I thought I would start a thread on my coming build, maybe to get some pointers/ideas/suggestions, even if the concept is pretty fixed already.
I have been reading about the bass from large woofers, even at lower listening levels, so this is why I have been dreaming about a build with 15" woofers. I thought that would be a fair compromise between what fits in my living room, and what would be considered a large woofer.
My living room is not that large, and it's all concrete, so I'm aiming for more directivity than my current speakers that are inspired by Troels PMS.
As in earlier builds I have again been inspired by Troels, but used other drivers and designed my own crossovers. The main inspiration here was the Faital 3WC-15 (http://www.troelsgravesen.dk/Faital-3WC-15.htm). First thought was to use the same woofer and mid, and replace the tweeter with a waveguided tweeter, since I believe there is some directivity 'mismatch' between the 8" mid and 1" tweeter..
However, it turned out the woofer (8ohm version) is on backorder, so I was on hold until I saw a used pair of 18 Sound15NMB420 advertised. I bought those, and have also ordered the 18 Sound 8NMB420 to use as mids. I have some Monacor WG300 waveguides and tweeters from previous projects, that sounded good to me, and measures good too, so the plan is to use them at least to start with: https://heissmann-acoustics.de/en/test-vifa-xt-300-xt25tg-waveguide-wg-300/
In my mind, they should make a pretty good match when it comes to directivity, even if XO point and steepness could be a bit sensitive.
I'm hoping to make a pretty simple box (monkey coffin style) with a volume around 150L. Will start with sealed, and possibly go with EQ or BR if needed, but I'm thinking room gain will add a significant amount to the low end.
When it comes to XO, I'm pretty convinced to go passive mid/tweeter, but I'm considering options when it comes to Woofer/mid (here I could use some advice):
1 Active DSP XO in computer (EQ APO) for woofer and mid HP
+easy to do
+cheap,
-but a very 'specific' setup, Needs a Win computer used as player, multichannel sound card
-only digital volume control
-a lot of cables and stacked amps
-Computer-trouble, and there is no sound.
-No way to run in any 'generic' setup.
2 Active 'analog' for woofer, passive HP for mid
+I like this option the most, amp and xo could be in the speaker, high level input used as signal source, only speaker cable and power to speaker needed) = could be hooked up anywhere to anything
+Low cost
-Don't know where I could find some DIY kits to build the filters
-Limitations in tuning possibilities
3 Active DSP for woofer, passive HP for mid
+ As above but more flexible
-Cost
-What is the delay in a Hypex amp module with DSP? If used like Troels does, the dealy would be applied to the woofer only.
Thoughts?
I have been reading about the bass from large woofers, even at lower listening levels, so this is why I have been dreaming about a build with 15" woofers. I thought that would be a fair compromise between what fits in my living room, and what would be considered a large woofer.
My living room is not that large, and it's all concrete, so I'm aiming for more directivity than my current speakers that are inspired by Troels PMS.
As in earlier builds I have again been inspired by Troels, but used other drivers and designed my own crossovers. The main inspiration here was the Faital 3WC-15 (http://www.troelsgravesen.dk/Faital-3WC-15.htm). First thought was to use the same woofer and mid, and replace the tweeter with a waveguided tweeter, since I believe there is some directivity 'mismatch' between the 8" mid and 1" tweeter..
However, it turned out the woofer (8ohm version) is on backorder, so I was on hold until I saw a used pair of 18 Sound15NMB420 advertised. I bought those, and have also ordered the 18 Sound 8NMB420 to use as mids. I have some Monacor WG300 waveguides and tweeters from previous projects, that sounded good to me, and measures good too, so the plan is to use them at least to start with: https://heissmann-acoustics.de/en/test-vifa-xt-300-xt25tg-waveguide-wg-300/
In my mind, they should make a pretty good match when it comes to directivity, even if XO point and steepness could be a bit sensitive.
I'm hoping to make a pretty simple box (monkey coffin style) with a volume around 150L. Will start with sealed, and possibly go with EQ or BR if needed, but I'm thinking room gain will add a significant amount to the low end.
When it comes to XO, I'm pretty convinced to go passive mid/tweeter, but I'm considering options when it comes to Woofer/mid (here I could use some advice):
1 Active DSP XO in computer (EQ APO) for woofer and mid HP
+easy to do
+cheap,
-but a very 'specific' setup, Needs a Win computer used as player, multichannel sound card
-only digital volume control
-a lot of cables and stacked amps
-Computer-trouble, and there is no sound.
-No way to run in any 'generic' setup.
2 Active 'analog' for woofer, passive HP for mid
+I like this option the most, amp and xo could be in the speaker, high level input used as signal source, only speaker cable and power to speaker needed) = could be hooked up anywhere to anything
+Low cost
-Don't know where I could find some DIY kits to build the filters
-Limitations in tuning possibilities
3 Active DSP for woofer, passive HP for mid
+ As above but more flexible
-Cost
-What is the delay in a Hypex amp module with DSP? If used like Troels does, the dealy would be applied to the woofer only.
Thoughts?
Last edited:
Hi, cool, big speakers are fun to listen to but pain in the *** to manufacture and measure
Its very tough to get full polar measurements on 150l speaker, which might mean there is no good measurement data for crossover / tweaks which might mean suboptimal end result. You can do it, its just cumbersome and thus error prone, I've built few and never again.
I suggest building separate structure for mid/hi and separate box for bass. This way you also can have more separation of concerns like edge diffraction secondary soubd source mitigation, if you are going to do some of that. Also could replace either without too much of a hustle should problems arise.
If you have DSP and amplifier power you could just go with simple closed boxes for the bass and EQ the bottom end and could go even smaller boxes if you wanted to. Build another box for mid/hi and if you are lazy and don't want to make huge roundovers for the top box (diffraction) the waveguide works just fine freestanding, even if it was made to be attached to a box.
Don't forget aesthetics of course, this advice is only form follows function. Also, full passive crossover might benefit thinking the stuff again, the baffle needs to play together with the crossover in order to get good results without making the xo too complicated and costly. Have fun!
Its very tough to get full polar measurements on 150l speaker, which might mean there is no good measurement data for crossover / tweaks which might mean suboptimal end result. You can do it, its just cumbersome and thus error prone, I've built few and never again.
I suggest building separate structure for mid/hi and separate box for bass. This way you also can have more separation of concerns like edge diffraction secondary soubd source mitigation, if you are going to do some of that. Also could replace either without too much of a hustle should problems arise.
If you have DSP and amplifier power you could just go with simple closed boxes for the bass and EQ the bottom end and could go even smaller boxes if you wanted to. Build another box for mid/hi and if you are lazy and don't want to make huge roundovers for the top box (diffraction) the waveguide works just fine freestanding, even if it was made to be attached to a box.
Don't forget aesthetics of course, this advice is only form follows function. Also, full passive crossover might benefit thinking the stuff again, the baffle needs to play together with the crossover in order to get good results without making the xo too complicated and costly. Have fun!
Last edited:
Thank you for the input!
Makes sense, maybe I should reconsider the plan to make one big box.. The initial plan was just to keep it simple with a square box, and make the best of the situation by playing with driver placement, maybe I should aim higher and make a separate box with a profiled baffle. Depends a bit if I can still access the same (magnificent) table saw I used before. If not, I have to keep it simple...
Anybody that has info on the processing delay in the Hypex DSP sub amps? I have not found this info online, and it could make me reject the option.
Makes sense, maybe I should reconsider the plan to make one big box.. The initial plan was just to keep it simple with a square box, and make the best of the situation by playing with driver placement, maybe I should aim higher and make a separate box with a profiled baffle. Depends a bit if I can still access the same (magnificent) table saw I used before. If not, I have to keep it simple...
Anybody that has info on the processing delay in the Hypex DSP sub amps? I have not found this info online, and it could make me reject the option.
Hypex FA digital processor global latency is 1,8ms (350us for analog pass-through), it for sure won't present any problems. 550Hz, 1,81ms cycle, 62,4cm in air
http://www.sengpielaudio.com/calculator-waves.htm
https://www.hypex.nl/documenten/download/1726
Traditional monkey box design will be OK, and getting same volume for bass with other shapes will make them more ugly IMO. Obviously you have quite a big room for them...
http://www.sengpielaudio.com/calculator-waves.htm
https://www.hypex.nl/documenten/download/1726
Traditional monkey box design will be OK, and getting same volume for bass with other shapes will make them more ugly IMO. Obviously you have quite a big room for them...
Thank you for that! So at 550Hz it's one cycle behind (summing with direct signal), which means at 275Hz it will be 180deg out and cancel the direct signal.. Not what I'm looking for it seems.
Room is just a standard living room, but I have enough space to place large speakers (and understanding girlfriend), and the room itself has openings to the rest of the apartment. Trial and error, as they say.. If it goes horribly wrong, I would like the speakers to be 'user friendly' to make them easier to sell, so that is why I don't like option 1 as a final solution. Maybe intermediate though, to do some listening tests etc.
Room is just a standard living room, but I have enough space to place large speakers (and understanding girlfriend), and the room itself has openings to the rest of the apartment. Trial and error, as they say.. If it goes horribly wrong, I would like the speakers to be 'user friendly' to make them easier to sell, so that is why I don't like option 1 as a final solution. Maybe intermediate though, to do some listening tests etc.
that is a very generalized statement, there is no way for you to know the actual AC of the drivers, or they're time / phase relationship without measuring them in the box, with the finished passive crossover in place.
Changing crossover order will change the phase relationship, same will physical placement.
If your planning on a flat baffle you will anyways have to adjust for it, if time alignment is a concern.
Also look into what TAD, JBL etc has done on speakers with large horn, using asymmetrical crossovers to "regain" time alignment etc.
And be aware the delay is not proportional with frequency, look it up on the forum.
The 15NMB420 (NMB= Neodymium MidBass) in a sealed box, will start to drop off fairly early -F3 100 hz or so. Your room gain i can't speak of.
42hz FS, 0,3 Qts. high sensitivity and fairly low xmax 6,5 mm, points to a BR box mainly.
It's a well behaved midbass, 72 gr mms, low le etc, the CSD is a little lively from the light cones, but overall a good driver.
Just not sure the bottom end will satisfy you, especially as sealed.
Last edited:
If I'm reading you correctly, it seems like you're missing a couple options:
4) Use something like a miniDSP for woofer crossover and midrange high-pass (this would also feed tweeter section). If you're bi-amping anyway it seems logical to let the DSP/active crossover do as much as it can for you instead of using a passive high-pass on the midrange (options 2 and 3). Since both sections would have the same delay, latency issues go away (for these 2 speakers, anyway). The miniDSP would also eliminate the need for the computer tie-in for function (still needed for setup/tuning, obviously)
5) Use Hypex modules on both the bass and midrange/treble. Same latency benefit as above.
To me it just seems to complicate things by adding almost 2ms delay to the woofer as a starting point, If there is a processing delay in the DSP, I'm pretty sure that delay will be constant with frequency. I had some play-time with DSP and delays/time alignments, and I was surprised I could hear a difference even in pretty short variations between woofer and mid (mid to tweeter even more sensitive of course). So, yes, I am aware of alignment issues, and this will probably not be perfect either once finished, but at least I can avoid adding extra delay to the woofer.
Yes, extension is something I will have to experiment with once they are in the actual listening room. I generally try to avoid BR since I don't like the GD, but maybe I will have to compromise on that too. I don't listen at extreme levels (I have neighbors), so I think X-max should be fine. I'm guessing woofer LP will be somewhere around 300-400Hz, with the baffle with I'm aiming for.
EDIT: I actually listened to the woofers 'wide band' with music, and they sounded pretty good, they could definitely play some midrange too.
Last edited:
4 Yes, it has been in my mind, but in previous projects I have ruled out minidsp because of the extra ADC and DAC conversion they add. I don't mind that in the bass, but mid-treble, I would like to avoid it. Who knows, maybe I will have to compromise.. It does make it more user-friendly. And the minidsp HD can be used as a DAC with USB input too, if I'm not mistaken. Definitely worth re-considering!
EDIT: looking at it again, I see there is now a minidsp flex, that would make it even more flexible and user friendly.
5 That is a no-go option for me. I like DIY amps, so for mid/treble I for sure want to 'cook my own', and I prefer to do that for bass too actually.
Last edited:
I think we have the same interest with large Woofer. a little story of my journey : i have used 2x15" woofer and 21" on open baffle, added
infinity kappa perfect 12VQ in dual-opposed and make them 2 for each left right channel
I have enjoyed it for almost 2yrs and thought it was end journey, then I visited 2 different local high end show 2018 2019 before covid came and listened to several expensive room like Eaglestoneworks,MBL,Sonus Faber, Tannoy, Raido, Focal etc and it reminded me of my young experience of music coming from Klipsch promedia 2.1 and Creative speaker.
what i had been missing those years was the slam/kick in midbass area especially if you play Call of Duty series game and those bang boom was gone with big drivers. then i finished this project last months and i can confess that i get back those midbass and enjoy it more for music
https://www.diyaudio.com/community/...ld-with-dynaudio-drivers.387930/#post-7098565
I have dissembled my open baffle and full range unused, now I'm going bigger with 2x9" woofer and Esotar mid+tweeter.
someday if budget permit i might build another 3way classic big box like ATC style
anyway, I'm using minidsp 2x4HD for mid & tweeter and behringer nu3000dsp for woofer, i have to upgrade my dsp to Zapco HDSP 8 channel someday because i ca't design passive xo.
however i have contacted one of my diy builder and he can design the passive xo for me with $100 charge (design only) so i might try someday with passive mid+tweeter and compare with my dsp approach which one is more preferable to my ears
ymmv
Attachments
I think the 'slam/kick/boom' you were missing with open baffles is a common problem, boxed speakers are able to 'pressurize' the room better and give you that 'slam/kick'. This is mostly based on what I read though, I tried 12" open baffle woofers once, but did not like it. I like closed boxes just for that slam, here I just want to go bigger and hopefully get more of the good stuff. I need the drums and snappy bass notes to sound just right.
I did some reading on the minidsp flex, and it is a really tempting package. Anyway, that will be for later.
Right now I'm playing around with baffle and driver location in vcad. I see some benefit from placing the mid on top as Troels did when it comes to floor reflections, so I will probably go for that. Rounded baffle at around R=100mm on mid/tweeter baffle is desirable, just have to figure out how to solve it. Thinking of 45deg bevels, and then adding some cut pipe to make the smooth rounding. It gets a bit complicated though, because I would also like the division between woofer and mid cabinets to be at an angle to reduce standing waves in both bass and mid cabinets, and when one angle meets another angle in another plane, the cuts with the saw get a bit complicated, and have to be very precise...
Right now I'm playing around with baffle and driver location in vcad. I see some benefit from placing the mid on top as Troels did when it comes to floor reflections, so I will probably go for that. Rounded baffle at around R=100mm on mid/tweeter baffle is desirable, just have to figure out how to solve it. Thinking of 45deg bevels, and then adding some cut pipe to make the smooth rounding. It gets a bit complicated though, because I would also like the division between woofer and mid cabinets to be at an angle to reduce standing waves in both bass and mid cabinets, and when one angle meets another angle in another plane, the cuts with the saw get a bit complicated, and have to be very precise...
Lazy man does separate cabinet for each driver, so small no roundovers fit
its about as good performance diffraction related ripple wise as big baffle with too small roundovers, except much essier to build. Roundovers are most effective when they start immediately beside the driver. Well, as always there emerges new issues. Anyway, just food for thought.
I love flat speakers with wide baffles, but they are even more difficult to place in the room.
Sonus Faber Stradivari
Troels Gravesen http://www.troelsgravesen.dk/PMS.htm
also kimmosto has one desing with curved/multiangled baffle, but he has hidden his constructions from his site
Sonus Faber Stradivari
Troels Gravesen http://www.troelsgravesen.dk/PMS.htm
also kimmosto has one desing with curved/multiangled baffle, but he has hidden his constructions from his site
I second this thought. I recently went through the full test regime of a tall bulky speaker. I did polar responses in the horizontal and vertical plane. This speaker was 46 inches tall, and 80 lb, so it is a fraction of the size of a 150 l box, and yet it was still a lot of work to measure properly. For the vertical polar measurements, I had to build a specialty rotating table to position the speaker sideways at a height of 48 inches from the floor. Lifting it up on the platform was a challenge to do solo.
For large speakers, I will always strongly lean towards a separate woofer cabinet. This way the mid-tweeter cabinet can be easily measured in the horizontal and vertical planes. There is little to be gained in making vertical polar measurements of a woofer cabinet that operates up to ~ 800 Hz or so. I will simply allow the simulation software to assume that vertical and horizontal polar responses are equivalent.
Attachments
I'm leaning towards a separate mid cabinet too, just struggling with sketchup, so I had to do something more interesting on one of the mids:
https://www.diyaudio.com/community/threads/the-elsinore-project-thread.97043/post-7116385
I might add that doing the full polar measurements might never be done anyway..
https://www.diyaudio.com/community/threads/the-elsinore-project-thread.97043/post-7116385
I might add that doing the full polar measurements might never be done anyway..