End Game DIY Loudspeakers

[Tangband]

A 4-way active dsp speaker with a 12 inch woofer, 6 1/2 midbass, 3 midrange and 0,75 inch tweeter will get very far If done right.

Crossovers at 80 Hz, 400 Hz, 3500 Hz with LR 4 crossovers.

This demands 8 amplifier channels and a dsp with 8 channels, optimal would be with digital inputs. This can be done with two minidsp HD, one for each channel. The soundsource and preamp can be a laptop with TIDAL or Apple lossless and a good DDC. ( Vintageflanker has done tests )

The baffle would be about 36 cm wide with rounded edges, the whole loudspeaker about 120 cm high.
Slotport or Linkwitz transformer ( closed ) for the 12 inch subwoofer.

If doing a 3-way, a 12 inch subwoofer and a 5 inch midrange and a 1 inch tweeter is thinkable.
Crossovers at 200 Hz and 2 kHz . This demands putting the 12 inch subwoofer close to the 5 inch mindrange, and away from the floor. A bit like ATC.

 

[Absolute]

Question for those of you with experience with DIY and rephase - and maybe an illustration of how difficult (or not) it is to better the performance of a good commercial speaker for those interested;

I've been wanting to try out a linear phase crossover on my DIY JBL M2 through the magic of rephase. How good does the measurements need to be? In-room measurements for a crossover between 700-800 hz is not exactly anechoic due to the distance required to properly sum the drivers in such a large speaker.
It's a bloody nuisance to drag 65 kg of speaker out of the house and up on a stand outside to measure, so any thoughts on how good is good enough for good results?
I'm using Minidsp flex 8 as a dsp.

 

[mcdn]

For a crossover in that upper midrange region around 800Hz there shouldn’t be any cabinet influences on phase. You’ve got DSP, configure the XO as you want it and measure in your room.

 

[Blockader]

Just reminding that I have listened 8361A and tens... maybe few hudred diy speakers. Also significantly better performers than 8361. I don't give a sht about sales and opinions in social media. Every listener should make own comparisons and conclusions if it is possible. It has been for me so no need to sympathize with the opinions of anyone.

>Also significantly better performers than 8361

would you mind giving some examples?

 

[kimmosto]

A 4-way active dsp speaker with a 12 inch woofer, 6 1/2 midbass, 3 midrange and 0,75 inch tweeter will get very far If done right.
Crossovers at 80 Hz, 400 Hz, 3500 Hz with LR 4 crossovers.

Agree. Years ago I designed 4-way prototype with 2x12" + 1x7" + 3.5" + 1" (Be) Scan-Speak drivers as total minimum phase with FIR XO. Total power 1 kW. Really nothing that couldn't be designed, measured in a living room and done as DIY in a garage. Acoustical design was very far from perfect with way too short roundings, but it is still one of the most nicest sounding conventional boxed speaker I know. But that kind of speakers for 40 k€/pair are impossible to sell nowadays on our local market so it was just a prototype for hifi show.

 

[Tangband]

Agree. Years ago I designed 4-way prototype with 2x12" + 1x7" + 3.5" + 1" (Be) Scan-Speak drivers as total minimum phase with FIR XO. Total power 1 kW. Really nothing that couldn't be designed, measured in a living room and done as DIY in a garage. Acoustical design was very far from perfect with way too short roundings, but it is still one of the most nicest sounding conventional boxed speaker I know. But that kind of speakers for 40 k€/pair are impossible to sell nowadays on our local market so it was just a prototype for hifi show.

Interesting.

Using dsp crossovers and building a 4-way is a good way to avoid compromises with beaming and directivity. One can also use cheaper drivers. Sometimes we see very expensive two way speakers when the real problem is the impossibility to have 20-20000 Hz reproduction at high spl with only two drivers, no matter how good they are.

 

[kimmosto]

would you mind giving some examples?

One of my DIY designs in 2010. 2x18" BMS woofers, 2" BMS coaxial in a horn, 1" BMS in rear horn. Cardioid 20-20 kHz. Nowadays with FourAudio PPA-1600 with Pascal M-Pro 1.5 kW/speaker. Acoustically flexible and tested in several rooms from 16m2 concrete flat up to at least 40m2 living room without need for room EQ. Not perfect and the best for every possible use case and individual for sure, but that is unfortunate fact for any other speaker - DIY or commercial.

 

[Absolute]

For a crossover in that upper midrange region around 800Hz there shouldn’t be any cabinet influences on phase. You’ve got DSP, configure the XO as you want it and measure in your room.

The problem is the woofer and its proximity to the floor, giving little reflection-free time and thus making the measurement for phase at and below the crossover rather untrustworthy. I'm asking because I don't know whether or not that makes an audible difference or what is preferable vs true anechoic crossover setup.
All these people using Acourate and Audiolense for crossovers brag about perfect crossovers, but is it really when measured in-room at the lp?

 

[abdo123]

A 4-way active dsp speaker with a 12 inch woofer, 6 1/2 midbass, 3 midrange and 0,75 inch tweeter will get very far If done right.

Crossovers at 80 Hz, 400 Hz, 3500 Hz with LR 4 crossovers.

That seems too close to what the KEF R3 is .

 

[voodooless]

One of my DIY designs in 2010. 2x18" BMS woofers, 2" BMS coaxial in a horn, 1" BMS in rear horn. Cardioid 20-20 kHz. Nowadays with FourAudio PPA-1600 with Pascal M-Pro 1.5 kW/speaker. Acoustically flexible and tested in several rooms from 16m2 concrete flat up to at least 40m2 living room without need for room EQ. Not perfect and the best for every possible use case and individual for sure, but that is unfortunate fact for any other speaker - DIY or commercial.
View attachment 254771

How is this thing remotely comparable to the 8361? Cool concept though

 

[abdo123]

How is this thing remotely comparable to the 8361? Cool concept though

Sssh... you're not supposed to question the validity of designs that need 5-6 meters to sum acoustically but somehow worked perfectly in a 16m2 room.

They're obviously vastly superior to the 8361, maybe, if you're listening in a warehouse.

 

[thewas]

One of my DIY designs in 2010. 2x18" BMS woofers, 2" BMS coaxial in a horn, 1" BMS in rear horn. Cardioid 20-20 kHz. Nowadays with FourAudio PPA-1600 with Pascal M-Pro 1.5 kW/speaker. Acoustically flexible and tested in several rooms from 16m2 concrete flat up to at least 40m2 living room without need for room EQ. Not perfect and the best for every possible use case and individual for sure, but that is unfortunate fact for any other speaker - DIY or commercial.
View attachment 254771

Is the photo showing the design that you describe as how can just two frontal woofers make it cardioid 20-20 kHz?

 

[kimmosto]

How is this thing remotely comparable to the 8361? Cool concept though

At this point I could finally ask why we ended up comparing DIY and commercial, and especially 8361 or Blade to DIY? Topic is "End Game DIY Loudspeakers", and in that context both 8361 and Blade are irrelevant or just some implementations IRL we can study and partly imitate - or NOT if they are not close enough to endgame.

Anyway, here is snapshot from one event where people compared Blade and very conventional 2.5-way, possible for almost any DIYer without much problem or challenge in acoustical XO design or manufacturing. Guess how many listener evaluated Blade better? Next questions would be why and how much result of comparison was related to compatibility of speaker concept and room+locations or personal preference. That part of showroom was build for Kef and McIntosh so intention was not to emphasize some local shack.

 

[mcdn]

The problem is the woofer and its proximity to the floor, giving little reflection-free time and thus making the measurement for phase at and below the crossover rather untrustworthy. I'm asking because I don't know whether or not that makes an audible difference or what is preferable vs true anechoic crossover setup.
All these people using Acourate and Audiolense for crossovers brag about perfect crossovers, but is it really when measured in-room at the lp?

What phase issues are you trying to solve? If you’re aiming to get a nice impulse response when the driver outputs are combined then it should be a simple matter of setting the right delay to compensate for their relative distances from the listener - but it will only be exact for one vertical angle at best. Only coax drivers can be phase accurate at multiple vertical angles.

 

[kimmosto]

Is the photo showing the design that you describe as how can just two frontal woofers make it cardioid 20-20 kHz?

Woofer box is resistance enclosure.

 

[kimmosto]

If some Genelec or Kef lovers doing nothing as DIY have come here to negate DIY I kindly ask them to f*ck off. This forum has already too much hype around those two brands. Open your ears or let DIYers think that they can cope this topic without your help.

 

[thewas]

Woofer box is resistance enclosure.

Do you have a full measurement spin of it as with such resistive materials it is not trivial to achieve it smoothly over a big frequency region, even Geithain who does it since decades isn't measuring perfectly there:

Source: https://www.me-geithain.de/de/rl-901k.html

 

[thewas]

If some Genelec or Kef lovers doing nothing as DIY have come here to negate DIY I kindly ask them to f*ck off. This forum has already too much hype around those two brands. Open your ears or let DIYers think that they can cope this topic without your help.

Because I really appreciate your great contribution to the DIY community with your superb VituixCAD software I don't think that with such argumentative style you are doing yourself or the DIY hobby a favour, maybe you should reconsider it.

 

[Absolute]

A cardioide speaker with distributed high-capacity woofers in height and compression driver combats a few key problem areas in any speaker, SBIR from front wall, floor, side-walls to a degree and dynamic capability - especially important in the bass to lower mids. Can't say anything about the horn, especially integration with a back-fired horn that probably aims to create more air up top, or the resonance and diffraction qualities of the cabinet, but the speakers pictured at least attacks some of the worst offenders to sound quality imo.

It could be a high-capacity Dutch 8c or a dud for all we know, but the ideas behind the design makes perfect sense to me.

 

[Vineeth Kumar V]

Is the photo showing the design that you describe as how can just two frontal woofers make it cardioid 20-20 kHz?

I see that Kimmosto has already answered this question while I was typing.
But if you are really interested in learning more about how the cardioid pattern was achieved, its impact on performance and some construction details from long back, please take a look here:

KS-1804 mk2

Some of us consider Kimmosto's old, now non-existent website as a gold mine of speaker design information and try to learn from it in whatever we can
I see that some people here in this thread disregard the opinions of someone as experienced in this field as Kimmosto and think that whatever brands they hold as idols/ideals are the best designs. Well, I can only pity them. Let ignorance be bliss.