Another Baffle-less design
Not my first attempt at Open Baffle, but probably my last.
After my first experience with Bastanis and countless dissatifactions with mainstream boxed speakers, I knew OB was the way to go.
The speakers here represent the culmination of years of real world experimentation with the format.
Many designs were tried from information gleaned on this & other forums. I can express my opinion & say from my many baffle experiments, the smaller ....the better. And no baffle is best of all for lack of resonances, edge diffraction artifacts, imaging and simplicity!
I am no expert in electronics, so designing a crossover was out. I also wanted to lose the tweeter and see if it was really necessary.....so my efforts always concentrated on active WAW configurations....Alpair, PRV & lately Faital....all contenders for the wideband. I am aware that not having a baffle on the widebands would cause their bass output to roll off sooner.....but who cares? What I lost in bass I would gain by imaging, plus the crossover is easily controlled by my Behringer iNuke with built in DSP to compensate for the wideband loss.
Open baffle is not the most efficient way to produce bass we all know. Following advice to increase efficiency, different shaped baffles & frames were constructed. They do increase bass....but also increased modes and nulls in my room. Some of my U-frame configurations totally cancelled the bass in my sitting position, and my sofa had to be moved 2’ forward to give bass. Smaller baffles caused me less trouble with modes, but I could not achieve the SPL’s I was after.
By now my ‘baffles’ had turned into wooden ladder frames....simple and cheap to build! I was, happy, however, with my widebands and did not miss a tweeter....perhaps because I cant hear past 14k! At this point I was using the PRV 5MR450 Neos.
Again, following advice forum advice, I had always been using efficient fabric surround bass drivers, AE Dipole 18, Goldwood 8518, Selenium 15” Neos (pretty good, actually....now discontinued) and similar. They never gave me deep bass at volume.....always bottoming out at realistic levels.
Departing from the usual recommendations, I decided to try heavy coned inefficient woofers, with rubber surrounds, large Xmax and low Fs. High Qts was not a priority as I was not sold on the woolly bass it gives (witness AE & Goldwood experience). Class D iNuke watts are also cheap.
The initial firing up of the 15” Dayton IB385 shocked me as the floor shook with Kraftwerk! This was the first time I ever experienced bass like this. I was relieved to know I had made a wise choice.
From there I moved onto the Dayton 390’s 15”, and now UM18. Most recent wideband is the Faital array using a mix of 3FE22’s & 4FE32’s. The other big break-though was getting rid of the wooden ladder frames and switching to a metal scaffold to support the drivers. This also allowed me to develop a cable suspension to fly the woofers & array giving total isolation from each other and the floor.
I mainly updated to the UM18’s to achieve higher SPL’s in the bass. This goal was accomplished, but I didn’t realise how much more controlled the bass would be. The effect is uncanny, as I now have HUGE bass scale.....but it does not encroach on the wideband territory whatsoever. Everything is more clean, tight and controlled all through the frequency range.
Before you ask....I am not getting any comb filter effects, and my bass drivers do not move with low frequencies until I hit 5hz. They are wired out of phase, but acoustically in phase so the cones move in unison. The speakers easily reach 105db SPL’s, but I don't normally listen at this level as it causes my ears to ring.
I have tried to produce a couple of YouTube videos for those of you that might be interested. Will post once ready....
Not my first attempt at Open Baffle, but probably my last.
After my first experience with Bastanis and countless dissatifactions with mainstream boxed speakers, I knew OB was the way to go.
The speakers here represent the culmination of years of real world experimentation with the format.
Many designs were tried from information gleaned on this & other forums. I can express my opinion & say from my many baffle experiments, the smaller ....the better. And no baffle is best of all for lack of resonances, edge diffraction artifacts, imaging and simplicity!
I am no expert in electronics, so designing a crossover was out. I also wanted to lose the tweeter and see if it was really necessary.....so my efforts always concentrated on active WAW configurations....Alpair, PRV & lately Faital....all contenders for the wideband. I am aware that not having a baffle on the widebands would cause their bass output to roll off sooner.....but who cares? What I lost in bass I would gain by imaging, plus the crossover is easily controlled by my Behringer iNuke with built in DSP to compensate for the wideband loss.
Open baffle is not the most efficient way to produce bass we all know. Following advice to increase efficiency, different shaped baffles & frames were constructed. They do increase bass....but also increased modes and nulls in my room. Some of my U-frame configurations totally cancelled the bass in my sitting position, and my sofa had to be moved 2’ forward to give bass. Smaller baffles caused me less trouble with modes, but I could not achieve the SPL’s I was after.
By now my ‘baffles’ had turned into wooden ladder frames....simple and cheap to build! I was, happy, however, with my widebands and did not miss a tweeter....perhaps because I cant hear past 14k! At this point I was using the PRV 5MR450 Neos.
Again, following advice forum advice, I had always been using efficient fabric surround bass drivers, AE Dipole 18, Goldwood 8518, Selenium 15” Neos (pretty good, actually....now discontinued) and similar. They never gave me deep bass at volume.....always bottoming out at realistic levels.
Departing from the usual recommendations, I decided to try heavy coned inefficient woofers, with rubber surrounds, large Xmax and low Fs. High Qts was not a priority as I was not sold on the woolly bass it gives (witness AE & Goldwood experience). Class D iNuke watts are also cheap.
The initial firing up of the 15” Dayton IB385 shocked me as the floor shook with Kraftwerk! This was the first time I ever experienced bass like this. I was relieved to know I had made a wise choice.
From there I moved onto the Dayton 390’s 15”, and now UM18. Most recent wideband is the Faital array using a mix of 3FE22’s & 4FE32’s. The other big break-though was getting rid of the wooden ladder frames and switching to a metal scaffold to support the drivers. This also allowed me to develop a cable suspension to fly the woofers & array giving total isolation from each other and the floor.
I mainly updated to the UM18’s to achieve higher SPL’s in the bass. This goal was accomplished, but I didn’t realise how much more controlled the bass would be. The effect is uncanny, as I now have HUGE bass scale.....but it does not encroach on the wideband territory whatsoever. Everything is more clean, tight and controlled all through the frequency range.
Before you ask....I am not getting any comb filter effects, and my bass drivers do not move with low frequencies until I hit 5hz. They are wired out of phase, but acoustically in phase so the cones move in unison. The speakers easily reach 105db SPL’s, but I don't normally listen at this level as it causes my ears to ring.
I have tried to produce a couple of YouTube videos for those of you that might be interested. Will post once ready....
Attachments
Last edited:
Glad you are closer to the end of your journey... never say never, though! 
I'd like to know the reasoning behind putting the 4FE32s in the middle, and the 3FE22s around it.
And if by any chance you have measurements.
I'm guessing you are crossing the big 18" around 500hz-ish?
Using the iNukes, it's only a 2-way, right? So, there is nothing between the 3FE and 4FE?
Sorry for so many questions!
Thanks for sharing.
I'd like to know the reasoning behind putting the 4FE32s in the middle, and the 3FE22s around it.
And if by any chance you have measurements.
I'm guessing you are crossing the big 18" around 500hz-ish?
Using the iNukes, it's only a 2-way, right? So, there is nothing between the 3FE and 4FE?
Sorry for so many questions!
Thanks for sharing.
Glad you are closer to the end of your journey... never say never, though!
I'd like to know the reasoning behind putting the 4FE32s in the middle, and the 3FE22s around it.
And if by any chance you have measurements.
I'm guessing you are crossing the big 18" around 500hz-ish?
Using the iNukes, it's only a 2-way, right? So, there is nothing between the 3FE and 4FE?
Sorry for so many questions!
Thanks for sharing.
Hi Perceval.....you know I am not into maths or measuring. With speakers I just use my ears. Out of curiosity, I did take a measurement though:
So, under recent lock-down I bought 8 x 3FE22 + 8 x 4FE32's.....readily available and cheap from BlueAran in UK. I played with 2 x 4 units / side arrays of each driver. The 3's couldnt handle the SPL's, but the 4's could. However, the 4's lacked the highs of the 3's. So I did various combinations. The 334433 combo popped image-wise and sounded just right with the highs and body of the mids.
Crossing @ 275hz. Again, I did this by ear in real time using the Behringer Connect software from the comfort of my listening position. I am quite used to this procedure and was able to dial in quite quickly.
I am wondering how it’s done withe the kyron audio gaia speaker. I thought making them play in phase in this setup would result in more output in front and back...
Since one driver is inverted if you run both in phase they will just cancel each other =)
I tested such a configuration a while ago.
hi , am i right in thinking you had prv drivers for mids and highs prior to this set up? if so do you prefer the faitals? what are the pros and cons of both from your experience? thanks
k9vap :
do you have measurements form different angles at about 2m distance?
i would think the dispersion around the crossover might be weird because your low frequencies don't seem to have a standard dipole radiation.
do you use a steep filter?
there is quit a dip between 63hz en 1k
also high frequencies above 1,5khz fall off
you might also say you have 600-4000hz boosted
is it on purpose to have a far from flat curve?
it would be quit interesting to use both dsp and measurements to improve your design - i myself use equalization to make it sound nice on top of my flat filtered speaker (measured on a 30 degrees angle)
do you have measurements form different angles at about 2m distance?
i would think the dispersion around the crossover might be weird because your low frequencies don't seem to have a standard dipole radiation.
do you use a steep filter?
there is quit a dip between 63hz en 1k
also high frequencies above 1,5khz fall off
you might also say you have 600-4000hz boosted
is it on purpose to have a far from flat curve?
it would be quit interesting to use both dsp and measurements to improve your design - i myself use equalization to make it sound nice on top of my flat filtered speaker (measured on a 30 degrees angle)
Last edited:
I don't see a dip between 63Hz and 1kHz, but a wide bell at 1kHz that would not make the sound pleasing to my ears... I'm very sensitive to that 1kHz region, sounding way too "nasal" for me.
But if the 1kHz bell could be EQ out, it would give a nice sloping FR that I would like.
Of course, to each his/her own. if k9vap adjusted the speakers by ears, and this is the preferred FR, then, it's his/her right!
I don't see a dip between 63Hz and 1kHz, but a wide bell at 1kHz that would not make the sound pleasing to my ears... I'm very sensitive to that 1kHz region, sounding way too "nasal" for me.
But if the 1kHz bell could be EQ out, it would give a nice sloping FR that I would like.
Of course, to each his/her own. if k9vap adjusted the speakers by ears, and this is the preferred FR, then, it's his/her right!
Indeed, we all have our audio kryptonite.
I nearly always build in a dip centred on 4000 hertz, because that frequency annoys me.
Architect:
I have had speakers in the past that measure flat. They sound boring and bloated to me.
I like the 64-1000hz dip. If I make it flat, vocals sound too chesty. The slope is also adjusted to integrate with the arrays.
Adjusting the slope in real time from the listening position can cause a number of effects. The secret here is to find the sweet spot that suits most recordings, has enough 'chestiness' to fill in the lower mids, but does not muddy the sound or obscure detail.
The upper frequencies I guess are a characteristic of the drive units themselves. My recently acquired DEQx can no doubt eq the rest, but I am not yet proficient in its use. I actually have it operating flat. I only have only engaged a HPF @200hz to protect the array and lighten the load on the amp.
I have played with the eq page in real time with spot frequencies along the slope but left it flat. Nothing needs changing to my ears.
I don't see a dip between 63Hz and 1kHz, but a wide bell at 1kHz that would not make the sound pleasing to my ears... I'm very sensitive to that 1kHz region, sounding way too "nasal" for me.
But if the 1kHz bell could be EQ out, it would give a nice sloping FR that I would like.
Of course, to each his/her own. if k9vap adjusted the speakers by ears, and this is the preferred FR, then, it's his/her right!
I know exactly what you mean. 1kHz is an annoying frequency that grates on me too. I would attack this with my DEQx if it were causing problems, but so far, I cant hear it.
Hi Gazza
You know I lived with the PRV's for a long time. I correspond with many enthusiasts curious about building a set of speakers in this way, but the PRV 5MR450's are not available in the EU, and importing makes them prohibitively expensive.
I had been looking for alternatives for a long time. A couple I tried were the Dayton RSS100 and the famous Tymphani TC9. I found although they both sounded sweet, they lacked dynamics. I attributed this to their relative inefficiency vs the PRV.
During this Covid lockdown period I looked again for an alternative. I chose the Faitals because they had nice curves on paper....but also had a higher efficiency than most.
I knew I was going to need an array because their power handling was not great....but the price was cheap. I went with the 16ohm 3FE25's and the 8ohm 4FE32's. These had the smoothest responses on paper. I went with 8 of each, the intention being to make 2 arrays to compare to my PRV's
Starting with the 3" array, I immediately noticed an increase in the highs. There was a reduction in efficiency as the bass was overblown, so I had to adjust my gains. I lived with these for a few days in complete satisfaction. They were dynamic like the PRV's, were smoother and had better highs. My only issue was overheating as sustained SPL's where I could start to smell the varnish on the coils.
I then tested the 4in array. These had the same characteristics of the 3's.....maybe a bit more body. They could also handle the SPL's. I lived with these for a few days....but I felt I was missing something the 3's were giving me. I guess you could call it 'sparkle' or 'air'.
I then figured why not combine the array? I settled on 4 x 3" + 2 x 4" in line as a 333344 with the 4's at the bottom.
This worked and gave me everything. Air from the 3's, body from the 4's, smoothness, dynamics and SPL's that handled my Kraftwerk.
After living with this combo and going through my music collection for days I couldnt believe how good these sounded. I decided to put my PRV's back.
Now.....the brain is an amazing thing. I had loved these PRVs for 4 years, and now I could hear holes in the frequency range, harshness in the upper mids and a lack of 'air'. They still had great dynamics though, and a single driver unit would handle the SPL of 6 x Faitals!
I restored my 6 x array happy with the successful outcome....but there was more. I wondered if there was anything to be gained by re-arranging the lineup of the array. I tried a 433334 configuration, but it was like the 4's were not contributing. Then I tried 334433..........
OMG.....this was stunning. Everything suddenly came to life and the imaging popped. Now it had been sounding great before, but this was another level.
This configuration became addictive and I had to force myself to go to bed.....only to wake up in the morning & start again!
Not sure where to head now....this speaker does it all for me!
thanks for the comprehensive reply. sorry i wasn't aware you had discussed the updates on your blog (and the forum links )too. so ill check in on those now and again too
Trans-Fi Audio - My System
Trans-Fi Audio - My System
Last edited:
Oh, this has to be you! YouTube
I asked some questions on Youtube but maybe here is better.
Just curious why you are using 2 x UM18 per side in that config? Just to reduce 2nd order HD? And why did you stop using the AEs?
Hi James
The design of this speaker evolved over a decade. I went through many designs.
The AE's did not give me the SPL's I wanted. There is further explanation at the foot of the video.
Eventually, I switched to inefficient, heavy coned, large Xmax drive units, and I fared much better with them. I worker through the 15" Dayton IB395, RSS390, and finally UM18.
The cone orientation also evolved from conventional facing forward baffle arrangement, to one behind the other, Legacy style, to face to face naked.
This seemed to integrate best in my room. I didnt know why, but turns out other people had done more scientific experiments found this arrangement increased directivity.
Dont forget, Celestion had the SL6000 system in the 80's, although it wasnt that great.
The design of this speaker evolved over a decade. I went through many designs.
The AE's did not give me the SPL's I wanted. There is further explanation at the foot of the video.
Eventually, I switched to inefficient, heavy coned, large Xmax drive units, and I fared much better with them. I worker through the 15" Dayton IB395, RSS390, and finally UM18.
The cone orientation also evolved from conventional facing forward baffle arrangement, to one behind the other, Legacy style, to face to face naked.
This seemed to integrate best in my room. I didnt know why, but turns out other people had done more scientific experiments found this arrangement increased directivity.
Dont forget, Celestion had the SL6000 system in the 80's, although it wasnt that great.
k9,
I have built 9-driver and a 16-driver arrays. Both were sealed.
Due to the nature of LA, they only performed well if your head was within the line of drivers. If I stood up and my head was above the 9 drivers line, then the highs would just drop.
So, walking around the room was not a pleasant experience because the tonality would drastically change.
Since my speakers also perform the ambiance music during the day and while eating, I have moved on from those sealed line arrays to various OB iterations, my latest being a WFW, with the FR driver mounted on a diy horn. I quite like them.
But, the arrays have always been been in the back of my mind because when sitting in front of them, it was effortless music... and punchy at the same time.
It got me thinking I could do OB line arrays.
But, I also remember the tonality change when walking around the room.
Having a line array with no baffle, does it alleviate that problem?
If you stand up with your head above the line, do you hear a tonality change, or the OB are diffusing the sound enough that you don't notice it?
Thanks!
I have built 9-driver and a 16-driver arrays. Both were sealed.
Due to the nature of LA, they only performed well if your head was within the line of drivers. If I stood up and my head was above the 9 drivers line, then the highs would just drop.
So, walking around the room was not a pleasant experience because the tonality would drastically change.
Since my speakers also perform the ambiance music during the day and while eating, I have moved on from those sealed line arrays to various OB iterations, my latest being a WFW, with the FR driver mounted on a diy horn. I quite like them.
But, the arrays have always been been in the back of my mind because when sitting in front of them, it was effortless music... and punchy at the same time.
It got me thinking I could do OB line arrays.
But, I also remember the tonality change when walking around the room.
Having a line array with no baffle, does it alleviate that problem?
If you stand up with your head above the line, do you hear a tonality change, or the OB are diffusing the sound enough that you don't notice it?
Thanks!