Hello,
First of all I would like to introduce myself, my name is Dominic and I'm on this forum since several years. I use to read and learn a lot here but I don't write so much usually since I'm not experienced enough to give advice to anyone 😆.
I would need your help about the building of my new speakers. I wanted to use the TAD 2001 drivers in the simplest design possible, so I decided to replicate a project that I found on the web (french forum), it's a 2way design with tad td2001 drivers on jmlc horns and supravox 285gmf woofers in MLTL enclosure, with Jean-Michel crossover. I don’t know if I can put the link here, but it's easy to find.
My first idea was to use supravox 215 woofers (similar to Cessaro entry level speakers) but I was not able to find an already existing design (including crossovers circuits) so that’s why I ended up with the 285.
Long story short, I don't want to make major changes in the design since I have not enough experience and I don't want to mess with it, I just would like to "adjust" a little bit the shape of the bass enclosure.
Attached the simulation published for the original 150L cabinet (internal 108x40x40), done with a software by Martin J. King (anyone used it?).
My question is: is it possible to reduce the height (around 90cm internal) and in case also the width (around 32cm internal) keeping similar results on the bass extension, maybe increasing the depth to compensate? Or this would spoil the entire MLTL design concept? (It would become a “normal” BR…?)
Can anyone help me recalculate the enclosure with reduced height and width?
If it’s not possible to reduce the height of the box with the supravox 285, maybe it would be possible using the 215 woofers? In that case I would need to recalculate the enclosure and adjust the crossover.
Thank you very much for your help!
Dominic
First of all I would like to introduce myself, my name is Dominic and I'm on this forum since several years. I use to read and learn a lot here but I don't write so much usually since I'm not experienced enough to give advice to anyone 😆.
I would need your help about the building of my new speakers. I wanted to use the TAD 2001 drivers in the simplest design possible, so I decided to replicate a project that I found on the web (french forum), it's a 2way design with tad td2001 drivers on jmlc horns and supravox 285gmf woofers in MLTL enclosure, with Jean-Michel crossover. I don’t know if I can put the link here, but it's easy to find.
My first idea was to use supravox 215 woofers (similar to Cessaro entry level speakers) but I was not able to find an already existing design (including crossovers circuits) so that’s why I ended up with the 285.
Long story short, I don't want to make major changes in the design since I have not enough experience and I don't want to mess with it, I just would like to "adjust" a little bit the shape of the bass enclosure.
Attached the simulation published for the original 150L cabinet (internal 108x40x40), done with a software by Martin J. King (anyone used it?).
My question is: is it possible to reduce the height (around 90cm internal) and in case also the width (around 32cm internal) keeping similar results on the bass extension, maybe increasing the depth to compensate? Or this would spoil the entire MLTL design concept? (It would become a “normal” BR…?)
Can anyone help me recalculate the enclosure with reduced height and width?
If it’s not possible to reduce the height of the box with the supravox 285, maybe it would be possible using the 215 woofers? In that case I would need to recalculate the enclosure and adjust the crossover.
Thank you very much for your help!
Dominic
Hi, I'm far from having expertise in this, in fact I envy you for access to MJK software. I do have a version of Supravox 215 in small 17L cabs dubbed "TLonken" (similar to ML-TL in principle), whose bass sounded rather like your simulated response curve. You can search for it on diyaudio. Of course bigger is better but probably not 9X better. By tapering the "line" you should be able to gain length and volume efficiency to offset the reduction you wish to make. But if you must stick with a regular ported box, either the software or an expert will have to tell you yes/no/maybe....
(added)
https://www.diyaudio.com/community/threads/supravox-8-tlonken-while-city-of-22m-laid-flat.393338/
https://www.diyaudio.com/community/...rkaudio-alpair-5-2-in-evansound-3-75l.393909/
(added)
https://www.diyaudio.com/community/threads/supravox-8-tlonken-while-city-of-22m-laid-flat.393338/
https://www.diyaudio.com/community/...rkaudio-alpair-5-2-in-evansound-3-75l.393909/
Last edited:
Hi Dominic
Haven't found the French design you might be referring to, but I found this one.
https://www.hifisentralen.no/forumet/threads/supravox-mltl-tad-2001-e-jmlc-600.96714/
IMO, your target width is too skinny for a 12 inch driver (reflections/inability to line/damp). Tried full horn width (518mm/20.4") rectangular box and then saw this pic and thought better of it (aesthetics) and rejockeyed everything to knock the corners off. I'd do something like this in your spot.
Preserves the driver-to-driver spacing from the pdf design so you have a chance to swipe the other crossover. If you use the same port, it's basically the same design save for a baffle step freq diff of about 40Hz down in the modal region anyway. Needs lining and full cross-section damping material between port and driver just like the pdf design (which the .no link doesn't adhere to if you look hard). There's nothing magical in the MLTL-ness of the pdf design (which likely has more filling than any builders used) -- they're tall 2-ways in which axial standing waves are accounted for. The filling/resistive damping material makes it a sort of continuum of mixed behaviors, but your ears always should win those choices. Plan to experiment with that. FWIW.
Attachments
I have a few pages from an old French book of sketchy cab drawings for the 215. But definitely not based on sound theory like MJK.
Last edited:
Thank you all for your support.
@wchang It could be useful, even if old versions of 215 have different parameters compared to the 215GMF, I suspect the response will vary, especially in the bass region, but why not.
@grindstone Here is the link of the design I mentioned (very similar to the one you found from the Norwegian forum): https://forums.melaudia.net/showthread.php?tid=3002&page=9
Your suggestion of design is interesting, the look is very similar to the Jensen version of the cabinet for 285GMF but without the Jensen vent, if I well understand. See discussion: https://www.diyaudio.com/community/threads/supravox-285-gmf-cabinet-size.137345/
I would like to explain the reason of my requests. The height topic is because I would like to have my ears (height about 100cm) at the same level of the area in between the woofer and the driver center, with current design (MLTL posted) my ear would be at the same level as the woofer, probably it’s a stupid concern I know…
The topic of the width instead is because normally I use to like and prefer sound of speakers with small baffle, since they are more inclined to “disappear”, maybe is mostly a psychological thing linked to eye more than the sound…
I use to like the sound of MLTL and TQWT charge, I agree with you about the idea to experiment, infact I just need to decide where to start from, then of course building a proto box is not a problem. First step is to choose the woofer to pick up, in this moment I’m more oriented on 285GMF but if the 215GMF would allow to use an enclosure with more reasonable dimensions (and shape) maybe is better to go for the 215…
It would be interesting also to know if someone had the chance to compare these two woofers, especially up to 1Khz, is the 285 good enough in this area despite its big diameter or the 215GMF has to be preferred for this frequency range?
@wchang It could be useful, even if old versions of 215 have different parameters compared to the 215GMF, I suspect the response will vary, especially in the bass region, but why not.
@grindstone Here is the link of the design I mentioned (very similar to the one you found from the Norwegian forum): https://forums.melaudia.net/showthread.php?tid=3002&page=9
Your suggestion of design is interesting, the look is very similar to the Jensen version of the cabinet for 285GMF but without the Jensen vent, if I well understand. See discussion: https://www.diyaudio.com/community/threads/supravox-285-gmf-cabinet-size.137345/
I would like to explain the reason of my requests. The height topic is because I would like to have my ears (height about 100cm) at the same level of the area in between the woofer and the driver center, with current design (MLTL posted) my ear would be at the same level as the woofer, probably it’s a stupid concern I know…
The topic of the width instead is because normally I use to like and prefer sound of speakers with small baffle, since they are more inclined to “disappear”, maybe is mostly a psychological thing linked to eye more than the sound…
I use to like the sound of MLTL and TQWT charge, I agree with you about the idea to experiment, infact I just need to decide where to start from, then of course building a proto box is not a problem. First step is to choose the woofer to pick up, in this moment I’m more oriented on 285GMF but if the 215GMF would allow to use an enclosure with more reasonable dimensions (and shape) maybe is better to go for the 215…
It would be interesting also to know if someone had the chance to compare these two woofers, especially up to 1Khz, is the 285 good enough in this area despite its big diameter or the 215GMF has to be preferred for this frequency range?
So what does happen to the response if the OP reduces the dimensions from 108 cm X 38 cm X 40 cm to 90 cm X 32 cm X 40 cm? As expected, you lose some bass response. The 1st attachment shows the changes. You can regain the loss if the depth is increased (90 X 32 X 57) so the original volume is maintained. See the 2nd attachment.
Do note that the amount of filling changes for each option.
Original (108 X 38 X 40) = 1.199 lb
90 X 32 X 40 = 1.009 lb
90 X 32 X 57 = 1.438 lb
Hope this helps.
Do note that the amount of filling changes for each option.
Original (108 X 38 X 40) = 1.199 lb
90 X 32 X 40 = 1.009 lb
90 X 32 X 57 = 1.438 lb
Hope this helps.
The MLTL-like cab in Fig 30 ter from that book, has about the same volume as OP would like. Its line length eyeballs about 1.7m for a quarterwave frequency of 50hz before considering ML from the narrow slot. A simple cab like I've drawn below, again of similar volume, has line length about 2m and taper effect (ML) about 1.4X expansion, for a quarterwave frequency of 30hz. I'd suggest a slot of 4-6cm which can be easily shimmed to tune the port output. (The divider board also provides critical bracing for so large a cab.)
And offset the TAD to align acoustic centers by ear.
And offset the TAD to align acoustic centers by ear.
Last edited:
Thank You!
That's exactly what I was looking for! 🙏
Is this a software or an excel sheet? Where is possible to download it?
The simulation with 90x32x57 looks promising, the response on bass is a bit different from the original one, looks smoother and it's missing the little bump below 40Hz but maybe it can be fixed by changing the port length or some other parameters... Or maybe is even better not to have it, who knows... 😅
The software is a Mathcad spreadsheet by Martin King. It requires the Mathcad program which is expensive. However, there's an old version (IIRC Mathcad 8) that can be downloaded for free. If you're interested check out Martin's website: http://www.quarter-wave.com/.
Thank you for the link to the French forum project (cool forum/people, but I'm never really 100% sure what they're saying through my English translator...seems to call horns pavillions...or maybe cabinets and horns). It's also funny how circular the forum world is as I followed the .fr link back here to the same 285GMF box sizing thread from '09. It is both good and bad that we have all found each other 🙂 It is also useful to see this is cedus' first MJK project and that philou is the one aiding him in his sims. What is perhaps funniest is that I manually translated all 11 pages before I saw there is a forum menu to change to English :|
Yes, my design was based on the same single port as in the original pdf you posted, I just didn't draw that circle because I assume you'd choose an equivalent port closer to a standard size than that one (4.724" dia/12cm) or tune it differently yourself. My design duplicated the original project as closely as possible in the available height with my arbitrary personal choices of width and depth and profile to better match Autotech horn curvature for aesthetics.
I'd lower the overall height, too, if it were my project--the height is fine. With an outboard horn, there is flexibility to physically move driver alignment after build. You are on your own on the cabinet width--my own experiences have shown the adjacent wall reflections to be more objectionable, but people differ. I used to like/build really skinny things too (that's how I learned this 🙂 ) In a 2-way that has to play high (or certainly for fullrangers), there is a cost in clarity that is too much for me. Speakers...everything costs something. Also, diffraction-wise, large bevels are smoother than sharp (or even small radiused) corners, FWIW. I do understand, however, that skinny is easier to get domestic approval on than wide and I understand your preference for skinny as it affects presentation of space.
I also have Martin's sheets but am not a commercial customer so limited to the personal use condition and have not shared results. If you are interested, try Hornresp or several versions of Akabak (all free) to experiment with these things yourself. We can get you started with the inputs and there are videos on YT and so on. All the different tools are convenient for certain things and not for others, but they all have value IMO. There are more free tools, too, if you are interested--it's a detail vs. convenience tradeoff.
If it were my project, I would choose the 285GMF. Bigger always has to move less than smaller (for same output), that one is quite-light ~20g Mms, and the builders report very favorably. Please keep us posted if you feel inclined to do so.
Update: I contacted Martin on the Facebook group, he told me that his Mathcad whorksheet is not supported anymore since it was running on old version of Mathcad (8) and to update it would require too much work. So basically he suggested me to use Hornresp.
So now I think I need to start exploring Hornresp and I see if I'll be able to make some simulations 😅
Anyway, how do you see these results? The original design is 5dB louder below 40Hz, this kind of response has to be preferred? Normally the calculation takes already in consideration the room, so my feeling is that I would prefer to have these few additional dB at low frequencies in the "real in room" response...
Any opinion?
So now I think I need to start exploring Hornresp and I see if I'll be able to make some simulations 😅
Anyway, how do you see these results? The original design is 5dB louder below 40Hz, this kind of response has to be preferred? Normally the calculation takes already in consideration the room, so my feeling is that I would prefer to have these few additional dB at low frequencies in the "real in room" response...
Any opinion?
Hornresp doesn't do reflections--for that you'll need Akabak or something else. Once you get into room reflections deeper, you'll understand how simple most of the tools (incl Martins 0->1 scale for all freqs) treat them; my point is that such can wait. Plan to get some "room gain" (actually you can use Basta! to explore that, maybe, as well as Akabak), but you don't need it built-in to be able to compare designs. Actually, you can look at the top of the MJK pdfs vs the bottom of the pdfs and see what rooms do...for now, just know that whatever you sim will not be pristine once it hits a real room.
Hornresp has a slightly quirky interface to most new users--we can help more with that, later. It models a lot of things but you only care about one of them right now. The biggest quirk to new users is that the interface is modal--what options you have depends on what configuration things are in. That used to be how a lot of software worked, but it seems newer/younger users find it foreign. Just know that you really have 2 basic tasks -- to get driver information entered, and to specify your geometry for the right model type. That's it. Further, the model (in your case), is really only going to consist of 3 segments. Each segment has only a cross sectional area and a length. That's it--you'll model the whole thing the long way (height direction). Same thing with Martin's sheets--there's just a lot more sections in them. You just need to get the cross sectional area and lengths--and out of all those, you group them into 3 pieces: (1) distance from top down to driver, (2) distance from driver to port, and (3) distance from port to bottom. All dimensions are internal, BTW, and HR works in cm & cm2.
Take a look at this (26 minute) YT video--there's some incorrect stuff but, from what I've skimmed, it'll show you (in English, if that's OK--is it or do you need .it stuff?) the basic process. Try to take the geometry of one of the designs in the pdf and see if you can get it into HR after viewing the vids. Found another 6.5 minute YT video for more of a sub-type TL.
Here's a screenie of a 4-segment TL out a v2 of @haraldkasper's manual (which I can't find--can only find v1) I guess (?) Harald's manual is now on Facebook, but I don't "do" FB.
Start there, anyway. If you have questions, post the input screen (main screen) of what you have and we can help.
PS While I'm here, I overlaid those two.
Hornresp has a slightly quirky interface to most new users--we can help more with that, later. It models a lot of things but you only care about one of them right now. The biggest quirk to new users is that the interface is modal--what options you have depends on what configuration things are in. That used to be how a lot of software worked, but it seems newer/younger users find it foreign. Just know that you really have 2 basic tasks -- to get driver information entered, and to specify your geometry for the right model type. That's it. Further, the model (in your case), is really only going to consist of 3 segments. Each segment has only a cross sectional area and a length. That's it--you'll model the whole thing the long way (height direction). Same thing with Martin's sheets--there's just a lot more sections in them. You just need to get the cross sectional area and lengths--and out of all those, you group them into 3 pieces: (1) distance from top down to driver, (2) distance from driver to port, and (3) distance from port to bottom. All dimensions are internal, BTW, and HR works in cm & cm2.
Take a look at this (26 minute) YT video--there's some incorrect stuff but, from what I've skimmed, it'll show you (in English, if that's OK--is it or do you need .it stuff?) the basic process. Try to take the geometry of one of the designs in the pdf and see if you can get it into HR after viewing the vids. Found another 6.5 minute YT video for more of a sub-type TL.
Here's a screenie of a 4-segment TL out a v2 of @haraldkasper's manual (which I can't find--can only find v1) I guess (?) Harald's manual is now on Facebook, but I don't "do" FB.
Start there, anyway. If you have questions, post the input screen (main screen) of what you have and we can help.
PS While I'm here, I overlaid those two.
Attachments
Hello Wchang, I noticed only now your message. Your suggestion of a folded TL with 215 woofer (RTF version) is extremely interesting. Which baffle size you considered? I saw in the book (fig 30 ter) the baffle is 50cm, way too big for me. I would like to keep it as small as possible, with the 215 woofer a baffle of about 30cm would be the best...
If I understand how to do it with Hornresp I will try to make some simulations with folded TL design
Last edited:
Thanks to the help of Grindstone (thank you again! 🙏) I've been able to simulate on Hornresp the MLTL box of the original project.
As you can see the response is very different from the one calculated with the Mathcad sheet from Martin.
How to interpretate these results? Is it just because Martin calculator takes in consideration the room behavior, at least more than hornresp?
As you can see the response is very different from the one calculated with the Mathcad sheet from Martin.
How to interpretate these results? Is it just because Martin calculator takes in consideration the room behavior, at least more than hornresp?
Attachments
For use as woofer I don't think baffle width mattered much (but stand ready to be corrected). Folding the line has the advantage of acting as low-pass filter on higher-order resonance spikes (much less stuffing!) as well as producing additional quarterwave and halfwave resonances on just part of the line delimited by boundaries. At least that's what I think is happening. Tapering a TL lowers depth by a lot (per MJK), a form of mass-loading (according to @Scottmoose) and saves significant space.
(1) See the Path 0.0 atop your power resp plot? Need to add path (dist betw sources driver & port) as "Auto" (Tools->Options->Auto)
(2) HR is displaying POWER response, MJK is displaying PRESSURE response. Pressure resp is avail by axial angle in HR under Directivity menu after choosing which output
The filling you add in the Loudspeaker Wizard does not apply after you return to the main window, but you can view the results by choosing Power instead of Schem in the same LS Wiz window--that will include the effects of the filling. Alternatively, maybe it's fast/easy for someone to zero-out the filling in the MJK sheets so you can see apples-to-apples? I can do it too, later--just have to get up off my tail & go start an old machine.
If the question becomes how do we get pressure + filling in HR, the answer is we don't--it doesn't do that. Akabak will. Each tool has it's own pros/cons.
(2) HR is displaying POWER response, MJK is displaying PRESSURE response. Pressure resp is avail by axial angle in HR under Directivity menu after choosing which output
The filling you add in the Loudspeaker Wizard does not apply after you return to the main window, but you can view the results by choosing Power instead of Schem in the same LS Wiz window--that will include the effects of the filling. Alternatively, maybe it's fast/easy for someone to zero-out the filling in the MJK sheets so you can see apples-to-apples? I can do it too, later--just have to get up off my tail & go start an old machine.
If the question becomes how do we get pressure + filling in HR, the answer is we don't--it doesn't do that. Akabak will. Each tool has it's own pros/cons.
Attachments
Last edited:
When I posted the 90 X 32 X 57.pdf I only filled out the Part 1 spreadsheet parameters to match the original MLTL parameters. That's why you see such a large difference in output level. Here's the updated system response and the complete .pdf with Part 1 and Part 2 parameters completed. To verify I didn't make any mistakes you should compare all the parameter values from both .pdfs (original and 90 X 32 X 57 Parts 1 and 2).
