Hi,
Trying to experiment for the first time. I've got some MAOP 10.2's well broken in, but currently homeless and I am considering grabbing some SB Audience 18SW450 18" subwoofers cause I've never really had "big displacement" drivers before.
Considering my first build with these to be an open baffle, cause I also haven't had an open baffle before, or even listened to one.
AND I'm also trying to design a speaker for the first time.
This is where I've gotten to currently.
Please note the shape of the baffle is a bit rough. Not deliberately wonky! Designed to be a winged OB.
1st order lowpass on SW at 100hz, 1st order highpasson MAOP at 600hz.
Without xovers:
I have a sneaking suspicion that the response on the MAOP is actually terrible. I don't believe BASTA models well above 1k though, so I'm not sure. Note the scale of the dB response.
I also don't like that the Full ranger is so far away from the SW driver (40cm).
I believe it should be within 1/4 wavelength of crossover to sum well? Which is actually kind of impossible with an 18" woofer anyway? With a non-symmetrical crossover, which should I be calculating for? The midpoint at 350hz? ~25cm
I don't actually really know what I'm looking at. I'm going through and trying to compare it all to the documentation that is online, and actually learn what I'm doing, but understanding what keywords I should be looking for in this instance, might accelerate the process significantly. I'd love anyone's tips or feedback here so I can move to my next phase of learning.
Is the peaking at the bottom end of the MAOP response "bad" or expected with OB?
What is a good goal response for the FR driver if this is actually as bad as I think?
How high in the bandwidth do I stop paying attention to BASTA's sim?
Does the SW actually look good like this with the lowpass in place?
Should I consider a H frame or U frame for the SW with a separate baffle on top for the MAOP?
Thank you for your time, I can't wait to hear your thoughts.
Trying to experiment for the first time. I've got some MAOP 10.2's well broken in, but currently homeless and I am considering grabbing some SB Audience 18SW450 18" subwoofers cause I've never really had "big displacement" drivers before.
Considering my first build with these to be an open baffle, cause I also haven't had an open baffle before, or even listened to one.
AND I'm also trying to design a speaker for the first time.
This is where I've gotten to currently.
Please note the shape of the baffle is a bit rough. Not deliberately wonky! Designed to be a winged OB.
1st order lowpass on SW at 100hz, 1st order highpasson MAOP at 600hz.
Without xovers:
I have a sneaking suspicion that the response on the MAOP is actually terrible. I don't believe BASTA models well above 1k though, so I'm not sure. Note the scale of the dB response.
I also don't like that the Full ranger is so far away from the SW driver (40cm).
I believe it should be within 1/4 wavelength of crossover to sum well? Which is actually kind of impossible with an 18" woofer anyway? With a non-symmetrical crossover, which should I be calculating for? The midpoint at 350hz? ~25cm
I don't actually really know what I'm looking at. I'm going through and trying to compare it all to the documentation that is online, and actually learn what I'm doing, but understanding what keywords I should be looking for in this instance, might accelerate the process significantly. I'd love anyone's tips or feedback here so I can move to my next phase of learning.
Is the peaking at the bottom end of the MAOP response "bad" or expected with OB?
What is a good goal response for the FR driver if this is actually as bad as I think?
How high in the bandwidth do I stop paying attention to BASTA's sim?
Does the SW actually look good like this with the lowpass in place?
Should I consider a H frame or U frame for the SW with a separate baffle on top for the MAOP?
Thank you for your time, I can't wait to hear your thoughts.
Last edited:
Here's my quick Basta impersonation of your sim. I used Alpair 10.3 parameters - that seems to be what's suggested for the MAOP 10.2. I reversed phase on the MAOP, too. I think you sim has far too much roll-off on the MAOP, which is probably to do with Le. WHat did you enter? MarkAudio give it as 71 uH, which is 0.071 mH, of course.
There's no reason why Basta sims stop being useful at higher frequencies, except that they are TS based sims, and assume perfect pistonic behaviour from the driver. So all the quirks of a particular driver's frequency response will be missing. Many drivers have a rising response trend, which will counteract the inevitable HF downward trend of the sim, due to voice coil inductance.
There's no reason why Basta sims stop being useful at higher frequencies, except that they are TS based sims, and assume perfect pistonic behaviour from the driver. So all the quirks of a particular driver's frequency response will be missing. Many drivers have a rising response trend, which will counteract the inevitable HF downward trend of the sim, due to voice coil inductance.
I think it's a very reasonable attempt, actually, and a decent plan. Don't worry that the nominal values Basta has suggested for the crossover reflect a wide spread - the actual response shape in practice is what counts.
Play with the baffle shape and crossover, aiming for smooth response and good phase tracking in the crossover region, and see how it develops. You know if you right-click and select "view options" you can bring up a dialogue box that allows you show the combined response?
Play with the baffle shape and crossover, aiming for smooth response and good phase tracking in the crossover region, and see how it develops. You know if you right-click and select "view options" you can bring up a dialogue box that allows you show the combined response?
Thank you for this! A heap of interesting things for me to note.
1. I need to learn more about phase
2. I definitely got the Le wrong 😑
I didn’t know I could see a combined response. That’s super useful.
I didn’t use suggested values for the crossover I believe; Are you referring to the lookup tables that are incorporated in Basta for the crossovers?
Thanks for the kind help and kind words, it’s exciting to know all my reading paid off a bit!
1. I need to learn more about phase
2. I definitely got the Le wrong 😑
I didn’t know I could see a combined response. That’s super useful.
I didn’t use suggested values for the crossover I believe; Are you referring to the lookup tables that are incorporated in Basta for the crossovers?
Thanks for the kind help and kind words, it’s exciting to know all my reading paid off a bit!
My first thoughts towards this is the width of the baffle for the Bianco. While I recognise the requirement for keeping decent bass extension, hence the wider the better, it does need to maintain dipole radiation high enough up in frequency to meet the MAOP.
I wouldn't be surprised if this would be dipole at low frequencies, transition to monopole at the top end of the Biano's passband, and then go dipole again, before going monopole at the width dictated by the MAOPs baffle width/cone diameter.
By the looks of things the dipole peak on the Biacno is at about 380Hz so you might be okay.
I wouldn't be surprised if this would be dipole at low frequencies, transition to monopole at the top end of the Biano's passband, and then go dipole again, before going monopole at the width dictated by the MAOPs baffle width/cone diameter.
By the looks of things the dipole peak on the Biacno is at about 380Hz so you might be okay.
Yes, I meant the lookup tables. I'm guessing that was what you meant when you said the lowpass was at 100 Hz and the highpass at 600?
A few more thoughts. I'm not sure if the SB 18 will be at its best crossing as high as you are. You might want to bring the crossover point down a bit?
It might be worth going second order on the woofer low pass. You can modify the slope by putting a resistor in series with the shunt capacitor. (To model this in Basta, you're best to use the "Advanced network" tab, rather than the "Passive filter" tab. Though actually you can fake it on the "Passive filter" tab by using first order low pass plus the conjugate link option, with custom values.)
The woofer is probably not quite going to give enough output to match the MAOP, especially when you allow for some resistance in the woofer inductor. So you might need to pad down the MAOP a little. Again, the Advanced network tab is the easiest place to model this.
Basta is a great tool for learning about all this stuff.
Anyway, here's an updated sim I did with adjusted crossover: on the woofer, 11mH series coil (0.5 ohms dcr) and 22 uF shunt cap with 4 ohms. On the MAOP, 2 ohms in series, and a 40uF series cap. MAOP phase inverted. (Crossover point not lowered, obviously. That's another challenge.)
A few more thoughts. I'm not sure if the SB 18 will be at its best crossing as high as you are. You might want to bring the crossover point down a bit?
It might be worth going second order on the woofer low pass. You can modify the slope by putting a resistor in series with the shunt capacitor. (To model this in Basta, you're best to use the "Advanced network" tab, rather than the "Passive filter" tab. Though actually you can fake it on the "Passive filter" tab by using first order low pass plus the conjugate link option, with custom values.)
The woofer is probably not quite going to give enough output to match the MAOP, especially when you allow for some resistance in the woofer inductor. So you might need to pad down the MAOP a little. Again, the Advanced network tab is the easiest place to model this.
Basta is a great tool for learning about all this stuff.
Anyway, here's an updated sim I did with adjusted crossover: on the woofer, 11mH series coil (0.5 ohms dcr) and 22 uF shunt cap with 4 ohms. On the MAOP, 2 ohms in series, and a 40uF series cap. MAOP phase inverted. (Crossover point not lowered, obviously. That's another challenge.)
Thanks again to you both.
How do I calculate when the woofer would go dipole vs monopole? Do we just reference where the dipole peak is in the unfiltered response?
Regarding the MAOP baffle, I can actually go quite small without sacrificing extension it seems, I'm assuming this is a good idea to maximising the dipole passband?
Ianbo,
1. why do you suggest that the woofer should go 2nd order? Just so I understand the line of reasoning. I assume it's to roll it off sooner and avoid the monopole response that 5th element is referring to?
1a. Does this then cause a phase issue with a 2nd order and 1st order crossover?
2. Would Basta be taking into account the rising response of the Maop in the mid bass? Or is that still subject to the TS modelling you mentioned earlier?
3. You'd pad down the MAOP to ensure a rising response into the bass frequencies for loudness equalisation?
I guess I need to start reading more about crossovers now. I wasn't really thinking that far ahead, but this actually seems achievable!
Also, I realised I have the exact parameters of my MAOPS in my photo album!
How do I calculate when the woofer would go dipole vs monopole? Do we just reference where the dipole peak is in the unfiltered response?
Regarding the MAOP baffle, I can actually go quite small without sacrificing extension it seems, I'm assuming this is a good idea to maximising the dipole passband?
Ianbo,
1. why do you suggest that the woofer should go 2nd order? Just so I understand the line of reasoning. I assume it's to roll it off sooner and avoid the monopole response that 5th element is referring to?
1a. Does this then cause a phase issue with a 2nd order and 1st order crossover?
2. Would Basta be taking into account the rising response of the Maop in the mid bass? Or is that still subject to the TS modelling you mentioned earlier?
3. You'd pad down the MAOP to ensure a rising response into the bass frequencies for loudness equalisation?
I guess I need to start reading more about crossovers now. I wasn't really thinking that far ahead, but this actually seems achievable!
Also, I realised I have the exact parameters of my MAOPS in my photo album!
I'm struggling to remember where I've found good reading on this. (Edit: I remembered! Linkwitz Lab, here: https://www.linkwitzlab.com/models.htm#A3) When the baffle is wide relative to the size of the cone, the off-axis dispersion broadens and gets messier above the dipole peak. The unfiltered response in the sim is a decent guide to where the dipole peak is - round about 350 Hz here, I think. So using the 18 up to over 500 Hz might be a compromise. (However, speaker design is always a set of compromises.)
Well yes, it's harder to roll it off low if you have a 1st order crossover. Also there's a peak in the woofer's response around 1500 hz, which wouldn't be well suppressed, and might be an issue.
I don't think so. It's the acoustic slope that matters, after all, not the electrical slope. When I played with the sim, I was able to get better phase matching by going second order on the woofer, and damping the shunt cap with a resistor.
No, Basta takes into account TS parameters, crossover, box type/size, and baffle size/shape. It can't know about the individual behaviour of a driver beyond those factors. As for that rise in the bass, I'm not sure what measurement setup MarkAudio use for those factory responses. (IEC baffle?) So I don't know if you'll get the same hump low down, in your baffle.
No, I was thinking you'd need to pad it a bit just to get a reasonably flat looking response overall. But it depends on what crossover you end up with at the end of the day. The problem at the moment is that all this is just simulations, as I'm sure you realise. It might or might not work as the sims suggest. Normally, if you're happy with the sim then you build the speaker and measure the drivers in the actual box or baffle, and then design the actual crossover on that basis. There's a catch with OB speakers, though: measuring them is not easy!
Last edited:
Hi @Mitch311, I've done something similar with smaller drivers 🙂
Guitarist OB Work in progress on a better crossover but need more works, but I am very satisfied with the result.
To complete/repeat an earlier response, with Basta you can emulate a complete passive crossover with the driver in the 3D space 😉
The result is very close of the measurement, but Basta doesn't take into account break up, phase shift for example. Measurement needs to see if it is good.
Guitarist OB Work in progress on a better crossover but need more works, but I am very satisfied with the result.
Yes if you use the H/U as bass helper < 150Hz, I don't recommend use H/U frame in the critical mid.
To complete/repeat an earlier response, with Basta you can emulate a complete passive crossover with the driver in the 3D space 😉
The result is very close of the measurement, but Basta doesn't take into account break up, phase shift for example. Measurement needs to see if it is good.
I don't really know what these crossover values actually equate to. I've just been fiddling from the values Ianbo provided.
They might not work at all, I'm not sure.
So points I'm worried about:
18SW450 likely goes monopole at around 400-500hz.
It has a breakup mode at approx 1500hz of 7-8db.
MAOP 10.2 has a rising response from 500hz down to 100hz.
Crossover numbers being completely arbitrary?
C-t-C distance of the two drivers causing lobing.
It's a pretty big baffle.
Some big capacitors & inductors.
Positives:
Looks pretty smooth for two wildly different drivers.
Good extension
MAOP 1.5khz dip filled by sim response curve
I think a good looking phase response?
Basta room gain model really extends the bass. - Not sure how believable this is?
Some questions:
Does BASTA take into account floor gain for the woofer?
Do I likely need a passive notch filter for the woofer?
Is this starting to look good enough that DSP will solve any other problems?
Is the modelled crossover completely bonkers or kind of realistic?
You're making progress. If you want to show the phase, right click on the FR window and select "Show phase".
You've made the wings bigger, so the effective baffle width of the lower section is now about 1.5m. That's a lot! Do you need the extra efficiency? (Low powered amp, maybe?) This pushes the dipole peak down to around 250 Hz, I reckon, and might make the radiation pattern more difficult around 4-500 Hz. It might also give you some cavity resonance. If the effective baffle wasn't so wide, the radiation pattern would probably be fine up to the crossover point.
I doubt if the woofer breakup will be an issue now - the level at 1500 Hz is much lower thanks to the second order crossover, and the reduced crossover point.
Because of the extra efficiency from the wings, the FR is now tilted down, a little, so you could well reduce the padding on the MAOP.
Basta doesn't take floor gain into account, so if you plan to have the baffle sitting tight on the floor, you will get some boost from that. (You can fake what that might look like by extending the baffle below the 0 point.) If I understand Basta correctly, however, the room gain function isn't helpful for modelling OB speakers - I believe is shows the room gain you might get with a sealed box, and OB won't behave the same way because there's no pressurisation of the room.
You're right - there are some chunky components in the crossover. This is where it become a difficult judgment call. Ideally you'd have a stash of components to play with, so you could try out a modelled crossover and tinker with it until you get it just the way you want.
P.S. On the curves tab, I usually select "Max output level" and "Cone excursion" - those are useful things to know with OB models.
You've made the wings bigger, so the effective baffle width of the lower section is now about 1.5m. That's a lot! Do you need the extra efficiency? (Low powered amp, maybe?) This pushes the dipole peak down to around 250 Hz, I reckon, and might make the radiation pattern more difficult around 4-500 Hz. It might also give you some cavity resonance. If the effective baffle wasn't so wide, the radiation pattern would probably be fine up to the crossover point.
I doubt if the woofer breakup will be an issue now - the level at 1500 Hz is much lower thanks to the second order crossover, and the reduced crossover point.
Because of the extra efficiency from the wings, the FR is now tilted down, a little, so you could well reduce the padding on the MAOP.
Basta doesn't take floor gain into account, so if you plan to have the baffle sitting tight on the floor, you will get some boost from that. (You can fake what that might look like by extending the baffle below the 0 point.) If I understand Basta correctly, however, the room gain function isn't helpful for modelling OB speakers - I believe is shows the room gain you might get with a sealed box, and OB won't behave the same way because there's no pressurisation of the room.
You're right - there are some chunky components in the crossover. This is where it become a difficult judgment call. Ideally you'd have a stash of components to play with, so you could try out a modelled crossover and tinker with it until you get it just the way you want.
P.S. On the curves tab, I usually select "Max output level" and "Cone excursion" - those are useful things to know with OB models.
Shrunk the baffle back down. I can actually build 2x of these minus kerf from a single sheet of timber. The MAOP has a 55uF capacitor only on it. Approx a 450hz crossover.
Without crossovers:
Dipole bump is approx 350hz.
When I extended the baffle 30cm below 0 I got a bump of 3db, really solidified down to 40hz.
I'm hoping the alternate sizing of the wings will help deal with any resonances. Should I make them different lengths?
100 watts / 30v pushes the woofer to xmax which is 112db at 100hz.
I'm actually wondering if I should consider going active for these.
4x big capacitors and 2x big inductors gets relatively pricey, might be worth the extra dollars to go active instead.
Attachments
Why do you need the big cap HPF? In past attempts I haven't liked their sonic cost (Jantzen Crosscap mixed with Standard, but never "burned-in"). Whereas big aircore inductor turned out just fine. Active LPF and straight-through semi-nude MAOP?
wchang, the answer is unflattering but simple.
Why? Cause I have no idea what I’m doing! 😂
Can you suggest reading material or suitable keywords for a Google search to send me learning with regards to inductor high pass filters?
With the Maop are you suggesting I let it run open without crossover?
Why? Cause I have no idea what I’m doing! 😂
Can you suggest reading material or suitable keywords for a Google search to send me learning with regards to inductor high pass filters?
With the Maop are you suggesting I let it run open without crossover?
The first photo above didn't actually have the HPF on for the MAOP. Here is what it actually looks like.
That is an option. And it models OK, potentially. However driving the MAOP full range it'll be subject to a lot of excursion because of LF content - around 2mm for 2.83V at 40 Hz, for example.
I meant aircore inductor LPF.
MAOP is fullrange relatively thick AlMg not fragile like some thin-paper-cone speakers that needed "air cushion" protection playing bass. Running it OB small baffle or baffleless will HPF it even without electrical parts -- as a baseline.
(edit after seeing latest chart) Response improvement may or may not compensate for the big cap SQ loss. The woofer can be tweaked further if needed in room.
MAOP is fullrange relatively thick AlMg not fragile like some thin-paper-cone speakers that needed "air cushion" protection playing bass. Running it OB small baffle or baffleless will HPF it even without electrical parts -- as a baseline.
(edit after seeing latest chart) Response improvement may or may not compensate for the big cap SQ loss. The woofer can be tweaked further if needed in room.
Last edited:
It's an option. I like passive, but I've built up a bank of crossover parts over time, many of them bought used, and including some chunky inductors. Active has the advantage that once you've bought a miniDSP or whatever, you can play with the crossover settings without having to buy lots of parts.