“Building the best 3-way full range speaker in the world”

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

"The filter is a normal 3 way second order LR design."

Properly designed, the MTM driver configuration with 6"-7" midranges generates a large center lob which is often favorted for seated home theater listeners because there is reduced room/wall interaction. Clear sound for close seated listners in a small room, even with the speakers against the rear wall.

For broader MTM usage, changing midrange driver size, C-to-C spacing, and the crossover circuit can produce a few different polar patterns.
 

Attachments

  • MTM crossover effects.jpg
    MTM crossover effects.jpg
    191.8 KB · Views: 1,406
Last edited:
Beautiful work and a very nice finish. I'm glad you explained the title in the first post.
We often see the word "best" used on this forum and it always makes me wonder... Best what? For whom?

Most suitable to your own specific needs is what people should be after i.m.h.o. :). It seems you've got that covered!

The front speakers costed approximately €12.000. This includes the outsourcing of the CNC work.
If I would have just built the fronts it would have been more.

This price for the mains pair scared me a bit, that's quite some budget to play with!
How much of that budget went into the CNC process?
 
"The filter is a normal 3 way second order LR design."

Properly designed, the MTM driver configuration with 6"-7" midranges generates a large center lob which is often favorted for seated home theater listeners because there is reduced room/wall interaction. Clear sound for close seated listners in a small room, even with the speakers against the rear wall.

For broader MTM usage, changing midrange driver size, C-to-C spacing, and the crossover circuit can produce a few different polar patterns.

I read that explanation as well, but must admit that the prototype we built just started to sound fantastic after two days tweaking. And that was with a dsp based active crossover to figure out what crossover frequency worked best.
But the 2nd order LR was where we came back to every time
 
Beautiful work and a very nice finish. I'm glad you explained the title in the first post.
We often see the word "best" used on this forum and it always makes me wonder... Best what? For whom?

Most suitable to your own specific needs is what people should be after i.m.h.o. :). It seems you've got that covered!



This price for the mains pair scared me a bit, that's quite some budget to play with!
How much of that budget went into the CNC process?

Best for my personal taste and application indeed

The CNC part was quite a challenge.
After we glued the first 6 baffles (2 big ones and 4 small ones) and went through the CNC process and the next day the layers came loose. Something went wrong glueing the 13mm panels on a vacuum table. The second time I glued them under a 30ton press. That proved to work but the CNC machine caught fire due to the drill getting blunt.
I spent in total € 1000 for just the baffles CNC without material
 
Here are some notes about the prototype and how we decided to go ahead with this project

The gluing of multiple layers and the leatherette was done on a vacuum table. For most applications this is a great too but less ideal for the HDF where the sides came loose.

Getting the leatherette smooth around the corners was something I never did before, but this was quite doable with some patience and a heat-gun. The positive surprise was how well this fairly low cost material (€5 per speaker) matched the drivers with respect to color.

I added feet to test the influence of the distance to the floor with respect to the down firing port. This proved to be not critical as long as there was enough 'breathing space'.

The end result looked okay for a "prototype". This unit ended up being in use for 3 months.

The impedance test showed quite well that the tuning frequency was calculated/simulated correct.

I also tested how well a threaded rod would stick in a threaded hole directly in this HDF material. also the RAMPA's for the speaker mounting worked OK, but maybe next time I would also just make threaded holes in the baffle as that is very solid.

Tuning was first done with an Electro Voice DC-one. This gave a good idea of how low you can cross to the Scan-Speak Beryllium unit. For me the lower the better but not that low where at higher volumes the tweeter had to work too hard.

Then we made a prototype filter with basic quality components. added a baffle-step compensation to compensate for the reduction in LF energy below 300Hz. At the end of day two we had real music coming from this speaker.
This was the point where we were getting confident that we were on the right track. Another weekend of tweaking proved that we made a musical design and that things were so good that the real thing would be great.

The graph below shows the frequency response measured at 1 meter tweeter height.

Hope this helps you guys in future decisions ....
 

Attachments

  • 9 Building Pics proto graph.jpg
    9 Building Pics proto graph.jpg
    188 KB · Views: 391
  • 8 Building Pics proto filter.jpg
    8 Building Pics proto filter.jpg
    171.2 KB · Views: 402
  • 7 Building Pics proto tuning.jpg
    7 Building Pics proto tuning.jpg
    110.8 KB · Views: 380
  • 6 Building Pics proto test threaded hole in HDF.jpg
    6 Building Pics proto test threaded hole in HDF.jpg
    143.1 KB · Views: 316
  • 5 Building Pics proto impedance with DATS.jpg
    5 Building Pics proto impedance with DATS.jpg
    97 KB · Views: 623
  • 4 Building Pics proto end result.jpg
    4 Building Pics proto end result.jpg
    88.2 KB · Views: 616
  • 3 Building Pics proto gluing the front.jpg
    3 Building Pics proto gluing the front.jpg
    161.6 KB · Views: 1,142
  • 2 Building Pics proto  feet.jpg
    2 Building Pics proto feet.jpg
    146.7 KB · Views: 1,168
  • 1 Building Pics proto testing leatherette.jpg
    1 Building Pics proto testing leatherette.jpg
    101.9 KB · Views: 1,173
  • 0 Building Pics proto vacuum table.jpg
    0 Building Pics proto vacuum table.jpg
    96.1 KB · Views: 1,210
Last edited:
This price for the mains pair scared me a bit, that's quite some budget to play with!
How much of that budget went into the CNC process?

I need to clarify something Ronald,

As we do not have a wood-shop with the tools and the space to make 6 speakers more or less at the same time I made a deal with a local wood-shop.
(i saw in your line array thread that you have quite some space to build something)

For a reasonable amount of money he delivered the MDF material and did all the CNC work for all parts except the fronts as that needed a 5-axis CNC. At the same time we could use his place in some evenings and the weekend to put it all together, glue, sand and mount the parts.

Of course the parts are expensive.The drivers (see below picture when they all arrived) Audio technology, the beryllium Scan-Speak tweeter and the Mundorf top of line parts add most to that total amount.

Even the feet were €35 each on Ebay. Stainless steel rods, Technocell dampening. and €2100 for the Walnut side panels (4 speakers) also influence the cost significantly.

I must admit that when I started I had a smaller budget in mind, but if a project like this takes over a year you don't feel it that much.... haha.

Cheers.
 

Attachments

  • Building Pics all drivers.jpg
    Building Pics all drivers.jpg
    101.5 KB · Views: 545
I can imagine the cost turns up quickly. You went all out (as did I) and made something you'll probably keep for a long long time.

I never dared to look at my own final cost, I started buying stuff in September 2011 and started to work on it in 2013, taking me well over a year to build it all. Even the little stuff will raise the price quickly when you need a lot of it :). Include man-hours at it will become crazy money.

You ended up with something to be proud of. That's probably worth a lot more than just buying new stuff in a store at a similar price. It is to me anyway.
 
I can imagine the cost turns up quickly. You went all out (as did I) and made something you'll probably keep for a long long time.

I never dared to look at my own final cost, I started buying stuff in September 2011 and started to work on it in 2013, taking me well over a year to build it all. Even the little stuff will raise the price quickly when you need a lot of it :). Include man-hours at it will become crazy money.

You ended up with something to be proud of. That's probably worth a lot more than just buying new stuff in a store at a similar price. It is to me anyway.

Hi wesayso,
I guess you are right there, though I must say that building speakers can be frustrating. I have been building them since I was 15, so 40 years now. And often they show flaws after the rush of excitement has settled. And then the tweaking starts or you are less satisfied.
Then again, I had the same thing with bought speakers as none of them are perfect either.
The best speaker I ever owened was the Avalon Eidolon, I payed less for it than these costed and I sold them to a friend because I wanted to include them in a surround setting and there was no matching center and surround.
My friend still uses them and they do still sound great.
This speaker sound better to me, even though my front-end is most likely a few notches below what my friend uses. The mids are just so gorgeous and the new speaker goes much deeper, and the top end is more transparent.

The best single word to describe them is "musicality" and you know, that is probably the most important virtue of any speaker.

Cheers.
 
Testing a different Amp:

My 7 channel Audionet AMP VII (that I use in bi-amp mode on the main speakers) has a problem and is currently being repaired and updated.

So a great moment to test some other amps. The AMP VII has been my amp for over 8 years and therefor my reference for many years.
My friend has two Hypex Ncore 500 based amps and they were used for the big speakers. So no bi-amp but plenty of power with over 500W in 4 ohms.
A very nice amp, very clean, dynamic and with a lot of time for the tiny details. I have not a lot of experience with class D amps but these small, lightweight cool running mono blocks were very impressive.

But first my next update that these speakers deserve is a hi-res front-end with a new streamer and DAC (MQA enabled)
The speakers sound even better after a good burn-in time and I am still extremely satisfied with the set. It is still a great sounding speaker.
 
PSchut,

Good to know that you are enjoying the speakers big time. It is always a good sign when equipment pass the test of time. That way you know for sure that it's not only a "new toy" thing. :)

That is exactly my point. (the dissatisfaction after the new toy is not new anymore).
Every time I fire up the system and sit down to relax and enjoy music the set delivers. I am discovering more music every week and when it is good it is awesome.
 
Thats a very nice speaker you have built and took a great deal of planning and time. Like you I dont think I could just drop $5000- $15,000 usd on a turn key speaker and get near the satisfaction, intrigue, and exclusiveness unless I had a strong influence on its construction. I built a two way MTM using SEAS w18 nextel drivers and a 29 beryllium tweeter. I also noticed that anything other than BW or LR second order just was not going to work no matter the phase,crossover points, and or dealy using a DSP. I settled with second order LR at 2kh. Would you mind telling me if your low frequency drivers are wired in series/ parallel and in-phase or out of phase?
 
Beautiful looking speakers! I bet they have great dynamics too.

Out of interest- is the front baffle removable with the threaded rods you're using? or was that just for alignment / fixing or to try some sort of resonance decoupling from the rest of the cabinet?

If removable - do you have a photo of the back to show how the rods are terminated / fixed?
 
Thats a very nice speaker you have built and took a great deal of planning and time. Like you I dont think I could just drop $5000- $15,000 usd on a turn key speaker and get near the satisfaction, intrigue, and exclusiveness unless I had a strong influence on its construction. I built a two way MTM using SEAS w18 nextel drivers and a 29 beryllium tweeter. I also noticed that anything other than BW or LR second order just was not going to work no matter the phase,crossover points, and or dealy using a DSP. I settled with second order LR at 2kh. Would you mind telling me if your low frequency drivers are wired in series/ parallel and in-phase or out of phase?

The LF and the MF are in parallel. I don't think speakers should be mounted in series as they will ruin the dampening. the impedance of one is always in series with the other. I did order these units with a higher impedance of Re 7,5 Ohm.
Because of the flat baffle and an extra large capacitor in the HPF of the midrange the MF and HF are in phase. It gave a perfect response mounted in phase and a nice gap when in anti-pahse.
The MF/HF set is out of phase with LF.
 
Beautiful looking speakers! I bet they have great dynamics too.

Out of interest- is the front baffle removable with the threaded rods you're using? or was that just for alignment / fixing or to try some sort of resonance decoupling from the rest of the cabinet?

If removable - do you have a photo of the back to show how the rods are terminated / fixed?

The main function of the M8 rods is resonance control of the enclosure by putting tension on the rods. No alignment was needed, it actually proved quite a challenge to get all 10 rods aligned and at the end I used 10mm pipes to get the rods trough all chambers.
In theory they are removable. The threaded rods are screwed into the front directly for approximately 20mm. This is an extreme solid fix that I tested separately and could not pull out with any normal force.
I planned to glue the front to the main enclosure. And next time I would.
But I decided to use a "sticky" foam as I didn't know how this would go. Would the leatherette hold and so on.
Because 10 M8 rods are used and the front is massively rigid with 5 layers of 13mm HDF glued under 30T pressure. The whole structure is very very solid. I don't think glue would have made a difference.

See pictures for how it was done.
Hope this helps
 

Attachments

  • front foam.jpg
    front foam.jpg
    250 KB · Views: 816
  • threaded rod nut 2.jpg
    threaded rod nut 2.jpg
    126.2 KB · Views: 822
  • threaded rod nut.jpg
    threaded rod nut.jpg
    64.3 KB · Views: 805
  • black nut cover.jpg
    black nut cover.jpg
    64.3 KB · Views: 806
A major update in my system

Hi Forum,
Happy new year to you all.

After some 6 years living with a modified (self designed) pro DAC sourced by Sonos I decided that this is my weakest link and needed an update.
This DAC was based around the AES21 input receiver, PMD100 HDCD upsampling filter, 4 PCM1704 BB DACS in balanced configuration, so not an entry level design. But a Sonos streamer is the best either.


In the mean time, my 7 channel Audionet AMP VII got a major overhaul. All caps replaced, the output relays replaced as they gave issues. An update in the feedback circuit and the broken channel repaired. After return there was a clear upgrade in sonic performance especially resolution in the mid-range.

So back to the DAC; I needed two additional features. Hi-res 192/24 capability (the current one was only 48/24 capable) and MQA compatibility.

After some research there was a shortlist of DAC's and after auditioning I went for the Meridian Ultra DAC. It arrived just before Christmas. I installed Roon and it was up an running in No-time.

Interesting at first notes I wasn't overwhelmed, but after two hours I couldn't listen to my old DAC anymore. The increase in details but at the same time the relaxation of the sound, the silk and smooth highs and the much better controlled low. A wider and deeper soundstage. The balck background in hires music (I added Tidal for extra music).
A curtain was removed and at the same time the DAC disappeared.

The important part of this update is that the speakers are ready for better sources, and probably better amps. I feel no need to tweak the speaker. Al the DAC did was make the already great sound a few steps better.

Cheers,
Peter
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.