Pretty Ugly but Sounds Amazing! - Designing A 4-Way Open Baffle Speaker

I settled on 150Hz, 520Hz and 3500Hz for now. This is all very, very preliminary and things will change.
For example, the new 8" is a Dayton Audio DC200-8 and I should have those early next week. They are 87db and the ones I'm using now are 92db, so a much better match for the other three drivers.
My intention from the start was to design and build a passive crossover for these, but digital has too many advantages these days in a system that's setup for the long term. It also has the flexibility to be tweaked are reconfigured without changing components. Or used for an entirely different speaker if I get tired of these.
One problem is that I don't have enough amplifier channels in my 6 channel amp for a pair of 4-ways. I do have enough when I include the two receivers I have. The Yamaha is a 7.1, so I can probably use 4 of its channels without much power supply sag. All off the drivers are 8 ohms, so that helps.
Ultimately I want to build an 8 or 10 channel amp with a minimum of 100 watts for each channel. And I mean from scratch - design the amp circuit, the pcb's and build it all from the ground up. I already have the transformer for such an amp, a monster toroid 1200VA 50-0-50. I also have a case I started making over 10 years ago that can easily fit 10 amplifier channels and a big power supply.
So in other words, there's some interesting stuff ahead if you are into that kind of thing.
You can help support the work I do in making these videos:
Project plans for sale: ibuildit.ca/pl...
Join the ibuildit community on Loacals: ibuildit.local...
Support this channel on Patreon:
www.patreon.co...
#diyspeakers
#johnheisz
#audio
My "Scrap bin" channel:
/ ibuilditscrapbin
My main channel:
/ jpheisz
Website: ibuildit.ca/
Facebook: / i-build-it-25804801424...
Instagram: / i_build_it.ca

[@IBuildIt]

I settled on 150Hz, 520Hz and 3500Hz for now. This is all very, very preliminary and things will change. For example, the new 8" is a Dayton Audio DC200-8 and I should have those early next week. They are 87db and the ones I'm using now are 92db, so a much better match for the other three drivers. My intention from the start was to design and build a passive crossover for these, but digital has too many advantages these days in a system that's setup for the long term. It also has the flexibility to be tweaked are reconfigured without changing components. Or used for an entirely different speaker if I get tired of these. One problem is that I don't have enough amplifier channels in my 6 channel amp for a pair of 4-ways. I do have enough when I include the two receivers I have. The Yamaha is a 7.1, so I can probably use 4 of its channels without much power supply sag. All off the drivers are 8 ohms, so that helps. Ultimately I want to build an 8 or 10 channel amp with a minimum of 100 watts for each channel. And I mean from scratch - design the amp circuit, the pcb's and build it all from the ground up. I already have the transformer for such an amp, a monster toroid 1200VA 50-0-50. I also have a case I started making over 10 years ago that can easily fit 10 amplifier channels and a big power supply. So in other words, there's some interesting stuff ahead if you are into that kind of thing.

[@dab9742]

It is better to put less quality loudspeakers, than to compensate for the weakness of bad loudspeakers by their quantity. And putting a 100 watt amp on a tweeter is like putting a big-block V8 on a skateboard...

[@stephenhall3515]

I am learning so much from your channel just in time for prepping my large music analogue archives for sale to universities in Japan or China by mid-2025. What I copy and retain for my sunset years will probably be played on speakers derived from your ideas. Thank you so very much.

[@Clobercow1]

Nice work John! Obviously there are many ways to do this and get good results. I would like to share what I do and I appreciate you sharing what you do! This is great for our community. When you do the tweeter, mid, and midwoofer, you might consider a near-field gated response and off axis testing so you can dial it in to your desired curve on as wide an axis as you can so your room reflections are more consistent with the direct speaker sound. This can help a lot with clarity, sound stage etc. This is what I do. I do a 1m measurement and then a near-field measurement and mock up the near-field measurements in software and compensate / adjust for their time distances so they match the 1m distance. Once I do that, I can design a crossover, digital or analog, and see what happens without having to do the actual test. Every time I've done this, the final result was bang on at 1m. It's a bit more effort up front, but it's overall quicker than doing sweeps over and over.

[@IBuildIt]

I measured all of the drivers to begin with, on and off axis, and posted those measurements here: kzbin.info/door/zGbp-rRVNwyFhn9gHoZr5gcommunity?lb=UgkxwYMzqsNSTQlBCmqJOvjnp8O4O_kVu162

[@GregsGarage]

It's interesting you bring up whether or not you should eq to a flat response... We argued with the guy I worked for all the time about hearing with your ears and not your eyes. He wanted flat response and we always put a few intensional bumps or dips in the eq to give the entire rig character and separation. When you're in a recording studio, you want flat Genelecs or something similar. They sound terrible for actual listening, but if you can make them sound good in the studio, your mix will sound great on "listening" systems. You are 100% correct, we like to hear things in bands.

[@IBuildIt]

Flat from the 1 meter at 1 watt standard measurement is still the target, but that's not the listening position. I know from flattening at the listening position you are losing some of the natural character of the music. Vocals are subdued and everything sounds less lively.

[@wolvenar]

It's nice though to have a known level value to start with so you can manipulate it elsewhere. But I understand what you mean.

[@adriantaylor1266]

You won't regret the purchase of the DSP. Many people try to iron out any humps in the frequency instead of iomproving via room, speaker or crossover optimization first. I also use a DSP since about 11 years. In the beginning I really squeezed out everything of the DSP board. Time after time I switched off different filters and effects. Then I started from the scratch again and set up my speakers with all given/built in features. After that I just made some slight touch ups to correct some modes I have in my listening room in mid range. And no, my wife didn't allow to put sound panels in our living room ;)

[@grantturley8600]

This series is so great! Your point about frequency response is interesting. As an audio engineer I definitely seek the flat response, but I think you’re probably right that most people are used to the sound of speakers plus the room in listening to music.

[@michaeltablet8577]

Very interesting! Thanks for posting the results. Glad you made this channel!

[@joeygsaudiochannel3972]

Very nice. My Wife hates the look of my OB's. But then my Carpentry skills always sucked. My OB's have an H-Frame stacked dual 12" GRS Woofers and whether I run them full range run the H-Frame's with my ASW-1000 ICE Amp's with miniDSP to tune the Bass. Evryone has always they sound great.

[@RobertJeffersonBased]

I literally just completed an H-Frame for two 12SW-4 drivers yesterday. I have no EQ so I know they are dropping off below about 40Hz but I don't care. They sound really, really good!

[@dab9742]

Passionate about high fidelity since my childhood, and I am now 60 years old, I can say that the more speakers you put in a speaker, the less realism it will be when listening. I have tried everything, low-slope filters, very steep filters, active filters, digital filters, DSPs, processors and others, I have never been so charmed, as by listening to it at its simplest. Like for example, a maximum three-way speaker, two would be better, but very difficult to obtain, to cover the bandwidth of our ears as much as possible. Afterwards, at the filtering level, the less stiff it is, and adapted to the natural lowering of the loudspeaker, the better it is, and finally, in amplification, the less corrections there are, the better it is. The grail, all with horns, with good performance, a small tube amp why not, or a class A of a few watts simply, and a good turntable, or failing that, CDs correctly recorded and mixed. The DSPs, processors, or other devices of this kind, only eat micro information, dynamics, warmth in the voices, and in the instruments. And pay attention to the curves revealed by your measuring devices, they only rarely reflect what our ears actually perceive. And prefer a speaker capable of 34 to 17Khz +-3dB, than another of 34 to 22Khz +-3dB the second will seem less balanced. And do not forget, that after the studies of Fletcher and Munson, our hearing capacities from the age of 40, greatly limit our faculties of perception of certain sounds, especially after 8Khz. premiere-scientifique.simulatus.info/TS/voix-oreille.pdf

[@IBuildIt]

You obviously haven't listened to a well designed multiway speaker if you think that. But then what we hear is largely influenced by what we assume to be true. If you expect a 4-way speaker to sound less real, then there's a good chance that's the way you'll hear it. If you think crossovers are evil, either passive or digital, that will bias you as well. Only by doing the listening blind can you take these biases away. Have you done any of that?

[@practicalguy973]

Very nice work! You might want to consider some pro 8" woofers and maybe compression drivers and waveguide for some future projects to have some extra head room efficiency wise teamed up with 15" woofers. Pro 8" would be less bass and xmax but usually 6db or so more efficient and for midrange could be effortless sound with the right crossover.

[@aman-mn5kc]

Those crossover bands are linked by default when searching, and mini dsp does not even offer that anywhere, it's very uesful! In my case i have to use old gear to find the perfect joint and then connect the mini dsp for the rest.

[@noahnovotny2100]

In addition to listening position measurements, you should take off-axis measurements. That will give you an idea of how the directivity and radiation patterns of the drivers match. If you are finding that a wavy response sounds better, I'd guess things aren't smooth in that regard. Radiation mismatches can cause wonky soundstage effects because certain objects will sound larger/smaller or wider/thinner if they play frequencies that overlap drivers. This is especially important to get right with a 4 way as the drivers will be sharing many instrument ranges. If you are planning on using an all-active crossover, directivity is the thing you should focus most if not all your attention on. Driver sensitivity and frequency response can be tweaked later to fit your space and taste with eq. Also, don't forget to time-align the drivers. You have obviously spent tons of time and money treating your room to have a very even reflection response by frequency. Those reflections will have more of an effect on the sound of the speakers than anything else. Super cool stuff! I always like to see what you are up to.

[@IBuildIt]

I measured all of the drivers to begin with, on and off axis, and posted those measurements here: kzbin.info/door/zGbp-rRVNwyFhn9gHoZr5gcommunity?lb=UgkxwYMzqsNSTQlBCmqJOvjnp8O4O_kVu162

[@noahnovotny2100]

@@IBuildIt oh nice I hadn’t seen that, but you should still do the off axis measurements with the crossovers applied. It’s not enough that the drivers measure smooth off axisby themselves, they have to measure smooth off axis together as well.

[@IBuildIt]

Off axis response of individual drivers won't change with the crossover. Total off axis should be even better after the crossover because these drivers are operating well below their beaming frequency (other than the tweeter). But, like I said in this video, I ain't done yet. Still a lot left to do and getting too granular with the measurements isn't productive at this point. When I settle on the final set of drivers and get the real baffle made, then I can run some more comprehensive tests.

[@gene8933]

I have 4 way system right now . I noticed different slopes get me flatter response. I agree that to flat makes it kind lifeless.I like peaks to get drum and cymbals to sound life like.

[@George10767]

You are obviously a true perfectionist. I wonder how "flat" is your test microphone? Also you would need to create an anechoic chamber, both for testing and for personal appreciation.

[@RobertJeffersonBased]

I finally built an H-Frame for a couple of the 12SW-4 GRS drivers based/inspired on your open baffle speaker videos. They sound dumb good! I'll release the design for the H-Frame too so people can build it if they want

[@vemsom]

I know you probably get a lot of suggestions from people who thinks they know a lot, but here's another 😂 You might get a bit less dipole cancellations if you use the lower of the cutouts for the bass. The frequency should be low enough not to cause any combfiler effects. And while I might have your attention, thanks for all your entertaining, inspiring and informative videos. The listening room looks awesome!

[@IBuildIt]

Not seeing much dipole cancelation, or at least not enough to cause a problem Surprised no one has suggested that I invert the polarity on the subs to try to solve that null at ~40Hz.

[@vemsom]

@@IBuildIt the fact that you don't have any space under the speakers for cancellations will also help a lot..phase inversion sounds like a creative way to tackle it. If it works, it works. If it sounds better it is better (often overlooked fact)

[@attainableaudio4533]

Awesome, looking forward to your use of the MiniDsp, I live like 15 minutes from madisound and may have to drive over and pick on up, hehe, hope you go in depth to the programming process of it. I've yet to make the jump into testing active setups.

[@Gersberms]

I'm super interested in the digital crossover. Ever since my active Dynaudios I've known that this is the way to go. Would you care to include a recording of your speakers at some point, maybe on a separate channel so if it gets flagged, no harm is done? I know it doesn't say a lot about the speakers because room acoustics play a big role and it's all through the lens of my headphones - but my headphones are pretty nice and I feel like I get at least some sense of the audio you're getting.

[@markduggan3451]

This is so confusing for me but I'm sure it makes sense to you.

[@steveducell2158]

You might want to try testing your rig outside in the back yard. This would eliminate the interactions with the room. Once you get a flat response from the speakers you can then work on room treatments.

[@rawsonte]

Agree on the peq smoothing didnt sound better in own projects as well, more crossover points and drivers used. Didnt catch which minidsp used, use a 8x10 one targeted at car audio which seemed to use the stronger dsp engines and has the active up +6db output (non hd minidsp and that 4x10 one seem to use older 56bit dsp), dunno it more mental and did for peace of mind to thrwart upgraditis.

[@wobblysauce]

Equalizing is a personal preference.

[@MichelLinschoten]

That's why every record we listen too is subjected to it at time of mixing

[@corystock1569]

Love this channel

[@wolvenar]

Nice. Waiting for all the comments telling you your wrong about ... Something.

[@IBuildIt]

I welcome the comments if the person doing the telling actually know what he's talking about.

[@wolvenar]

@@IBuildIt Yeah, those are nice. It's the people who complain but add nothing of benefit to the conversation that bug me.

[@johnviera3884]

That Behringer is awesome. Sucks that it’s not 4 way stereo

[@deanwilliams93]

But, John, what do you listen to?

[@Gioxtream]

May I know why you did not use a 4 way passive crossover?

[@yttean98]

Use digital xover then Room equalizer(if you like) built-into the xover, eg. Minidsp. then your system will sound good. I built a similar system they sound great and better still cheap.

[@Leondc123]

Please my friend. Do MORE videos Editing is no needed. You can just live recorder and then make the upload. Share your thoughts during the process of creation and take us with you. Do MORE video. With tweaking and mods, even in the part of woodwork and cutting wires, and screwing the drivers on the plate. I love DIY speaker, and in these exactly moment i am all about OB and two 15" speaker with a fullrange driver or a compression driver and a carved wood horn. That's why i am saying... do more. Reach some sponsors . And let us hi the like button more ofter, TNX.

[@vangelo2252]

Not sure whats going on with your sub woofer trace where its rising @ 20 hz! I would die for a measurement like that, esecially from a 15" woofer. I am running 4 x UM18's in OB clamshell arrangement. I think you are picking up floor resonances..........

[@IBuildIt]

The subwoofers are infinite baffle, four 12" in the front wall and that is the measured response. They easily go down well below 20Hz, something that's almost impossible to do with boxed subs.

[@felixm.8910]

I myself bought a miniDSP 2x4 HD (?), the one with XLR inputs and outputs. I wanted to use it as a crossover from my 3-Way Stereo speakers to my subwoofer. I couldn't do that tho, since my tweeters have such high efficiency (107 db/W/1m) that you could audibly hear the hissing coming from the D-A-converter stage, or somewhere else in the circuitry. Now I only use it as an active highpass-filter for my 18" sub (24dB/oct. @40Hz) or for testing purposes to figure out crossover points like you did. From what I could find in hifi- and the miniDSP-support forums is, that it is not possible to get rid of the hissing.

[@sudd3660]

il am in the same boat, hiss after going active crossover speaker. but there is no hiss in passive crossover so what passive parts can we add to attenuate the problem? otherwise active speakers are dead tech until they fix it.

[@felixm.8910]

@@sudd3660 the problem with attenuating the hiss is, that it also attenuates your midrange and tweeter drivers over all. So either use the active crossover/dsp to figure out how you want to design your passive x-over, or go a bit more expensive and get a higher quality soundprocessor. At moment I don't need another DSP, but if I'd need one, I would propably get a "dbx driverack PA2".

[@sudd3660]

@@felixm.8910 the dsp part is not the problem, i get hiss as soon as i plug in a amps to the tweeter and midrange. i dont mind adding extra part or loosing power or volume, i got 100watts for a tweeter and that is more than i need. subwoofers got separate amps.

[@IBuildIt]

All amplifiers (and all electronic audio gear for that matter) produces some amount of noise - it's the noise floor of the device. If you can only hear it when you are close to the speaker and without music playing, it isn't a problem worth fixing. Even music played at a fairly low level swamps any noise that comes from most competently made audio equipment.

[@felixm.8910]

@@sudd3660 in that case I would set the volume of the dsp to the maximum and set the power amplifier of your speaker as low as you can afford. That way you get the signal-to-noise ratio to the best possible result. That worked for me for prototyping/figuring stuff out. But in my small living room the hissing was unbearable for me, since I have very sensitive hearing and also didn't want to get distracted by it. In the end I built a passive crossover and hooked the speakers up to my trusty Denon class AB monoblocks. And what can I tell you? Not a single sound or hissing or noise comes out of the speaker when nothing gets played.

[@vicg5323]

Can you do a sound demo please?

[@jannevellamo]

I think a dipole tweeter would be better. If half the speaker is already dipole, why not the rest of it? Yeah, I've had a pair of ESS AMT 1bs, which of course were fully dipole and they did sound glorious, till somebody ruined them.

[@danhorton6182]

What DSP unit did you pick up? One from Mini DSP or did you grab Dayton’s unit? I have a couple of the Dayton units and for what they can do, especially a stereo pair of 4 ways, they are an absolute steal. I also love being able to make changes on the fly using the app on your phone.

[@IBuildIt]

The miniDSP 4x10.

[@chetleonard169]

Each of the 4 drivers seem to display similar "humps" in their response curve. Am I right in thinking those similar humps are due to room nodes?

[@IBuildIt]

The room is very well treated (as shown in previous videos) so this is mostly the natural output from the speaker at that distance. Believe it or not, it's actually really good in-room response. But there's room for improvement and I'm far from done.

[@ezjuliani]

Well executed

[@gizmobowen]

I've got a mini-dsp that I used to dial in my subwoofer and I can't recommend it enough.

[@Steve_Just_Steve]

Will you talk about your thoughts on amps? Or have you and I missed it? Thx Edit: My bad I found a few vids =)

[@IBuildIt]

I'm in the "all competently designed amps sound the same" camp. Guys fret over this or that amp, while listening to a system set up in a room without any acoustic treatment. Properly done acoustic treatment is a much, much bigger bang for the buck investment.

[@Steve_Just_Steve]

@@IBuildIt Almost exactly what I wanted your thought on. Thx

[@julioaperales1228]

Hi. I know You just opted for a miniDSP... I did my 3 way Open Baffles using analog crossover from Ashly. an XR2001 which is 4-2way active or 2 3way active. Now I will add subs and go 4 way with an Ashly XR4001 which is 4way/7way. I have a miniDSP DAC8 and is good and versatile but keeping it analog sounds better. No EQ, just tuned like you did

[@dilbyjones]

What dsp mini dsp model did you get? Thanks BTW.

[@IBuildIt]

The miniDSP 4x10 Hd

[@mraarone]

Super impressive on all fronts. I do have one suggestion to help you. You see, I’m a data scientist and saw your charts and sensor set up. I noticed you didn’t have an actual listener’s body present during your sensor data collection. Your presence will change all of your fine tuning. Your body will dampen and modify the reflectivity of sound represented in your data. I recommend rerunning a data collection with you actually sitting in the chair, if possible. This is relevant unless you are planning to have the perfect room for listening while not present.

[@rewind9536]

Minidsp 4x10HD is too old, but they have newer units that sound okay. The Behringer unit is just terrible.

[@wesselingaudiodesign5031]

The rise below 20hz looks like noise in the measurement!

[@IBuildIt]

If it was noise! then a high pass filter wouldn't get rid of it!

[@abdo-dr1tu]

Speaker designers are crying in a corner.

[@KravchenkoAudioPerth]

Nope John's ideas are plenty solid. Mark

[@abdo-dr1tu]

@@KravchenkoAudioPerth adjusting the crossover based on in-room measurements is a capital crime.

[@KravchenkoAudioPerth]

@@abdo-dr1tu Nope. It is smart. And should be possible if required in many designs. An active speaker is the pinacle of the loudspeaker art. Done correctly it is about as good as it gets.

[@abdo-dr1tu]

@@KravchenkoAudioPerth you should adjust the crossover based on the direct sound, not the mess that is the acoustics of a small room. Our ears are not microphones.

[@hiveofactivitysussex6924]

Which slopes did you use John? I used the Behringer for a few years and yes it's good IMV. I now use a DEQX which is a lot more expensive but a lot better in terms of features and sound quality.

[@IBuildIt]

This unit only has the one option, 24db L-R's I think.

[@hiveofactivitysussex6924]

@@IBuildIt Yes, I seem to recall that. It was a good few years ago. Many thanks