you making a frame or a folded baffle? A frame has four sides, so a U-frame has two side walls plus a top and a bottom. Looks like you are doing a folded baffle?
Hornresp allows sim's for unfloded baffles and U-frame designs with the driver of your choice. It only seems to allow a floor plan sim, as if looking down and changing x and y but not z aspects. Thats understandable when you consider a U-frame needs a lid.
A U-frame is enclosed like a tunnel. It has roof. The roof is needed to stop dipole cancellation. If you are not using a roof you are not making a U-frame. You are making a folded baffle only. The baffle will have less resonance but much more dipole cancellation with the waves taking the path of least resistance and cancelling over the top.
If you want to see how a U-frame extends the low frequency wuth resonance costs vs a baffle, you can do that with hornresp and your selected driver.
Hornresp also allows modelling of a driver in a plain baffle with SPL and Xmax. It will allow you to sim at what power and frequency you exceed Xmax or run risk of damage. Woofers in OB can easily bottom out at low freq and higher power.
The AE dipole series is hard to beat for performance in OB being designed specifically for OB with quality and performance prioritised over cost.
Hornresp comes with a Help file for instructions but I just play around with options with the mouse e.g.
One possible idea that others will know far more about more then me, would be to use digital crossovers with active drive. That could allow a rapid iterative design process without buying any parts at the click of a mouse from your listening chair. When you reach a stable crossover design then you go shopping with confidence.
Digital crossovers are freely offered with info on this site for PC use. Active drive hardware could perhaps be a repurposed AV receiver with six channel digital input and six channels amplified out. You might even have one or know someone with one on a shelf? You just need to sort out that you can feed it six digital channels. Sound quality will be limited but for testing of crossover configurations it could be illuminating.
Thanks, I will have a look at hornresp. I actually planned a "roof" for the U-frame but I am not sure hom much that will make a difference. As far as I know, the dipole effect is most pronounced for low frequencies. For that reason I plan to have wider "side-wings" on the bottom where the subwoofer is placed.
I am already constructing a prototype setup wher I can easily change some parameters once all parts have arrived. I think without that and measurements I will not come to good results.
I am already constructing a prototype setup wher I can easily change some parameters once all parts have arrived. I think without that and measurements I will not come to good results.
You can model the influence that a wider baffle side wing deign will have on low frequency extension using Basta. That might help you with your prototypes. It may be if you are using sloping side baffles that increases depth doesn't make too much change. The cancellation may be dominated by the path of least resistance eg over the top of the front baffle?????
Great. It will be interesting how close your sims come to the measured.
Last edited: