I am a fan of waveguides and BEM optimized baffle shapes etc. However, these things are not trivial to make. The OSMC has a single waveguide that can be had by uploading a 3D file to a CNC service, and that's already too much hassle for most builders (people buy the waveguides off my stock instead). I'd strongly recommend to avoid such 3D parts if you want people to actually build the thing.
I think times are changing. At least here in Europe a lot of 3D print services have established since the OSMC came to life. And would-be builders will need support in many ways, on applying damping, constructing resonance-free and testing of the result. I can imagine the ‘how to’ in using 3D-parts is part of that support.
I dont want to come across as rude or upset anyone but this thread seems to be more heading down the road of 'difficult to build and rather expensive monkey coffin'.
I was under the impression that a big part of the thought process was to avoid many of the hurdles many saw in the osmc (cost, difficult finding drivers/waveguides).
To my mind there are a lot of drivers that measure very well objectively and punch way above their price point. Many of course are harder to use and need more attention in xo etc but surly this group would be more than up to the challenge?
I would have thought choosing desired limits in size, cost eyc and trying to build within that framework for maximum performance would be the logical starting point.
I was under the impression that a big part of the thought process was to avoid many of the hurdles many saw in the osmc (cost, difficult finding drivers/waveguides).
To my mind there are a lot of drivers that measure very well objectively and punch way above their price point. Many of course are harder to use and need more attention in xo etc but surly this group would be more than up to the challenge?
I would have thought choosing desired limits in size, cost eyc and trying to build within that framework for maximum performance would be the logical starting point.
In my opinion, additionally to the more sophisticated Andy's retro 3 way loudspeaker, for the other group of interested builders,
I'd suggest a simpler to execute and cheaper to build, an entry level version of it.
As an example:
Visaton W250S-8= 90€
Visaton W130S-4= 39€
SS H2606/920000=44€
passive x/o per side=90€
SUBTOTAL=263€
Cabinet vented 70 lit (Fb=33 Hz)
F3/F6/F10=39/31/27 Hz
SPL/2,83V/1m=89dB (-3dB already calc. baffle step loss, sensitivity estimated by the TSP)
I'd suggest a simpler to execute and cheaper to build, an entry level version of it.
As an example:
Visaton W250S-8= 90€
Visaton W130S-4= 39€
SS H2606/920000=44€
passive x/o per side=90€
SUBTOTAL=263€
Cabinet vented 70 lit (Fb=33 Hz)
F3/F6/F10=39/31/27 Hz
SPL/2,83V/1m=89dB (-3dB already calc. baffle step loss, sensitivity estimated by the TSP)
Last edited:
Speaking of a modular setup, here is the sketch of an idea I had some time ago. This concept would allow comparison of narrow and wide baffles.
I was not referring to 3D printing specifically. The prints I see at work are cool and stuff, but they are not pretty in the sense that their finish would be up to snuff for something that I'd put in my living room. The OSMC parts are made by 3D CNC machining, which gives a very nice finish.
Also, as I already wrote in my previous message, it's absolutely not difficult to upload a file to some 3D maker service and press the "order" button -- but that's not the point. The majority of OSMC builders (all?) don't get their 3D parts from a 3D service provider, but instead they buy the parts from me. This leads me to conclude that a design without any custom 3D parts might be better if you want people to pick up the project and build the speaker.
I could not find the cad file when I checked a while ago, I think the link had broken. Also, would a local maker really charge less than you?
@andy19191 in your last round up email you mentioned the aim was to produced a "modern midfield monitors" and this Genelec etc as examples. Doe this by definition presume a flat on axis frequency curve and a power response that slopes downwards approx. 6db from 100hz to 10khz and also well controlled off axis directivity i.e. nice even spacing of off axis frequency curves.
I think once the general form factor is agreed on, and potential a budget, then some of these type of parameters may be next on the list. As a starter for ten -
On axis frequency - Flat +/-2db from 40Hz to 20khz
Power response - +/-2db from a line with 6dB drop over 100hz to 10khz (approx -0.9dB/octave slope)
Off axis response - (this is a stab on the dark but ....) - 0 to +/-30degs vertical and horizontal maximum 5dB from on axis between 100hz and 10khz with no overlap of 0,10,20,30 deg frequency curves and no significant peaks/troughs.
Minimum sensitivity - 86dB/W/m
Desired sensitivity - 90dB/W/m
Nominal Impedance - greater than 6ohms
Minimum Impedance - greater than 3ohms
Phase coherence - at least 1 octave either side of the crossover frequencies
There are probably a load more parameters but this would be my starting "ideal" point to work towards.
However, one options may be to look at different "voicings" to suit - close to wall use, BBC dip, Smiley Face, Bass Boost etc. Just a wild card to throw into the mix, I for one would like to see a close to wall option given that most UK living rooms are approx 12ft x 16ft with 8-9ft ceilings so pretty small and so getting 3ft form any boundary for the ideal sound is almost impossible.
Just a few thoughts to add into the mix and as a systems engineer it sounds like we need a systems requirements document that can then be used as the baseline to judge any designs against (and tested for compliance).
I think once the general form factor is agreed on, and potential a budget, then some of these type of parameters may be next on the list. As a starter for ten -
On axis frequency - Flat +/-2db from 40Hz to 20khz
Power response - +/-2db from a line with 6dB drop over 100hz to 10khz (approx -0.9dB/octave slope)
Off axis response - (this is a stab on the dark but ....) - 0 to +/-30degs vertical and horizontal maximum 5dB from on axis between 100hz and 10khz with no overlap of 0,10,20,30 deg frequency curves and no significant peaks/troughs.
Minimum sensitivity - 86dB/W/m
Desired sensitivity - 90dB/W/m
Nominal Impedance - greater than 6ohms
Minimum Impedance - greater than 3ohms
Phase coherence - at least 1 octave either side of the crossover frequencies
There are probably a load more parameters but this would be my starting "ideal" point to work towards.
However, one options may be to look at different "voicings" to suit - close to wall use, BBC dip, Smiley Face, Bass Boost etc. Just a wild card to throw into the mix, I for one would like to see a close to wall option given that most UK living rooms are approx 12ft x 16ft with 8-9ft ceilings so pretty small and so getting 3ft form any boundary for the ideal sound is almost impossible.
Just a few thoughts to add into the mix and as a systems engineer it sounds like we need a systems requirements document that can then be used as the baseline to judge any designs against (and tested for compliance).
I think with some Italian brand drivers (aforementioned LaVoce 12”, Sica 5.5H1.5CP) and that H2606 (excellent choice!) you could get better results at about the same price.
The file is attached to some post buried somewhere in diyAudio, possibly in the OSMC thread. I should add it to the github repo. Yes, I keep buying larger numbers of waveguides, which helps with the cost per part.
This is getting off-topic here. Please take it to the OSMC thread if further OSMC-specific discussion is needed.
That will not make much sense depending on the off-axis dispersion and power response. But I guess we shouldn't jump the gun before we don't know if / how the project will go forward, and what it want's to achieve.
I want to address the risk of cost creep. There has been some discussion about whether the SB34NRXL is too expensive for this project.
If we assume the SB34NRXL75-6 is the chosen woofer, at $315 USD each, then what is the budget available for the other 4 drivers, and is this a reasonable budget?
If the upper end driver budget is $1275, that leaves $645 to purchase two tweeters and two midrange drivers.
Some reasonable mid drivers might be:
ScanSpeak Discovery 15M/4624G $ 94
Seas MCA15RCY (H1262) $109
SB15NBAC30 $ 81
Many other options from Wavecor, Visiton, Dayton
I am focusing on 5.5" drivers because it is possible to match this size driver with a non-waveguide 1" dome and get a good DI and power response if the baffle shape is optimized. Of course the design team will look at all options.
There are a lot of reasonable tweeters:
SB26ADC or SB26CDC $ 60
The SB26 metal dome tweeters are very adaptable to waveguides, and several excellent waveguides have already been developed
Seas 27TAC/GB $ 93
SEAS 27TBCD/GB-DXT $106
Waveguide is included
Other options from Wavecor, Morel, Seas, ScanSpeak, some of which have waveguides.
So if we assume the mid driver will be $150, and the tweeter w/guide will be $120, the overall driver budget is $1170... well within the requirements.
If we assume the SB34NRXL75-6 is the chosen woofer, at $315 USD each, then what is the budget available for the other 4 drivers, and is this a reasonable budget?
If the upper end driver budget is $1275, that leaves $645 to purchase two tweeters and two midrange drivers.
Some reasonable mid drivers might be:
ScanSpeak Discovery 15M/4624G $ 94
Seas MCA15RCY (H1262) $109
SB15NBAC30 $ 81
Many other options from Wavecor, Visiton, Dayton
I am focusing on 5.5" drivers because it is possible to match this size driver with a non-waveguide 1" dome and get a good DI and power response if the baffle shape is optimized. Of course the design team will look at all options.
There are a lot of reasonable tweeters:
SB26ADC or SB26CDC $ 60
The SB26 metal dome tweeters are very adaptable to waveguides, and several excellent waveguides have already been developed
Seas 27TAC/GB $ 93
SEAS 27TBCD/GB-DXT $106
Waveguide is included
Other options from Wavecor, Morel, Seas, ScanSpeak, some of which have waveguides.
So if we assume the mid driver will be $150, and the tweeter w/guide will be $120, the overall driver budget is $1170... well within the requirements.
When did the SB34 become the only woofer choice???
This immediately rules out mid-domes, as well as skews high both the price and size of the whole speaker in a way I personally don't think is at all of a broader appeal. If we're jumping somehow to cone mids and tweeter lists, well, the thread is losing structure and any order.
This immediately rules out mid-domes, as well as skews high both the price and size of the whole speaker in a way I personally don't think is at all of a broader appeal. If we're jumping somehow to cone mids and tweeter lists, well, the thread is losing structure and any order.
SB34NRXL is not the only woofer choice. But it is at the upper end of the cost constraint for this project. A $400 - $500 woofer is just not going to be affordable with this project, but a $315 one will work, barely. I was just pointing out that a $315 woofer should not be excluded at this point.
We are not jumping to anything, and nothing has been decided.
When the requirements for some new project/product are established, it is critically important that the people establishing the requirements have a detailed understanding of what is feasible. Many projects have been doomed from the start because the requirements were overly ambitious or even physically impossible. This is a common problem with large complex projects like new aircraft designs, power plants, new microprocessors... A speaker design is far far less complex than these kinds of things, but still we need to be vigilant that our requirements are achievable.
The way to avoid unrealistic requirements is to preliminary conceptual studies, or "what if" studies. That is what I was doing... Assuming we use the most expensive woofer on the table, is it possible to come up with a reasonable 3-way speaker concept that is within the budget? Answer: yes.
We are not jumping to anything, and nothing has been decided.
When the requirements for some new project/product are established, it is critically important that the people establishing the requirements have a detailed understanding of what is feasible. Many projects have been doomed from the start because the requirements were overly ambitious or even physically impossible. This is a common problem with large complex projects like new aircraft designs, power plants, new microprocessors... A speaker design is far far less complex than these kinds of things, but still we need to be vigilant that our requirements are achievable.
The way to avoid unrealistic requirements is to preliminary conceptual studies, or "what if" studies. That is what I was doing... Assuming we use the most expensive woofer on the table, is it possible to come up with a reasonable 3-way speaker concept that is within the budget? Answer: yes.
If the SB34NRXL75-8 at $315 is the high end option, then for a 12", the @waxx recommended SB34NRX75-6 at $223.60 (again Madisound) has to be considered the more practical option. That's a near-$200 savings on a pair, which is significant overall.
Plus it can be used in a sealed box <75 liters, which offers greater flexibility, as sealed box volume can be adjusted up/down 10% or more with nary any effect on bass response. It's also preferable for vinyl users vigilant about subsonic grunge from warps, less-than-perfect turntables, etc.
Plus it can be used in a sealed box <75 liters, which offers greater flexibility, as sealed box volume can be adjusted up/down 10% or more with nary any effect on bass response. It's also preferable for vinyl users vigilant about subsonic grunge from warps, less-than-perfect turntables, etc.
It's always good to look at a lot of options and think about even the craziest id in the prestudy phase we're in, that is certainly true and even essential in my opinion to get to a good result and find real innovation when needed.
My guess...Once the technical requirements are decided, the likely next step will be to narrow down the narrow list of potential 12" drivers.
Yes agreed. It is a very good woofer, both from objective measured standard and a subjective sound quality (based on my personal experience). Sealed box simplifies a lot of things, and if it were my decision, I would go sealed box.
Another very nice sealed-box 12" we will need to include on the list is the Dayton RSS315HF, both the 4 and 8 Ohm versions.
SB34RNXL is 240 euros at THLP (the price here in France when you are registred). I don't know the average incomes in USA but I suspect it is maybe higher, at least for the population who DIY in that hobby. So feasible for North AMerica at 315 USD ?
So the SB34RNXL is perhaps not a bad choice if the difference in qualilty matters (i.e. cut off higher than 200 hz or lowish sealed volume).
That's terrible I am totally stopped on the naming phase from one day..... Yesterdays' Monkey Coffins Always (YMCA) ; more genttle than the coffin word but a blink eyes : Where Are My Monkeys (WAMM) ; The Afforbale Reasonably Zenithal Astonishing Neutral (TARZAN)... ; My Open Wonderful Gratefully Loudspeakers Inspection : (MOWGLI... very 70's too as totally jungle like)
So the SB34RNXL is perhaps not a bad choice if the difference in qualilty matters (i.e. cut off higher than 200 hz or lowish sealed volume).
That's terrible I am totally stopped on the naming phase from one day..... Yesterdays' Monkey Coffins Always (YMCA) ; more genttle than the coffin word but a blink eyes : Where Are My Monkeys (WAMM) ; The Afforbale Reasonably Zenithal Astonishing Neutral (TARZAN)... ; My Open Wonderful Gratefully Loudspeakers Inspection : (MOWGLI... very 70's too as totally jungle like)
Last edited:
I suggest to let @andy19191 figure out the design targets. Without these it's meaningless to discuss specific drivers, parts and their "numbers".
Yeah the combo bass + midwoofer choice first inside that targeted pricing... It will be the fundation of what the rest depends (size, total price with passive filter). If Andy doesn't know, people should discuss the trade off of the feasible combos as the illustrated examples of this thread.
From the beginning I think the drivers should be chosen to be compatible with 12 dB slope because of the price. Sometimes you can find crossovers simple enough thankst he drivers. You can see that sometimes from T Gravsen, P Carmody, K Campbel, etc.
Edit: this 12" Dayton is expensive in Europe...
From the beginning I think the drivers should be chosen to be compatible with 12 dB slope because of the price. Sometimes you can find crossovers simple enough thankst he drivers. You can see that sometimes from T Gravsen, P Carmody, K Campbel, etc.
Edit: this 12" Dayton is expensive in Europe...
Last edited: