I'm just thinking about improving my loudspeakers: Braun LS200. I'm curious if the drivers' positions on the speaker baffles were rearranged to follow D'Appolito's MTM layout. Is it absolutely considered an upgrade? I mean, is there any obvious improvement?
Here's a beautiful image of the LS200 from its brochure.
My idea for converting them to MTM style is shown below.
Here's a beautiful image of the LS200 from its brochure.
My idea for converting them to MTM style is shown below.
Last edited:
The idea behind D'Appolito's MTM arrangement is the control of the vertical polar radiation pattern. In general it works best at the higher Midrange to tweeter frequencies, particularly if you can get the mid driver separation down close to the cross-over frequency wave lengths. It's less useful for woofers, simply because the wavelengths are such that it doesn't make a significant difference. You asked if it was an "absolute ... upgrade", in this case, the answer is no. It might be, but it would require more investigation, regarding cross-over frequencies, slopes, driver separation etc....
Do you know xover freq of said loudspeaker, as well as all drivers physical dimension?
I disagree with DCtodaylight on the fact mtm works best on higher midrange to tweeter freq. At least stated that way. In fact the higher you try to achieve it in freq the more you'll have to accept side lobes and their nastys in radiation pattern.
In advanced verticaly stacked design inspired by D'Apolitto approach it is possible to really control directivity over wide bandwith but it's outside the scope of the initial question...
One of the often overlooked advantage of MTM is it can help to mitigate floor/ceilling bounce circa 200hz (+/-100hz) because it will generate two smaller notch at different freq rather than one big at one center freq.
The rule of thumb if you want both drivers to behave as one ( omnidirectional radiation pattern) is to use centre to centre spacing = 1/4 wavelength of xover frequency.
This can be relaxed to 0.33 and still offer an omnidirectional behavior. You can 'shape' the vertical polar pattern at xover freq by varying the ratio of wavelength at xover freq to define the center to center distance of a pair of omni source:
a 0.5 ratio you would have a 83* coverage on vertical pattern at xover freq, 0.462 offer 90*, 0.43 circa 100*, all this value offering no lobes ( secondary main 'axis' of radiation which add to 'main on axis radiation pattern ) hence an even 'main' lobe radiation.
Past 0,5 things change at 0.53 you would have a main 77,5* coverage lobe on axis but with side lobe ( pointing up (toward ceiling/+90*) and down (toward floor/ -90*) -20db relative to on axis, 0,6 would give 67,1* with side lobes -10db ( and lobe axis around +/-85*) and finaly with 1 wavelength you would have a 38,7* coverage on axis but with ( very wide) side lobes pointing toward +/-70* at same level than on axis (0db).
Presscot, i fear the weak link will be the mid dome: they rarely accept to go lower than 600hz.
At 600hz one wavelength equal ~57cm. 57cm/4~14,3cm.
This mean centre of dome to centre of woofer= 14,3cm for 1/4wave length., approximately 20cm if 1/3.
More doable.
There will be another issue in the way you locate tweeter. If on side then you might have issue with xover, or at least have to redo one.
If using dsp and FIR with very steep slope it would less be an issue...
I disagree with DCtodaylight on the fact mtm works best on higher midrange to tweeter freq. At least stated that way. In fact the higher you try to achieve it in freq the more you'll have to accept side lobes and their nastys in radiation pattern.
In advanced verticaly stacked design inspired by D'Apolitto approach it is possible to really control directivity over wide bandwith but it's outside the scope of the initial question...
One of the often overlooked advantage of MTM is it can help to mitigate floor/ceilling bounce circa 200hz (+/-100hz) because it will generate two smaller notch at different freq rather than one big at one center freq.
The rule of thumb if you want both drivers to behave as one ( omnidirectional radiation pattern) is to use centre to centre spacing = 1/4 wavelength of xover frequency.
This can be relaxed to 0.33 and still offer an omnidirectional behavior. You can 'shape' the vertical polar pattern at xover freq by varying the ratio of wavelength at xover freq to define the center to center distance of a pair of omni source:
a 0.5 ratio you would have a 83* coverage on vertical pattern at xover freq, 0.462 offer 90*, 0.43 circa 100*, all this value offering no lobes ( secondary main 'axis' of radiation which add to 'main on axis radiation pattern ) hence an even 'main' lobe radiation.
Past 0,5 things change at 0.53 you would have a main 77,5* coverage lobe on axis but with side lobe ( pointing up (toward ceiling/+90*) and down (toward floor/ -90*) -20db relative to on axis, 0,6 would give 67,1* with side lobes -10db ( and lobe axis around +/-85*) and finaly with 1 wavelength you would have a 38,7* coverage on axis but with ( very wide) side lobes pointing toward +/-70* at same level than on axis (0db).
Presscot, i fear the weak link will be the mid dome: they rarely accept to go lower than 600hz.
At 600hz one wavelength equal ~57cm. 57cm/4~14,3cm.
This mean centre of dome to centre of woofer= 14,3cm for 1/4wave length., approximately 20cm if 1/3.
More doable.
There will be another issue in the way you locate tweeter. If on side then you might have issue with xover, or at least have to redo one.
If using dsp and FIR with very steep slope it would less be an issue...
Last edited:
Yes, the xover frequencies are published at 450/4,500 Hz. But, according to the simulation, the electrical xover points are 350/3,500 Hz. So, I’m not sure which is more accurate.
Electrical is only a fraction of the outcome: the driver natural behavior ( into the box) is took into account and the real meaningful value is the acoustical one ( the driver+box+electrical xover).
So this could be the 450hz/4500hz which are 'the one' that matters... what size are the woofers? It can matter about directivity.
So this could be the 450hz/4500hz which are 'the one' that matters... what size are the woofers? It can matter about directivity.
Actually, I have two plans for the improvement project.
Plan I is to convert them to 3.5 ways.
Plan II is to rearrange drivers’ positions following D’Appolito’s style (this thread).
However, I rethink and realize that maybe plan I and II cannot be done simultaneously because if performing plan II first, the distance between each woofer and midrange driver will be equal. Therefore, each woofer should produce the same sound, or have the same xover. Hence, the original 3-way system may be preferred.
And if performing plan I first, the equal distance between each woofer and midrange may not become matter. And it might be better to have unequal distance as each woofer will play on different xover, for the 3.5-way system.
Plan I is to convert them to 3.5 ways.
Plan II is to rearrange drivers’ positions following D’Appolito’s style (this thread).
However, I rethink and realize that maybe plan I and II cannot be done simultaneously because if performing plan II first, the distance between each woofer and midrange driver will be equal. Therefore, each woofer should produce the same sound, or have the same xover. Hence, the original 3-way system may be preferred.
And if performing plan I first, the equal distance between each woofer and midrange may not become matter. And it might be better to have unequal distance as each woofer will play on different xover, for the 3.5-way system.
don't do it.
it makes perfect sense how braun laid out the drivers:
tweeter should be at ear height.
NEVER put tweeter and midrange side by side. that creates all sorts of comb filtering when moving around (or even just moving your head).
a real d'appolito concept is very difficult to realize because of required crossover frequency wavelength / max driver distance restrictions.
it makes perfect sense how braun laid out the drivers:
tweeter should be at ear height.
NEVER put tweeter and midrange side by side. that creates all sorts of comb filtering when moving around (or even just moving your head).
a real d'appolito concept is very difficult to realize because of required crossover frequency wavelength / max driver distance restrictions.
You’d need a second mid driver, then you could turn it into WMTMW.
As that is rather unlikely, it’s probably best to leave the thing alone. Never place mid and tweeter horizontally.
If you feel an urgent need to fiddle, measure all drivers individually 360 degree horizontally and vertically, feed the data into VituixCad and see if you can come up with a better crossover than Braun did.
As that is rather unlikely, it’s probably best to leave the thing alone. Never place mid and tweeter horizontally.
If you feel an urgent need to fiddle, measure all drivers individually 360 degree horizontally and vertically, feed the data into VituixCad and see if you can come up with a better crossover than Braun did.
I am confused why people didn’t recommend placing the midrange and tweeter horizontally despite the fact that many manufacturers including Braun did that in a lot of models.
This is the example of the LS200’s siblings—same series found in the same brochure.
Or these are its predecessors.
This is the example of the LS200’s siblings—same series found in the same brochure.
Or these are its predecessors.
Putting drivers up in the corners that way does give good looking single axis baffle simulation results, however it also tends to produce variance between angles so it's not always recommended. Not only that but some prefer the mid to tweeter lobing to happen vertically.
This doesn't mean that these can't sound good but you probably won't find many suggesting you do this.
This doesn't mean that these can't sound good but you probably won't find many suggesting you do this.
Many brands did that. The reasons were that they didn't know (not possible at Braun 😉), they wanted to save space or - like it was quite popular at that time - did it intentionally because a lot of people thought the IDEAL place for a speaker was on top of their living room cabinets and lay them on its side.
These speakers were designed at a time when real speaker science just started to evolve. In the pre- computer, pre- "measure anything" times. The few speaker developer where very unique, often strange people, that used a mixed measure/ listen /experimenting technique to do their magic. I have not heard a single German speaker at that time, that produced a good bass to my taste. If they could do one, they where not allowed to. I was spoiled by Altec, JBL and EV speaker, not to forget some British, like Celestion. There where some German "boutique speaker" builder, but that was more like a kind of snake oil cult.
Then, please see the economic background, these speaker had to be build very cheap and were sold at ridiculous high prices. Braun needed to have some audible improvement in the drawer, which they presented with huge marketing ballyhoo when sales of the old model dropped. Usually at the IFA (Internatiolnale Funk Ausstellung). So the medical docter, architect, lawer and whatever status seeking rich client would buy the new one and pass the older model to their children. That marketing worked quite a while, but not forever as we know. Braun, like B&O in Denmark, was about high quality, design and complicated electronics, but not high end sound. Sound was somehow secondary and not the aim of German engineering. Electronic guys thought of speakers as something primitive. I don't know if anyone can understand the mindset of that time.
About the placement of chassis, the marketing peoples opinion was always more important than a better, new construction. Germany was very much in a "we always did it like that, customers want it like that" mindset. The engineer doing the Braun speaker stuff could have done much better!
IMO you will not get a worse speaker if you realize your plan to move the chassis to W-M/T-W. If you mirror the M/T you keep the very small sweet spot, but that is how they where build. Staging should improve a little, because the two large woofer, placed like they are, don't help with that. Anyway, expect no miracle. Also, change the electrolitycs in the crossover, they add distortion when they age, even if not in the signal path.
Experimenting with a large capacitor (500-1200microF) in series with the woofers would be interesting, to maybe extend the bass and make it sound "dry".
It enables the voice coil to pull more current from the amp by lowering the impedance, in the low region before the drop off. Extending bass. A very rewarding, but often not fully understood principle.
To be honest, keeping the cabinet design with front grill and installing some nice 3-way kit would be the best option. A neat two way at the top with an active sub (or two) lower down, would be a nice idea, too. Sure, you will change a good kit in some way if you can not realize the baffle as planed, but compared to these old, in most cases aged and dried out chassis, the difference will be incredible.
Anyway, you may sell the original chassis as spare's for good money.
Then, please see the economic background, these speaker had to be build very cheap and were sold at ridiculous high prices. Braun needed to have some audible improvement in the drawer, which they presented with huge marketing ballyhoo when sales of the old model dropped. Usually at the IFA (Internatiolnale Funk Ausstellung). So the medical docter, architect, lawer and whatever status seeking rich client would buy the new one and pass the older model to their children. That marketing worked quite a while, but not forever as we know. Braun, like B&O in Denmark, was about high quality, design and complicated electronics, but not high end sound. Sound was somehow secondary and not the aim of German engineering. Electronic guys thought of speakers as something primitive. I don't know if anyone can understand the mindset of that time.
About the placement of chassis, the marketing peoples opinion was always more important than a better, new construction. Germany was very much in a "we always did it like that, customers want it like that" mindset. The engineer doing the Braun speaker stuff could have done much better!
IMO you will not get a worse speaker if you realize your plan to move the chassis to W-M/T-W. If you mirror the M/T you keep the very small sweet spot, but that is how they where build. Staging should improve a little, because the two large woofer, placed like they are, don't help with that. Anyway, expect no miracle. Also, change the electrolitycs in the crossover, they add distortion when they age, even if not in the signal path.
Experimenting with a large capacitor (500-1200microF) in series with the woofers would be interesting, to maybe extend the bass and make it sound "dry".
It enables the voice coil to pull more current from the amp by lowering the impedance, in the low region before the drop off. Extending bass. A very rewarding, but often not fully understood principle.
To be honest, keeping the cabinet design with front grill and installing some nice 3-way kit would be the best option. A neat two way at the top with an active sub (or two) lower down, would be a nice idea, too. Sure, you will change a good kit in some way if you can not realize the baffle as planed, but compared to these old, in most cases aged and dried out chassis, the difference will be incredible.
Anyway, you may sell the original chassis as spare's for good money.
Hey @presscot - You really need to build something, and I strongly encourage you to start with a pre-built kit.
Simple…..horizontal dispersion is what takes precedence and drivers arranged horizontally have a very uneven dispersion pattern across the passband they’re in.
While an MTM does reduce floor and ceiling bounce, it comes at a price of overall resolution since there are multiple forward lobes whose phase doesn’t fully align outside of the crossover region.
@presscot Your answer can be found in the Stereophile off-axis plots of traditional multi-way speakers. Even those that have excellent horizontal dispersion suffer above the tweeter axis. The issue has to do with the distance between the midrange and tweeter (or mid-woofer and tweeter). Consider a standard vertical alignment of a tweeter and mid-woofer or tweeter and middrange. As the listening angle goes up the midrange gets further and further away from the listening position, causing destructive interference. On the other hand, moving the mic horizontally the relative distance between tweeter and mid changes far less.
In a perfect world with perfect crossovers of infinite slope this is not a problem. If your low pass filter went to 2999.9999999 Hz and your Tweeter started at 3000 Hz and played NOTHING below that frequency then there's no problem. The limitations of crossovers however mean that the vertical alignment is often better.
In reading your other posts lately I get the impression you've not really understood how the effective driver distance is part of the final frequency and phase response. If you understood that you'd want to avoid having a horizontal mid/tweeter. 🙂
In a perfect world with perfect crossovers of infinite slope this is not a problem. If your low pass filter went to 2999.9999999 Hz and your Tweeter started at 3000 Hz and played NOTHING below that frequency then there's no problem. The limitations of crossovers however mean that the vertical alignment is often better.
In reading your other posts lately I get the impression you've not really understood how the effective driver distance is part of the final frequency and phase response. If you understood that you'd want to avoid having a horizontal mid/tweeter. 🙂
Last edited:
There would arise another problem from that. Okay, let's exclude the group delay, phase, abrupt dispersion step etc, a very steep x-over does indeed sound bad in the most cases because you start to hear each driver 'separately', the speak becomes very inhomogenous. Unless you have drivers with the same membrane material, not too different in diameter etc, that will be very obvious and annoying and not enjoyable. Sorry for nitpicking and I know it's not your point and while you would solve the interference problem perfectly but that is an aspect which is almost always ignored at first when someone plans very steep filters. To solve one problem perfectly unfortunately usually means a bunch of others poke their head out instead. Speakers are always about making compromises.
Yes, almost always.
"But.. but it measures perfectly!" 😳
😉
You're really trying hard to find examples that confirm something you like, right?
You could also try to understand why people did not recommend horizontal placing. But of course that would mean learning about physical principles ...
Best way to learn is to build, as @eriksquires suggested!
If you've ever watched what Danny Richie from GR Research does with factory speakers, then you know there's a way to upgrade those speakers, and that's not what you're planning. Strengthen the box with internal stiffeners, add damping to the sides (bitumen?), damping the speakers with polyester wadding, replace the connectors and wiring, measure the frequency range and improve or rework the crossover with better parts. Perhaps the ferrofluid in the drivers should also be replaced. If they are 20-30 years old, and have ferrofluid, it is probably already dried.