Help? designing a DIY loudspeaker

[DanielT]

Nobody said it can’t be done. It’s just not easy .

Exactly and the solution to it spells time and a great deal of commitment.

I'm completely dumbfounded when I see all the test baffles and measurements, modifications and so on Maarten made to arrive at the final result. I don't have the time that he does.

 

[Markie Warkie]

Quite a bit has been said about the logic (of lack thereof) of making DIY speakers, so let’s focus on the design specifics:

2 midwoofers playing the same frequency range is only a good idea as an MTM. If it’s going to be an TMM, you’ll need to make it a 2.5-way. It saves you the baffle step compensation . MTMs have stricter constraints on c-over frequency. Generally the lower the better.Also generally, you want a PR with more surface area than your woofers. So I would recommend at least 3 or 4. And save up for an active crossover, it will make things a lot easier, and you’ll save a ton of money on crossover components you’ll never use again.

Thanks so much for your response

A MiniDSP Flex should work quite well for an MTM. You can save a bit on the PR’s if you take some cheaper ones (check on AliExpress for instance), or just go reflex for a first project. As for the tweeter, I’m not sure if that’s the best choice for an MTM. The recommended crossover point seems to be > 2 kHz. I think for an MTM, I would choose a quite a bit lower X-over of 1.5 to 1.8 kHz, and to prevent interference lobing of the woofers, you would actually need something even lower, more like 1 to 1.3 kHz. You can cheat a bit by offsetting the tweeter a bit, to bring the woofers closer and raise the crossover frequency. Maybe SB TW29D-B or TW29R-B are a better choice, or some lower FS seas or ScanSpeak tweeters. Or something smaller, like SS D3004.

… and yes, imho, most MTMs have a way too high crossover frequency. It’s a compromisers, and not the best one.

thanks so so much for the replies and not straight up discouragement so I was looking at mini dsp 2x4hd since they use the same processor as the flex

also the MtM thing I'm gonna have to do abit more research on that and decided if the compromises are Worth the upside's

 

[voodooless]

thanks so so much for the replies and not straight up discouragement

Not at all. It's great fun. Just know what you start with

so I was looking at mini dsp 2x4hd since they use the same processor as the flex

Yes, they are very similar. The flex primarily has better DACs on board, otherwise, it's quite comparable. You could always upgrade later on, they have quite a good resale value.

also the MtM thing I'm gonna have to do abit more research on that and decided if the compromises are Worth the upside's

You could just make an active MMT with a passive .5 filter. It's not that many components, and saves you on DSP channels and amps. It will be simpler to get right, and you can use the tweeter you already envisioned.

 

[Markie Warkie]

Not at all. It's great fun. Just know what you start with

Yes, they are very similar. The flex primarily has better DACs on board, otherwise, it's quite comparable. You could always upgrade later on, they have quite a good resale value.

You could just make an active MMT with a passive .5 filter. It's not that many components, and saves you on DSP channels and amps. It will be simpler to get right, and you can use the tweeter you already envisioned.

i have a vintage denon AVCA11SR that has 125 X7 in 8Ohm or 180x7 in 6Ohm so amp channels is not the problem its mainly DSP channels i saw you could use a Rpi4 with Camila dsp and stuff but from what i see its very hacky

​

 

[fluid]

Here is a couple of well worked out designs with included passive crossover schematics that will provide good bang for your buck.

https://www.audiosciencereview.com/.../revel-m105-copy-diy-build.29465/post-1320364
This post has the Scanspeak woofer version and looks much better than it did with the SB15NBAC, not that there is anything wrong with the SB driver.
The port will give a reasonable amount of bass that can always be plugged later on if you get a subwoofer or bass modules.

https://www.donhighend.de/?page_id=6762
This one uses the GHP principal to give decent low bass without a port or passive radiator

There is also the Italian Stallion from the same page that uses a Sica coax.

Building a proven design that already has good measurements is a safe way to get good value for your time and money.

You can learn a lot from measuring and equalising any speaker you have or build to see how and what changes, finding the best position for the speakers, room treaments etc.

There are so many good existing small 2 way speaker designs that reinventing the wheel has limited value, unless it is purely for interest or the challenge.

 

[Markie Warkie]

Here is a couple of well worked out designs with included passive crossover schematics that will provide good bang for your buck.

https://www.audiosciencereview.com/.../revel-m105-copy-diy-build.29465/post-1320364
This post has the Scanspeak woofer version and looks much better than it did with the SB15NBAC, not that there is anything wrong with the SB driver.
The port will give a reasonable amount of bass that can always be plugged later on if you get a subwoofer or bass modules.

https://www.donhighend.de/?page_id=6762
This one uses the GHP principal to give decent low bass without a port or passive radiator

There is also the Italian Stallion from the same page that uses a Sica coax.

Building a proven design that already has good measurements is a safe way to get good value for your time and money.

You can learn a lot from measuring and equalising any speaker you have or build to see how and what changes, finding the best position for the speakers, room treaments etc.

There are so many good existing small 2 way speaker designs that reinventing the wheel has limited value, unless it is purely for interest or the challenge.

honestly, I'm a bit confused rn I do wanna build something of my own but at the same time that invites a lot of uncertainty the correct way would be to have the right equipment and so on

if I do go on I'm not gonna use an MTM design cause that would not fit my use case HT + Hifi and I would use a digital crossover
Now the problem is I don't know how to proceed honestly I'm completely lost

 

[fluid]

There are a number of ways to design a speaker, from purely by ear, on axis only, using a USB mic. The most up to date sensible method is to use an analogue mic with soundcard loopback for accurate time measurements. Full dual plane polar measurements and the design being done in Vituixcad. This does not have to cost a lot of money but it is a good chunk of what a speaker would cost to build. If it is a hobby you want to get into then the cost may not be a problem and can eventually be amortized.

If you want to build a good speaker with minimal design effort and cost, then build one of the proven designs. If you want to build your own speaker, do that and use the proven designs as inspration on how to do it.

 

[dualazmak]

Let me share one suggestion of different approach.

To start with your (dream-)SP exploration, how about to convert existing commercially available (used or new) HiFi 2-way or 3-way passive (LCR-network plus attenuators) SP system into fully active configuration? And how about adding super-tweeters and L&R subwoofers? Through this approach and exploration, you can learn a lot, I believe.

Only if you would be interested in this approach, my multichannel project would be of your reference, and you can find here the latest system setup of my audio system.

 

[Markie Warkie]

Let me share one suggestion of different approach.

To start with your (dream-)SP exploration, howe about to convert existing commercially available (used or new) HiFi 2-way or 3-way passive (LCR-network plus attenuators) SP system into fully active configuration? And how about adding super-tweeters and L&R subwoofers? Through this approach and exploration, you can learn a lot, I believe.

Only if you would be interested in this approach, my multichannel project would be of your reference, and you can find here the latest system setup of my audio system.

Oh that is one cool system drawing ! i think my best option will be to follow a kit ! im hoping Directiva R2 has some updates

 

[Fredygump]

I would only build a loudspeaker myself if I wanted to realise a certain design with it that does not exist. If I have to ask for advice for that, it only shows that I don't have any good ideas myself. Besides, you can't explain everything that needs to be considered in the short form of postings in a forum. I had to read thick books first.

In my opinion, it makes no sense to imitate commercial designs; it won't be cheaper or better.

I wanted to agree that DIY makes the most sense when you want something that either doesn't exist, or is extremely niche and is therefore quite expensive. In those cases DIY can be a great value. I believe it is almost always impossible to DIY an inexpensive product for less than the mass produced product costs.

But I disagree with the implication that asking questions means you don't have a good idea. Good ideas often start out vague, and need to be explored to discover their final form. And asking questions is a valid way to tease out what the core idea is.

 

[Fredygump]

honestly, I'm a bit confused rn I do wanna build something of my own but at the same time that invites a lot of uncertainty the correct way would be to have the right equipment and so on

if I do go on I'm not gonna use an MTM design cause that would not fit my use case HT + Hifi and I would use a digital crossover
Now the problem is I don't know how to proceed honestly I'm completely lost

In my opinion the next step is to buy the drivers you want to use and to build a box out of cheap materials to test. Probably buy just what you need for 1 speaker at first, because once you see the result you might quickly realize that you want to change things. Maybe the drivers don't perform as desired? Maybe they don't look right? I think it is best to tell yourself you are building a prototype, so build one, test it, and then decide where you want to go with the design.

Don't be intimidated. At the end of the day you are just building a box. The people who designed the drivers did the hard part, and now the challenge is to find components that work well together. Think of it like an expensive Lego kit. It is as simple or as complicated as you want to make it.

But a warning--these things always cost more than expected!

 

[MAB]

im not really trying to imitate a commercial design i wanna make something of my "own" but since im limited on equipment and budget base it on something already established so im not just 100% guessing

Fluid gives some good advice. I built a small active desktop 2-way + dual subs out of drivers from a previous DIY project and a MiniDSP Flex EIght... using same method Fluid mentions above: a soundcard + mic, REW, a turntable, and Vituixcad. It sounds good. I am sure there are better desktop speakers on the market that cost less. But it's very rewarding. And there are expensive sounding speakers that sound horrible, so I avoided that.

But as Fluid mentions, hard to beat some of the top commercial designs. Some do have drivers available: I just DIY'ed a pair of JBL M2 based on some contributions from other ASR members and some external DIY sites. Those sound about as good as anything I have ever heard, possibly my ego getting in the way though.

None of this hobby should be considered cost-effective. But there are reasonably good DIY and commercial speakers that you can copy. And with a DSP crossover, you can realistically measure and make improvements without rebuilding crossovers, or even advanced knowledge of filter circuits and compromises.

 

[Fredygump]

Fluid gives some good advice. I built a small active desktop 2-way + dual subs out of drivers from a previous DIY project and a MiniDSP Flex EIght... using same method Fluid mentions above: a soundcard + mic, REW, a turntable, and Vituixcad. It sounds good. I am sure there are better desktop speakers on the market that cost less. But it's very rewarding. And there are expensive sounding speakers that sound horrible, so I avoided that.

But as Fluid mentions, hard to beat some of the top commercial designs. Some do have drivers available: I just DIY'ed a pair of JBL M2 based on some contributions from other ASR members and some external DIY sites. Those sound about as good as anything I have ever heard, possibly my ego getting in the way though.

None of this hobby should be considered cost-effective. But there are reasonably good DIY and commercial speakers that you can copy. And with a DSP crossover, you can realistically measure and make improvements without rebuilding crossovers, or even advanced knowledge of filter circuits and compromises.

Not to play devil's advocate, but..... I build speakers that were copying Genelec's W371A cabinets, added coaxial drivers, and ended up with a "mini" 8381A. I didn't pay even close to $70k, so it was cost effective.

My primary goal was great multi-sub bass, and my total build cost me less than a set of 4 decent quality subwoofers.

Speaking of that, building subs is a good place to start for DIY. To the OP, maybe buy decent MTM bookshelves, and then build some subwoofers? Then learn about multiple subwoofer optimization.

 

[computer-audiophile]

... I build speakers that were copying Genelec's W371A cabinets, added coaxial drivers, and ended up with a "mini" 8381A.

Interesting - you seem to have taken a highly developed commercial design as a model. Can you post a picture of how it turned out? That would be nice!

 

[Thomas_A]

Well. It can be done , this compete with commercial speakers that is, BUT it may require a lot of time and study (depends on the level of knowledge at the start). Then a lot of tests and measurements. Maarten did it when he designed his, which he called "Räv-L" (Fox-L) DIY. I don't know how many test boxes/baffles and measurements he did, many anyway. Fortunately, he then posted the entire construction, with detailed drawings for the lazy to copy. Which then many did.
Even if you don't know Swedish, you can check all the graphs in this thread. Where it compares to commercial speakers:

Can cheap DIY speakers compete with the best speakers on the market in terms of sound quality?
How difficult is HiFi and what can you do yourself? What should be considered to achieve good reproduction?

The short answer to the question in the title is an unequivocal 'YES', DIY is highly competitive! It took me as a beginner 13 weekends over the course of a year (it would have gone much faster if the measurement results below had been less surprising, and if the simulation tools had been more accurate) and material costs approx. SEK 5,000 ( €£$ 500) to achieve the result in this post in the form of a speaker like this ( called Räv-L):
View attachment 318233

Comparison reference axis Räv-L vs Revel M106,Revel 226be, Magico s5, Kef 1 Meta, Ino i14s
View attachment 318235

faktiskt.io • Visa tråd - Är DIY konkurrenskraftigt? SPIN-off Räv-L EPS2 piP i14s mfl

Actually Maartens Räv-L speaker name was based on of how you pronounce ”Revel”. I think.

 

[computer-audiophile]

... At the end of the day you are just building a box.

That would never have been my goal with DIY. But it is interesting to discuss such different approaches. For example, no one prevents you from modifying the drivers yourself or designing really unusual enclosures that are not available in this or that form.

 

[DanielT]

Whether DIY can compete with commercial speakers depends on what you compare it to.

The Tekton M-Lore costs $750
That speaker also looks like a real beginner DIY. That speaker consists of bass driver:

ALPHA-8A 8

Recommended for professional audio mid-range applications in a sealed cabinet, or as a mid-bass in a vented satellite enclosure.

eminence.com

And

Tweeter:

Peerless BC25TG15-04 | HiFiCompass

hificompass.com

The Peerless BC25 is a really good budget tweeter IF used correctly, which it is NOT in the Tekton M-Lore. A hopelessly misconceived speaker. Drivers that don't match, wrong crossover point and so on. Any beginner can make a better speaker than that, for that money. For less money than that. It doesn't take a lot of DIY tips and advice to create a better, lower cost, floorstanding speaker than the Tekton M-Lore.

Tekton M-Lore:

Tekton M-Lore Speaker Review

This is a review, listening tests, equalization and detailed measurements of the Tekton (Mini) M-Lore speaker. It is on kind loan from a member and costs $US 750 a pair. As you see, this is a mini-tower speaker if there is such a thing. I had to put a 6 inch stand under it to get the tweeter...

audiosciencereview.com

Can a DIY beat a $10,000 bookshelf speaker? Yes, and at a much, MUCH, lower cost. I am thinking of these:

Wilson Audio TuneTot Review (high-end bookshelf speaker)

This is a review and detailed measurements of the Wilson Audio TuneTot stand-mount/bookshelf speaker. It is on kind loan from a member and costs around US $10,000 (varies due to color). I am not a fan of the way the front looks without the grill. The sides though sport probably the best...

audiosciencereview.com

A well built DIY with quality drivers from Purifi plus help from knowledgeable DIY plus digital crossover, tested and measured and then tuned (in a test/evaluation process)plus if needed PEQ would blow those Wilson Audio TuneTot off the track. That for maybe around a tenth of the cost of a pair of Wilson Audio TuneTot (maybe a little more but about that cost). In and of itself, it would probably take some time to test and measure, but it would work.
Or you just take the blueprints from the Räv-L" (Fox-L) DIY that I described in #39 and build them and you have a pair of speakers that are easily better than the Wilson Audio TuneTot.

 

[Fredygump]

Interesting - you seem to have taken a highly developed commercial design as a model. Can you post a picture of how it turned out? That would be nice!

I made a build thread over here: https://www.avsforum.com/threads/di...axial-multi-subwoofers.3281891/#post-62760334

I saw that the cabinets were simple, but driver configuration was unconventional on the W371A. They seemed to claim multiple sub room optimization, but with subwoofers clustered too close together to work like that. So I started experimenting to see if I could figure out what they are doing. I came up with something that works, although I can't say it is how the Genelec system is working.

 

[computer-audiophile]

Thanks for the visuals!

I came up with something that works, although I can't say it is how the Genelec system is working.

Ok, so you made your own, different subwoofer.

 

[Fredygump]

Thanks for the visuals!


Ok, so you made your own, different subwoofer.

I believe that is inaccurate.

My design is a full range 4way/ 4.5 way floor standing speaker with active crossovers. It is true that I placed a special emphasis on fixing the low frequency issues that most designs ignore, but the fact is that only 1 of the 4 drivers plays in the "sub" frequency range. The design eliminates the need for separate subwoofers, but they are not subwoofers.

My reason for sharing is to say that a person should focus on the aspect of speaker design that they are interested in.