Small 2-way speakers with linear on-axis and power response characteristics (Scan Speak and SB Acoustics drivers). H&V off-axis measurements included

[HighFutility]

One possible drawback i can think of is distribution of internal modes. As the height becomes close to depth the frequencies of relevant modes start to overlay:
View attachment 449471
for comparison, the modes of the original box:
View attachment 449470

On the other hand, internal modes didn't turn out o be a problem in the original design (due to heavy damping) so perhaps this won't be a big problem here either.

Thank you for your ongoing contributions and for taking the time to reply to my post. I am a complete beginner, so I don't comprehend the aforementioned. Will the box depth increase the resonant frequencies to a higher frequency or create more resonant frequencies or both?

 

[Paweł L]

I might have gotten a little carried away. Is this too much dampening…?

View attachment 446551

Actually you could put some Dacron, Rock wool or similar in space behind the tweeter to absorb the height mode. And yeah I would put piece of MDF 1.5-2in wide as a brace - keeps side panel resonances dawn and helps keep the additional damping/stuffing in place.

 

[Paweł L]

Thank you for your ongoing contributions and for taking the time to reply to my post. I am a complete beginner, so I don't comprehend the aforementioned. Will the box depth increase the resonant frequencies to a higher frequency or create more resonant frequencies or both?

Friend of mine build pair of enclosures where the width and depth were similar and almost twice the height - I didn't like sound of midrange.
The shorter the distance between panels, the higher in frequency the resonance that will occur. 0.8 multiple seem to work OK. But the simulation like the one provided by Mechanik does show the modal frequency distribution so the dimensions can be adjusted to spread the resonances and to avoid amplification at certain frequency(ies).

 

[Groove01]

Actually you could put some Dacron, Rock wool or similar in space behind the tweeter to absorb the height mode. And yeah I would put piece of MDF 1.5-2in wide as a brace - keeps side panel resonances dawn and helps keep the additional damping/stuffing in place.

Thanks for the suggestions, I try that!

I went back and forth on the brace, but couldn’t decide on the best location for it. I’ve put the front baffle on now, so need to see if I can figure a way to add it now.

 

[The French]

Mechano23 got stands.

Great, and nice job !
Is it plywood as well ?
Is the front face of the speakers just painted ?

 

[Mark411]

Hi--Here are some photos of my recently completed speakers. I'm very happy with both the looks and sound. Shout outs to:

@XMechanik for the awesome design and testing/refinement;
@amirm for putting the design "on the map;"
@wineds for the PCB layout and .rar design files;
@Woody99 for the amazing how-to manual for his men's shed group in Oz; and
everyone else on the forum who shared information from whom I drew tips and inspiration!

I had the speaker boxes themselves built and finished (with Osmo Polyx) by the amazing Michael Keymont of Machined In Wood in Massachusetts. He does great work. His prices are good, he's able to ship items, and he offers many finishing options including some pretty exotic ones. I believe he also does unfinished flat pack work, too.

 

[Paweł L]

Thanks for the suggestions, I try that!

I went back and forth on the brace, but couldn’t decide on the best location for it. I’ve put the front baffle on now, so need to see if I can figure a way to add it now.



View attachment 449694

Just put 3/4inch thick MDF 2" wide around 55-60% of the height in the middle of side panels and it can on the angle. For the enclosure this size it will be sufficient to reduce the main resonance. You could also add another one just behind the front baffle between midwoof and tweeter. Oh, and I forgot uppss - the port is on the back towards the tweeter so not a good place for more stuffing behind the tweeter, so just some Dacron or like behind the midwoofer will do. Next time you cut the holes on the baffle for the midwoofer, round them of on the inside side or at 45° to reduce the diffraction from the back of the woofer's cone.

 

[rp86]

I am a complete beginner, so I don't comprehend the aforementioned. Will the box depth increase the resonant frequencies to a higher frequency or create more resonant frequencies or both?

If a resonant frequency exists in more than one dimension it is likely to be higher in amplitude than if it was only present in one. This could be a problem if that frequency is likely to be excited (generally meaning with the passband of the bass or mid driver).

 

[HighFutility]

One possible drawback i can think of is distribution of internal modes. As the height becomes close to depth the frequencies of relevant modes start to overlay:
View attachment 449471
for comparison, the modes of the original box:
View attachment 449470

On the other hand, internal modes didn't turn out o be a problem in the original design (due to heavy damping) so perhaps this won't be a big problem here either.

What software can I use to see this? Currently im familiar with WinIsd and slowly learning HornResp.

 

[XMechanik]

What software can I use to see this? Currently im familiar with WinIsd and slowly learning HornResp.

This is one of the Auxiliary tools in VituixCad (Tools>Auxiliary or F8).

 

[tripu11]

I just finished my Mechano23 build (but for painting the box). They sound fantastic. I used a port made of black PVC with an internal diameter of 40mm. I had to use slightly different component values in places because I purchased my crossover parts from Parts Express, but I was able to model the effect of the changes using XMechanik's VituixCad project files to confirm the changes wouldn't have a significant impact on the outcome. Total cost all-in was US$296. This does not include box material and I had 0.82uF capacitors on-hand when I started. I also made my own PCB boards using the ferric chloride etching process with leftover copper boards from another build. A wonderful little project.

Below is the VituixCad six-pack of my final measured results. Note that the SPL is relative, I did not use 2.83v specifically.

I did not use any bracing at all inside the cabinet. I took full measurements of the speaker (1) without any stuffing or damping, (2) with 70g of rockwool insulation and (3) with 70g of pink fiberglass insulation. The three near field port measurements are below (separated for clarity). Green has no stuffing or damping, blue has 70g of rockwool and yellow has 70g of pink fiberglass insulation. Pink fiberglass insulation gave me the best result, so that is what I am using in my final speaker.

I made the rookie mistake of changing the depth as a result of using a slightly wider port, resulting in similar height and depth dimensions, exacerbating the internal box resonance at that frequency (see at about 680Hz in the green measurement). I was originally planning on adding internal hardwood strips glued to the sides and top of the box on the inside (to act in place of bracing, in addition to stuffing) to suppress the box resonance, but it was essentially completely eliminated by the pink fiberglass stuffing so I ended up not bothering. All that is left is the port resonance.

Below is the near field port measurement, rockwool v. pink fiberglass (without separation).

And, finally, the nearfield woofer measurement, no damping or stuffing v. 70g pink fiberglass insulation (separated for clarity):

Thank you XMechanik for a wonderful design.

 

[Heinrich]

One possible drawback i can think of is distribution of internal modes. As the height becomes close to depth the frequencies of relevant modes start to overlay:
View attachment 449471
for comparison, the modes of the original box:
View attachment 449470

On the other hand, internal modes didn't turn out o be a problem in the original design (due to heavy damping) so perhaps this won't be a big problem here either.

How could it display internal resonances without considering the driver's position, and not the least, not explaining where the point of reference is for the sensor to detect that resonance - strange, and the width of the peaks, depending on damping? Won't ever trust it.

I went back and forth on the brace, but couldn’t decide ...

Bracing is a good idea in general. Best found during many extensive experiments with two accelerometers is to connect opposing walls with, yep, just a brace glued on close to the middle. I glue small edges to the walls and then glue the stick (2x2cm^2) in onto that edges successively, so it doesn't have to be of any precision. It works allright in every case, and it won't look at you later, because it's hidden on the inside, right?.

 

[Groove01]

I glue small edges to the walls and then glue the stick (2x2cm^2) in onto that edges successively, so it doesn't have to be of any precision.

Thanks!!

love the idea of gluing edges on each side tas a mount for the brace. That’s totally a “why didn’t I think of that” solution- lol. Thanks for sharing!

 

[HighFutility]

I am in the process of building three variants of this speaker. I have taken in-room measurements and EQ'd a room curve, but I have never taken proper full measurements of a speaker. I have read the forum post https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/ I would be interested in taking full measurements of all three speakers. Other than rew and the mic which I already own. What other items do I need? It's rainy season right now, so measurements would have to be done indoors.

Ikea sells lazy suzans like these for relatively cheap. Would this help speed up measurements at different angles? I could probably carve out or draw in the various degrees.https://www.ikea.com/ca/en/p/snudda-lazy-susan-rubberwood-80543263/?utm_source=google&utm_medium=surfaces&utm_campaign=shopping_feed&utm_content=free_google_shopping_clicks_Eating&gad_source=1&gad_campaignid=1056264848&gclid=CjwKCAjw87XBBhBIEiwAxP3_A1H8Wrmt5r0EREYECXKXpaRHXThv4VNBEn6zzcKzhnjVsYbmQ9OsghoCMm4QAvD_BwE

SNUDDA lazy Susan, rubberwood, 39 cm (15") - IKEA CA

SNUDDA lazy Susan, rubberwood, 39 cm (15") Put a spin on everyday life with this round wooden tray. Give it a twirl to get closer to your favourites. Comfortable and practical – and an easy way to upgrade the kitchen or dining area. Turning function in base makes it easier to serve from.

www.ikea.com

 

[tripu11]

Yes - I made my own using a lazy susan swivel plate and leftover driver cutouts from my IB subwoofer install. Made the process extremely quick. I took my measurements in an empty bedroom. This is the lazy susan swivel plate I used: https://www.lowes.com/pd/RELIABILT-6-in-Square-Lazy-Susan-Turntable/5014535721

 

[The French]

There is a reinforcement inside that connects the centers of the side panels (piece of plywood 135x40x18mm).

Hi,
Sorry if it has already been mentionned, but the reinforcement piece is supposed to be 135mm long, while the internal width is 138mm.
Am I missing something ?

 

[HighFutility]

Hi,
Sorry if it has already been mentionned, but the reinforcement piece is supposed to be 135mm long, while the internal width is 138mm.
Am I missing something

You may or may not want to leave a few mm ~1-2mm for space for glue. I have cut things with no tolerance, and when you apply glue, it just slides off. Even with the precision of a CNC, you may want to leave a gap of 1mm.

 

[carbidetooth]

You may or may not want to leave a few mm ~1-2mm for space for glue. I have cut things with no tolerance, and when you apply glue, it just slides off. Even with the precision of a CNC, you may want to leave a gap of 1mm.

For clarity, filling a 1mm space with conventional wood glue will not add any significant strength at the gap. The bond really needs to be between the parts being glued which requires a minimal glue line. It is possible to starve a joint of glue but hard to imagine that would be the case here. For a better structure, I'd suggest a slightly loose pocket for the brace and utilize something like PL Premium in that joint only. PL Premium X3 is a slow-curing polyurethane construction adhesive that expands a little when curing. I have lots of experience with this product specifically and while they don't claim the stuff is structural, I have used it that way with success in a variety of applications, including speaker building.

 

[HighFutility]

For clarity, filling a 1mm space with conventional wood glue will not add any significant strength at the gap. The bond really needs to be between the parts being glued which requires a minimal glue line. It is possible to starve a joint of glue but hard to imagine that would be the case here. For a better structure, I'd suggest a slightly loose pocket for the brace and utilize something like PL Premium in that joint only. PL Premium X3 is a slow-curing polyurethane construction adhesive that expands a little when curing. I have lots of experience with this product specifically and while they don't claim the stuff is structural, I have used it that way with success in a variety of applications, including speaker building.

Thanks for the clarification.

 

[HighFutility]

@carbidetooth I figure you're the go-to person for this. This morning, I tried putting a round over on my cabinets. Disaster struck. I got a nice tearout and gouged the front baffle. Cabits are a write-off. Will post pictures shortly. I have only worked with MDF, so I could easily patch with Bondo, sand, and veneer over.

Baltic birch is going to be a bit of a pain for me. What are good round-over bits that prevent tearout? A Frued Quadra bit? I have a 1/4-inch shank.