EVA foam for performance speaker enclosures

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here are some new measurements of the 3cm deep waveguide. With the close distance to the woofer, the slope on the bottom of the waveguide is steeper so it is not perfectly symmetric. This is causing some problems in the vertical off axis response. For the speaker to sound right, I have to keep the tweeter higher than my ears and have a very narrow "good" vertical height position.

When I align the microphone perfectly with the mid point between the woofer and tweeter, the measurements are fine.
495345d1437804602-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-bessel-3700hz.jpg


495346d1437804602-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-3500hz.jpg

495347d1437804602-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-4000hz.jpg

495344d1437804602-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-no-eq.jpg


But when I align the microphone directly in front of the tweeter it causes some problems
495348d1437804602-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-4000hz-centered-tweeter.jpg


But outside of some possible problems above 10khz in vertical off-axis and alignment, this is not bad.
I think I have convinced myself now to build a new front baffle with a 2cm shallow "waveguide".
 

Attachments

  • rs28f-3cm-waveguide-no-eq.jpg
    rs28f-3cm-waveguide-no-eq.jpg
    71.2 KB · Views: 954
  • rs28f-3cm-waveguide-bessel-3700hz.jpg
    rs28f-3cm-waveguide-bessel-3700hz.jpg
    74.4 KB · Views: 1,917
  • rs28f-3cm-waveguide-lr12-3500hz.jpg
    rs28f-3cm-waveguide-lr12-3500hz.jpg
    77.2 KB · Views: 964
  • rs28f-3cm-waveguide-lr12-4000hz.jpg
    rs28f-3cm-waveguide-lr12-4000hz.jpg
    80.8 KB · Views: 971
  • rs28f-3cm-waveguide-lr12-4000hz-centered on tweeter.jpg
    rs28f-3cm-waveguide-lr12-4000hz-centered on tweeter.jpg
    79.1 KB · Views: 959
Last edited:
OK here are some new measurements of the complete Satori MW16+RS28f EVA build
In this case there is no internal bracing, no filling, no additional lining. Just a raw EVA box, 20mm laminated panels, taking advantage of the natural damping and sound absorptions properties.

Very linear phase response, nice fast decay in the CSD waterfall plot with minimal resonances.

some of the specs of the DSP for these measurements:
rs28f eq 9000hz +4db q0.8 high-self
rs28f level -4db
satori LR12 2300hz crossover
rs28f LR12 3300hz crossover
Satori LT + HP BW6 37hz
satori flattening eq around 1100hz

The walls do vibrate when you touch them, but I think this might actually be better with EVA since I suspect the vibration is absorbed and damped (like a limp membrane bass trap) rather than resonating it back out like a stiff material (wood, metal) would do.
But I am still experimenting with the build and the material.
 

Attachments

  • satori-rs28f-20150725-rt60.jpg
    satori-rs28f-20150725-rt60.jpg
    40.1 KB · Views: 113
  • satori-rs28f-20150725-ir.jpg
    satori-rs28f-20150725-ir.jpg
    46.7 KB · Views: 969
  • satori-rs28f-20150725-gd.jpg
    satori-rs28f-20150725-gd.jpg
    37.9 KB · Views: 127
  • satori-rs28f-20150725-csd.jpg
    satori-rs28f-20150725-csd.jpg
    156.5 KB · Views: 154
  • satori-rs28f-20150725.jpg
    satori-rs28f-20150725.jpg
    70.8 KB · Views: 757
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Jeshi,
Nice work! Can you please show harmonic distortion up to 5th at 1m and 85dB reference level? I wonder why there is ringing in your impulse response ? The silk domes are usually pretty well controlled or is this the aluminum dome? Also, can you expand the vertical scale on the group delay to max 50ms with 5 ms minor divisions? If you want to show plots inline with text, use the image link icon and paste the URL of your uploaded image thumbnail from below. Makes it easier to discuss figures.
Thanks.
 
Great thanks sharing jeshi.

Have some comments to share :).

I could be wrong but looks like your phase turn is not replicating what a LR2 aught to show. In REW find and hit the bottom "Estimate IR Delay", this would move your negative step (tweeter reversed polarity) seen in your IR plot to time zero instead of now it's approx. at minus 60uS. Think if you do this would flatten much the positive wrapping phase trace from 7kHz and up and at same the 90º (180º/2) turn relative to zero º phase scale would much more replicate your real acoustic XO frq point which is supposed be 2300Hz. To clear what i mean is the 180º phase turn centered around choosen frq XO point is not reflected in your plot, so either above would help phase align or the real acoustic slopes is not as what was targeted.

It's my experience never use that bottom "Estimate IR Delay" when measuring driver as select woofer or tweeter by itself but when the whole system is measured then to have correct phase plot IR need alignment at point zero.

To compare below is FR/Phase/IR/SR for a electric LR2 2300Hz where tweeter have reverse polarity via electric DSP loopback listened with REW in loop. System bandpass is set to 40-23000Hz but at HF area my soundcards 48kHz sample rate would be a limiter and press a little on phase turn up there. That 90º phase turn relative to zero º is at 2300Hz is what i miss in your plot in real world.

Last two plots are Harsch XO filter type as xrk971 suggest you should try.
The HP should be a Bessel slope but in below i keept a 2300Hz HP LR2 and BW4 2300Hz for woofer, then a 0,21739mS delay on tweeter give below nice FR/Phase/IR/SR. If trying this XO type remember polarity should be same for both drivers and the physical real world offset for woofer and tweeter acoustic center offset at 2300Hz should be added or subtracted the 0,21739mS delay on tweeter.
 

Attachments

  • FR.png
    FR.png
    33.5 KB · Views: 102
  • IR_SR.png
    IR_SR.png
    32.5 KB · Views: 77
  • IR2.png
    IR2.png
    26.6 KB · Views: 70
  • FR2.png
    FR2.png
    29.6 KB · Views: 65
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Byrtt,
I love your loop back sound card simulations (or should I say, idealized experiments)! You have shown that with proper textbook response shaping on drivers that one can achieve essentially transient perfect response with the Harsch XO topology. Very nice and thanks for showing. The contrast with a pure LR2 is obvious - the initial tweeter kick is in the wrong direction! How can that possibly sound as dynamic? Yet so many speakers use LR2 as the XO topology.
 
: ) agree those real world loopbacks are great to have confirmation a topology or use as textbook target curves that are overlaid in REW with real live measurements to hit them most precise.

Hehe see below the initial tweeter kick for LR2 can be repaired to right direction by reverse polarity for woofer instead of tweeter but then other compromise from perfect show up :rolleyes:
 

Attachments

  • IR_SR2.png
    IR_SR2.png
    32.4 KB · Views: 78
Founder of XSA-Labs
Joined 2012
Paid Member
Yes but this is worse because the real impact of the woofer kick is now a rarefaction instead of a compressive wave. Less impactful I believe. I mean you will see the woofer cone suck into the cabinet on a kick drum. That seems opposite of the pressure wave in reality. Can we detect difference between compressive pressure wave vs rarefaction wave? Not sure but does not seem right.

I want to get a low cost USB sound card so that I can try these "experiments" like you.

I wonder if this unit can help me realize these loop back tests?

Amazon.com: BEHRINGER U-Control Uca202 Ultra Low-Latency 2 In/2 Out Usb/Audio Interface With Digital Output: Musical Instruments

I will start a new Harsch XO thread so we don't pollute Jeshi's EVA foam thread.
 
Last edited:
thanks! Yeah this raw white foam does look very nice. I was originally planning to paint the speakers, but the the white looks so nice I think I will keep it like this.

For surfboards, do you use EVA or a polystyrene foam?

Hi jeshi,
For boards, most of the industry uses open cell PU foam as it can be blown into blanks cheaply.
The Homeblown closed cell PU foam is more expensive, but has none of the disadvantages open cell - one of it's major advantages is that it's density is far more consistent through the blank ie:
It's 'cut feel' does not change from skin to core.

I've not thought about finishes other than glass (or some form of 'fabric') / poly resin , but the foam will yellow with age and get dirty Very quickly if left untreated.
Traditionally board makers have used either cellulose or water based poster paint sprayed on or / and tinted resin - I Brush paint blanks with Daler Rowney 'System7'* artists acrylics thinned with good old tap water and a splash of 'flow enhancer' to stop beading / titting (globules) happening.
Acrylic hardly smells, a bonus for us allergy sufferers;)

Question: How do You mount your drive units, do you use 'hard inserts' let into the foam and if so, what do you use?

*Daler Rowney White is just as 'snow white' as foam ;)
 
Last edited:
Question: How do You mount your drive units, do you use 'hard inserts' let into the foam and if so, what do you use?
thanks for the info on surfboard making techniques.

Yes I use hard inserts. I use a local japanese version of this for softwood.
Threaded Inserts for Soft Wood | E-Z LOK

Softwoods.jpg

I use the style on the left without the top lip so I can recess them. I use M5 for the drivers and M6 for my removable front baffle. These can screw into the foam. So far they hold really well, but I originally though I would need to glue them into place.
 
H


Yes but this is worse because the real impact of the woofer kick is now a rarefaction instead of a compressive wave. Less impactful I believe. I mean you will see the woofer cone suck into the cabinet on a kick drum. That seems opposite of the pressure wave in reality. Can we detect difference between compressive pressure wave vs rarefaction wave? Not sure but does not seem right.

I want to get a low cost USB sound card so that I can try these "experiments" like you.

I wonder if this unit can help me realize these loop back tests?

Amazon.com: Behringer UCA202 Audio Interface: Musical Instruments

I will start a new Harsch XO thread so we don't pollute Jeshi's EVA foam thread.

Because off topic will not comment much on regarding rarefaction instead of a compressive wave makes difference but remember how wesayso in PM said only some Van Halen tracks is really good and some even benefitted from phase reversal and it seemed every time it sounded as David Lee Roth sang from behind the drums and that is not right.

Funny you point to cheap UCA202 at this desktop computer used for previous loopback plots i use a cheap big value UCA222 USB soundcard (the red painted one) and it measures just as fine as a more expensive M-Audio PCI card or much much more expensive ASUS PCI card or a middle priced ASUS USB card.
 
Jeshi,
Nice work! Can you please show harmonic distortion up to 5th at 1m and 85dB reference level? I wonder why there is ringing in your impulse response ? The silk domes are usually pretty well controlled or is this the aluminum dome? Also, can you expand the vertical scale on the group delay to max 50ms with 5 ms minor divisions? If you want to show plots inline with text, use the image link icon and paste the URL of your uploaded image thumbnail from below. Makes it easier to discuss figures.
Thanks.

I am pretty sure the ringing is because of my physical time-alignment (maybe not perfect). Or maybe the waveguide, or the DSP. Here are some IR of just the RS28f in the baffle.

Here is the IR of just the RS28f in the deep 3cm waveguide with no EQ. It is showing some strangeness.
495420d1437840057-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-noeq-ir.jpg


And here is just the IR of the RS28F in the waveguide after the DSP is doing a LR 12db/oct crossover at 3500hz (satori is muted). No other eq is applied here, just the CO. It looks like all of the ringing is happening at this point. Maybe some of the additional ringing is caused by a mis-adjusted time alignment of the satori and rs28f.
495421d1437840057-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-3500hz-ir.jpg


And for comparison the IR from both driver with the full DSP crossover and EQ. I invert the phase of the rs28f for the crossover.
495386d1437825628-eva-foam-performance-speaker-enclosures-satori-rs28f-20150725-ir.jpg


I suspect the foam carved waveguide might be causing some problems.
I will make up another test baffle soon. I picked up some cheap EVA floor tiles so I can experiment more freely with some alternate baffles.

I am not sure how to do harmonic distortion in REW. Or should I be using another program for these kinds of measurements.

Oh I placed an order for a pair of ScanSpeak 10f/8424 full range drivers. So I will be doing some builds with them in the near future.
 

Attachments

  • rs28f-3cm-waveguide-noeq-ir.jpg
    rs28f-3cm-waveguide-noeq-ir.jpg
    47.5 KB · Views: 768
  • rs28f-3cm-waveguide-lr12-3500hz-ir.jpg
    rs28f-3cm-waveguide-lr12-3500hz-ir.jpg
    49.2 KB · Views: 997
Last edited:
Yes but this is worse because the real impact of the woofer kick is now a rarefaction instead of a compressive wave. Less impactful I believe. I mean you will see the woofer cone suck into the cabinet on a kick drum. That seems opposite of the pressure wave in reality. Can we detect difference between compressive pressure wave vs rarefaction wave? Not sure but does not seem right.

If you're reproducing bass-range instruments with asymmetrical waveforms, it does make a difference. My electric bass sounds sounds "pillowy" and "punchy" with deep bass one way, and "hard" and "grunty" the other way (reversed polarity).
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
I am pretty sure the ringing is because of my physical time-alignment (maybe not perfect). Or maybe the waveguide, or the DSP. Here are some IR of just the RS28f in the baffle.

Here is the IR of just the RS28f in the deep 3cm waveguide with no EQ. It is showing some strangeness.
495420d1437840057-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-noeq-ir.jpg


And here is just the IR of the RS28F in the waveguide after the DSP is doing a LR 12db/oct crossover at 3500hz (satori is muted). No other eq is applied here, just the CO. It looks like all of the ringing is happening at this point. Maybe some of the additional ringing is caused by a mis-adjusted time alignment of the satori and rs28f.
495421d1437840057-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-3500hz-ir.jpg


And for comparison the IR from both driver with the full DSP crossover and EQ. I invert the phase of the rs28f for the crossover.
495386d1437825628-eva-foam-performance-speaker-enclosures-satori-rs28f-20150725-ir.jpg


I suspect the foam carved waveguide might be causing some problems.
I will make up another test baffle soon. I picked up some cheap EVA floor tiles so I can experiment more freely with some alternate baffles.

I am not sure how to do harmonic distortion in REW. Or should I be using another program for these kinds of measurements.

Oh I placed an order for a pair of ScanSpeak 10f/8424 full range drivers. So I will be doing some builds with them in the near future.

The 28F looks as I expected it to by itself. Maybe you have a resonance in the waveguide or diffraction somewhere.

HD data is a tab at top in REW. Later version 5.0 and later so maybe if you can't find it then download latest version. Click just the 2nd thru 5th harmonic components. Don't need to see to 10th harmonic.

Here is example of HD in REW from my 10F/RS225.
495051d1437631905-10f-8424-rs225-8-fast-ref-monitor-harsch-10f-rs225-hd.png
 
Last edited:
The 28F looks as I expected it to by itself. Maybe you have a resonance in the waveguide or diffraction somewhere.

HD data is a tab at top in REW. Later version 5.0 and later so maybe if you can't find it then download latest version. Click just the 2nd thru 5th harmonic components. Don't need to see to 10th harmonic.

I think the final IR and the "RS28f with DSP LR2" have the same ringing. The full speaker has a tiny bit of pre-ring, but it looks like most of the ringing is happening after I apply the DSP LR2 3500hz. What do you see that I am not? These two look basically the same to me (outside of the inverted phase in the full speaker)
495421d1437840057-eva-foam-performance-speaker-enclosures-rs28f-3cm-waveguide-lr12-3500hz-ir.jpg
495386d1437825628-eva-foam-performance-speaker-enclosures-satori-rs28f-20150725-ir.jpg


OK my version of REW was 5.0 and didn't have the HD. I updated.
495437d1437843016-eva-foam-performance-speaker-enclosures-satori-rs28f-20150725-hd2.jpg
 

Attachments

  • satori-rs28f-20150725-HD2.jpg
    satori-rs28f-20150725-HD2.jpg
    96.6 KB · Views: 964
Last edited:
thanks for the info on surfboard making techniques.

Yes I use hard inserts. I use a local japanese version of this for softwood.
Threaded Inserts for Soft Wood | E-Z LOK

Softwoods.jpg

I use the style on the left without the top lip so I can recess them. I use M5 for the drivers and M6 for my removable front baffle. These can screw into the foam. So far they hold really well, but I originally though I would need to glue them into place.

Ahhh, I never thought of threaded metal inserts!
Used in the denser foams, those would hold some fair size mid/bass drivers!

If I ever find the time, I would like to make a video of 'foam working techniques' - things like how to make radiused edges that go round curved panel lines, how to make them symmetrical from one side of the 'form' to the other and 'sanding blocks', lengthy posts just don't cut it for descriptions of use / techniques.
I'm not 'just' making surf boards, anything that can be made from foam / fabric goes..... motor car / bike body work panels, boats, aircraft.
Even skateboard decks can be made from these materials, a friend and I have made foam core / multi carbon layup racing downhill decks - they are hugely expensive to produce, but then we're verging on F1 / Aerospace tech.
They transmit the 'surface feeling' (what the boards rolling over) and cornering grip in a Very special way, but dont transmit the vibes that wooden decks do - the ones than make your feet go numb and make the board 'jitter' and wander off line in corners.

You are seeing the benefit of this type of foam already, IF you / we can find a way of laminating without the nasties of carbon/glass and poly/epoxy on a kitchen table, the gains in controlling resonance of panel / damping of resonances the gains could well exceed ones wildest dreams.

I have some small off cuts of 120kg and 200kg foam laying about, I'll start sticking 'stuff' to it.... will have to wait till Monday to buy lighter weight foam / bonding agents BUT there might just be some stuff laying about some where :scratch:
 
You are seeing the benefit of this type of foam already, IF you / we can find a way of laminating without the nasties of carbon/glass and poly/epoxy on a kitchen table, the gains in controlling resonance of panel / damping of resonances the gains could well exceed ones wildest dreams.

awesome. I still want to explore the EVA floppy walls a little more (is it really like a limp membrane absorber). But wow you might be right about carbon/glass/epoxy/foam enclosures and controlling damping/resonances. Unfortunately I don't have any experience with glass/kevlar/carbon layup. I might consider learning it though.

I will be trying neoprene inner-wall lining once it arrives from aliexpress. The extra mass/damping of the neoprene combined with the EVA walls might also give a large performance gain. I may also try an EVA/neoprene/EVA laminate wall.

I think the basic concept of sound damping is to absorb the sound waves and convert to heat. That means that the material has to vibrate (capture the sound, or resonant with the sound) and then to damp the vibration quickly (turn into heat) and not to resonate/ring back into sound.
 
Just a thought that it might also be advantageous to use the same corning fiberglass board to line the inside of the cabinets just as is done with wood enclosures. Trying this or other materials like No Rez just might have some magic to them in making a reference level EVA enclosure.
 
awesome. I still want to explore the EVA floppy walls a little more (is it really like a limp membrane absorber). But wow you might be right about carbon/glass/epoxy/foam enclosures and controlling damping/resonances. Unfortunately I don't have any experience with glass/kevlar/carbon layup. I might consider learning it though.

I will be trying neoprene inner-wall lining once it arrives from aliexpress. The extra mass/damping of the neoprene combined with the EVA walls might also give a large performance gain. I may also try an EVA/neoprene/EVA laminate wall.

I think the basic concept of sound damping is to absorb the sound waves and convert to heat. That means that the material has to vibrate (capture the sound, or resonant with the sound) and then to damp the vibration quickly (turn into heat) and not to resonate/ring back into sound.

The high tech laminates / carcassing end has been left out of speaker design* - yes, things like the SL600's, OverKill Audio 'Prey' enclosure, B&W's silly money pieces have been done, but on the whole it's a vastly under visited area of material use mostly because the components are quite horrid to work with.
You would not catch me making a carbon + resin / foam object on MY kitchen table - but if you have a place to work these things, an open mind and 'the dare to go there', worlds your oyster.

Regarding your last point:
I 'think' CSPU foam has Everything needed in the way of 'required properties', as your 'laminating' the sheet to gain thickness, the glue layer it's self will have an impact on how resonances are absorbed / controlled.

Maybe we should be thinking of this foam in the same way as sand fill between a box inside a box..... but without the silly weight!
The sand is the 'foam bubbles' and the 'wall' is the foam structure it's self!
Looking at it this way is quite 'Wow'!

Maybe there will / wont be any sonic gain with neoprene between layers - maybe it could work the opposite way round ie:
A single layer of giftwrap paper laminated between two sheets of foam 'might' be 'better'.... maybe better to (from outside > inside) lay-up foam>paper>foam>foam.
:confused::scratch:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.