The Schlafmutzes Mini MT’s design and progress

[JonW]

In earlier posts I described how there might be some cabinet reflections happening inside, so the helpful advice was to add some felt to absorb things up. Here is what I did and also what I recommend others do if they are going to build these speakers. I will add these steps and pictures to the build tutorial on page 1 of this thread.

I used F-13 felt, 1/4” thick and 2” wide strips purchased from McMaster-Carr. 3 feet should be plenty for 2 speakers. It's about $3 plus shipping. Maybe order this with the various black screws that you will need for fastening the drivers to the cabinets.


The W16 woofer has a rubber boot on the back. Peel it off. Don’t worry. It comes off easily. The ER15 woofer does not have this boot so no need to remove it. Otherwise, do everything the same as shown here for the W16.


See. It’s off.


Cover the back with felt. I used one strip (2” wide) and then another strip that was cut to half width (1”). I glued the felt to the driver using hot melt adhesive. It stuck very well.


Then cut a strip of felt that will wrap about 2/3 of the way around the magnet. We do not want to wrap around the entire magnet or else the assembly won’t fit through the cabinet hole. I also have the woofer leads on the bottom here, in the area not covered by the felt. You might want to have your cabinet pilot holes for the woofer screws drilled and positioned with all of this in mind.


I also glued down, with hot melt adhesive, a strip of felt on the bottom of the cabinet, touching the front.


After gluing the felt to the magnet, here is what it looks like.

[JonW]

The really small cabinet size of these speakers means that space is at a premium for the crossovers. So I’m using 20 gauge inductors (and occasionally 18 gauge when the 20 gauge is not available in a specific value). Here is a photo showing two 0.3 mH inductors. On the left is a 14 gauge example. On the right is a 20 gauge inductor. A CD case is in the background for scale. The size difference is huge. So we’re definitely going with the 20 gauge for this project. There will be a little more DCR to contend with. But looking at the crossover, we are not loosing much at all. As an extra bonus, the 20 gauge inductors are cheaper.

[JonW]

Here is the voicing setup. I have been relegated to the basement since there is a little kid running around the house now. I can’t exactly leave this contraption in the living room for days or weeks while I do the voicing. "Come here little guy. Put your finger right there." Zap!

[JonW]

Progress is happening more quickly than I can post about it all, at the moment. A good thing. The latest updates…

-The crossover for the deluxe (W16 woofer) speaker for in room placement is really done. No more changes from anything described above. I’ve listened to it more and I am really happy with where it is at. Some day I’ll post a detailed writeup with all the relevant plots and explanation of the circuit and how each part influenced the sound from voicing.

-I started on the computer modeling of a crossover for the basic (ER15 woofer) speaker for in room placement. It looks great so far. In the end, it’s quite similar to the deluxe crossover. Even the woofer part. Go figure. I want to give it another round of modeling but it’s probably done. Then I might soon move on toward voicing this crossover before I post up all the plots and circuit. However, if anyone happens to be building this flavor of the speakers and needs to see all the data soon, speak up and I will gather up everything and post it here. (I’d be really surprised if anyone was building these but you never know.) The basic on wall crossover is still to be made.

-Generally speaking, I am finding that this Scanspeak D3004/602010 tweeter in this cabinet is a little more difficult to work with than the Scanspeak 6600 tweeter in a larger cabinet. In the end, everything works out very well. But more parts are needed to get things under control. Or perhaps I am being more picky this time around.

-Stuffing or cabinet dampening can be an important part of any speaker build. Some dampening may be needed to tame internal reflections in the cabinet and such. But with my last project, I found that too much dampening hurt things. I overstuffed the cabinet (using OC703) and the speaker sounded dead and dull. I removed some stuffing and that air and magic returned. So I have voiced the deluxe in room speakers here with varied amounts of stuffing. As mentioned earlier in this thread, the cabinets are small and the measurements showed the possibility of minor problems from internal reflections. So I retook measurements again with some 1.25” Sonic Barrier from Parts Express. I used a piece to cover the whole rear of the cabinet except for a section cut out to make room for the binding posts (I have 4 binding posts.), about 5” x 9”. Photos below show how it all looks.

Voicing went well with the Sonic Barrier there. But I wanted to see if the cabinet was overstuffed. So I removed this big piece and replaced it with a smaller 3” x 5” piece. I also listened with no Sonic Barrier at all. With none in there, a tiny amount of air showed up. But the speakers also sounded a bit busy and congested. Almost distorted. I’m probably splitting hairs here because they still sounded great. And all of this is likely just in my head. But it seemed as though pressure waves from the woofer were bouncing off the rear and hitting the woofer to confuse the driver while it was playing the next bit. Hence the congestion and apparent distortion. You don’t believe me? I wouldn’t’ either, if I was reading this. But they sounded better, overall, with the Sonic Barrier in there. So we’re keeping it. When the 3” x 5” piece was in there, the congestion sound persisted. But it was gone with the full 5” x 9” piece. So I’m happy to have found a good outcome here. The Sonic Barrier is $27 per sheet and one sheet should be enough for 4 pairs of speakers. This Sonic Barrier is in addition to the felt, shown above.

[technodanvan]

It's looking really good man, keep it up!

I like seeing these smaller speakers, I wish more people would concentrate on this and even smaller form factors.

[cjd]

Looking great Jon!

I'm curious if it's the corner treatment that's helping congestion - would a single strip of sonic barrier in the corners do the trick? Stirring the pot, I know! But if you feel they're a little overstuffed still... Of course, I also wonder if the felt on the magnet would be better placed on the walls snug up against the baffle - trying to manage first reflections off the sides. I guess stuff I can experiment with in my own small designs. If I had any...

It's looking really good man, keep it up!

I like seeing these smaller speakers, I wish more people would concentrate on this and even smaller form factors.

Small is tough at some level, and a little boring. I'll be giving the Pecorino a more thorough revisit along the way I think, and I've not seen a smaller overall build (though some of the M3 designs get dang close.) But in the end I wish they dug deeper than they do - I guess that's why I use the sub (upfiring RSS210 crossed ~120Hz I think)

[technodanvan]

Small is tough at some level, and a little boring. I'll be giving the Pecorino a more thorough revisit along the way I think, and I've not seen a smaller overall build (though some of the M3 designs get dang close.) But in the end I wish they dug deeper than they do - I guess that's why I use the sub (upfiring RSS210 crossed ~120Hz I think)

Hey, I'm okay with a 2.1 system!

[JonW]

It's looking really good man, keep it up!

I like seeing these smaller speakers, I wish more people would concentrate on this and even smaller form factors.

Thanks! Small form factor introduces some interesting design constraints. Small is easy. But small with bass extension to reach a sub (at, say, 80 Hz) is a whole different matter. Low distortion is another. I'll post the near field woofer measurements at some point. Both both the Deluxe and the Basic speakers here play down to 80 Hz and should mate with a sub well.

[JonW]

Looking great Jon!

I'm curious if it's the corner treatment that's helping congestion - would a single strip of sonic barrier in the corners do the trick? Stirring the pot, I know! But if you feel they're a little overstuffed still... Of course, I also wonder if the felt on the magnet would be better placed on the walls snug up against the baffle - trying to manage first reflections off the sides. I guess stuff I can experiment with in my own small designs. If I had any...



Small is tough at some level, and a little boring. I'll be giving the Pecorino a more thorough revisit along the way I think, and I've not seen a smaller overall build (though some of the M3 designs get dang close.) But in the end I wish they dug deeper than they do - I guess that's why I use the sub (upfiring RSS210 crossed ~120Hz I think)

Thanks, Chris! They're not overdampened now. I was a tad concerned about that, but playing around with the different amounts of Sonic Barrier and felt have me happy with where they are. No idea if just strips in the corners would be enough. Interesting question. The rear and front baffles are parallel to each other, for whatever that is worth. I bet the whole reflection game would change if I made the rear baffles perpendicular to the floor and not parallel to the front baffle. The cabinets would look a bit odd, though. I've taken the woofer on and off so many times, at this point, that I'm concerned about stripping the wood. I'm quite happy with the stuffing level now, so I'm hoping for only one more on and off of the woofers, to install the final crossover. Some day.

Yes, felt on the sides could work. Probably about as well as what I have now. I prefer it this way only for practical reasons. The woofer hole is only a few mm's from the inside cabinet edges. Having the felt around 2/3 of the magnet allows me to ease the assembly into the cabinet, with some wiggling. I tried a full strip of felt around the whole magnet circumference and it just would not go in. I tell ya, keeping these speakers small presents some challenges. At least they sound good.

Please start a thread on the Pecorinos when the time is right. I'll be interested to read along on another small speaker build. For me, mating with a sub at a good level of 80 Hz was important. I don't know if speakers can be too much smaller than these and still play fully down to 80 Hz. Well, there are probably drivers out there capable of that. But probably with higher distortion as a tradeoff.

[cjd]

There is always a day when good enough is done, and nothing more is helpful unless you have a plain academic interest in continued experimentation.

As for the Pecorino:

https://www.htguide.com/forum/showthread.php?39093

http://divine-audio.com/pecorino

[JonW]

There is always a day when good enough is done, and nothing more is helpful unless you have a plain academic interest in continued experimentation.

As for the Pecorino:

https://www.htguide.com/forum/showthread.php?39093

http://divine-audio.com/pecorino

Yup, I've found a good amount of stuffing and I'm going with that. I tried several variations and I'm happy with where it ended up. And it's good to know that I actually tried several things. It wasn't just a guess. As for the academic interest thing, I get plenty of that at my day job.

Those Pecorinos look cool! Using the metal box is very clever. I like that idea a lot. Was that wood bracing really needed- are those boxes not already free of vibrations? I wonder if there are larger metal boxes available for larger speakers. Provided that the metal box is sufficiently strong and free of vibrations. Could open up some interesting ideas.

[JonW]

Chipping away a little more... I’ve got a crossover modeled for the basic, in room speakers. Clipped it together and did a little listening. They do sound nice. But, to be honest, not as nice as the deluxe, in room speakers. I was expecting them to be pretty close, closer than what I’m finding. But the bass is not at all lacking versus the W16. Actually, the ER15 might be reaching a little deeper. Despite the predictive modeling suggesting otherwise. I need to look back at the near field measurements I took, but I think that they might show a tiny bit more bass extension with the ER15 than the W16.

I can tell where all the money went, going from the basic ER15 woofers at ~$82 each to the deluxe W16 woofers at ~$225 each. The ER15 sounds more distorted, in comparison. Looking back on it, I probably should have listened to the basic speakers first. So as not to set my expectations too high.

I’ve got a little of the voicing done, trying to work with the ER15. It’s taking a different touch than the W16. In the end, the W16 got a very flat frequency response. We’ll see what works best with the ER15. But so far, it’s actually looking like a sloped frequency response, higher in the bass and lower in the treble. Interesting. I’ll post more ER15 versus W16 impressions when I have the basic crossover voiced and complete. It might be premature to say too much now from listening without a finalized crossover.

[JonW]

Final crossover for the deluxe (W16 woofer), in room speakers

Here is my masterpiece. I am quite happy with the way these things turned out. All voiced and everything. Let’s take the tour.


The circuit:


Both the tweeter (C2, C3, L2) and the woofer (L4, L5, C6) are third-order electrical filters. Acoustically, they both hit fourth-order Linkwitz-Riley (LR4).

R2, R4, R8, and R9 are there to account for the DC resistance of the inductors. They are not physical parts of the crossover, so do not add them in. Using the small 18 and 20 gauge inductors, to fit into these tiny cabinets, makes for high DC resistance, so it’s been accounted for.

R1 is a tweeter padding resistor. Changing the value here is an easy way to tune the speakers to have more or less treble. Add an Ohm for less treble, take one away for more. As it stands, there is a nice balance of highs versus mids. (Not much bass to speak of in these speakers.)

The C5 cap helps kill the woofer resonance/floppiness around 5,000 Hz.

There are two Zobels (C7, R11 and R12, C8 ) here. And an LCR filter (L1, C1, R3). Plus another filter (R10, L6). They each contribute a little bit with regard to keeping the frequency response flat and matching phase of the two drivers. Some of them help to keep the flattish impedance. I know that it’s a lot of parts, but they each really do help contribute to maximizing the performance of these drivers.



If you want to see how the voicing went, here is a summary of what I found:

-I listened with and without the tweeter LCR filter. It provided a little better imaging and less sibilance. A small but noticeable effect.

-Same story with the tweeter Zobel. Better imaging and less sibilance with it there.

-The tweeter tiny inductor (L6): Without it, things were a little less lively.

-The tiny woofer cap (C5): Without it, there were too many highs. Like a tweeter running a little hot and noisy.

-Woofer Zobel: A tiny bit hollow and less full sounding without it. But a very small effect.

-R1 at 7 Ohms: I tried various values here and 7 hits the balance well. I measured these resistors with my cheap LCR meter and both came out to 7.4 Ohms. No idea which is more accurate- the label on the resistors or my meter.

-C2 at 32 uF: Lower values like 29 uF are livelier but also more fatiguing. At 35 was too dull, a tiny bit.

-C3 at 12 uF: A higher value like 14 uF could work, a little livelier but might get fatiguing.

-C6 at 19 uF: Higher values like 22 uF thinned out the sound. Lower like 18 uF was a little livelier but might get fatiguing.

-L2 at 0.2 mH: 0.1 is too bright and 0.3 is too thin.

-L5 at 0.5 mH: 0.6 is pleasant but a tad too busy.

-L4 at 3.0 mH: Just what we are after. 2.7 and 3.3 mH were both not as good.

All in all, everything sounds just wonderful to my ears. Balanced and super, super clear. Not fatiguing. Just what I was looking for.



Frequency response:
These speakers sound best very flat. No smoothing here and they are almost at +/- 1 db.



Phase:
Pretty well matched up.



Impedance:
It’s pretty smooth and flat, with no big peaks or swings above 100 Hz. That’s what I like to see. The lowest point is about 6 Ohms. These speakers are easier for an amp to drive than the Basics (details on those coming soon).



Tweeter reverse null:
Almost -30 db, which looks good to me.



Off-axis response:
Using measurements taken 30 degrees horizontal.

[JonW]

Final crossover for the basic (ER15 woofer), in room speakers

The general crossover topology here is similar to the deluxe, in rooms. But they sound different and a lot of things changed. This ER15 woofer definitely needed more work to sound good than the W16. Let’s walk through it all.

The circuit:


R1 is a tweeter padding resistor. Adjust it to taste, to get more or less treble. I like it where it is at 6 Ohms.

The electric filters for the tweeter (C2, C3, L2) and woofer (L4, L5, C6) are electrical third-order. Acoustically, they came out right at LR4.

The Zobels (C7, R11 and R12, C8 ) and filters (L1, C1, R3 and R10, C9, L6) all helped bring things under control and sound good.

The resistors R2, R4, R8, and R9 are not real. Do not add them in. They are there to account for the high DC resistance created by using these small 18 and 20 gauge inductors, needed to fit the crossovers into the tiny cabinets.

Lots of parts but I think that they turned out sounding good.


Voicing:
So here is the story with voicing these speakers. When I first started playing them I was not happy at all. Kind of dull, lifeless, bloated bass, and definitely more distortion relative to the deluxe speakers. After a lot of component changes, I brought them back to life and I quite like them now. Some thing that I found along the way:

-R1: Modeled flat at 4 Ohms but ended up at 6 Ohms for a proper level of treble relative to the woofer.

-L4: Lower values thinned out the bass. Started out at 3.0 from the model but ended up at 3.3.

-L5: Started at 0.6 but 0.5 is a little more clear, a little more balanced.

-C6: Started at 15 uF and creeped up to 20 uF for better balance, less harsh highs, and generally a nice sound. After all the changes mentioned so far, the speaker started to come together and sound like something nice.

-C2: Started at 28 and ended up at 32 for improved clarity.

-C3: Started at 10 and 12 provided more clarity.

-L2: Stayed at 0.2. Lower and clarity was lost.


Frequency response:
Hey look, it’s not flat. The starting model had them as flat as the deluxe one above. But this is how they sound best. The deluxes, however, sound better flat. Go figure.



Phase:
A decent matchup of the drivers.



Impedance:
It looks OK, but I did not prioritize the impedance plot with this speaker design as much as I did with the deluxes. Nothing below 6 Ohms but a little lumpy above 2,000 Hz. Looks OK to me. But beware. This speaker is hard to drive. I played them through a receiver- a Yamaha RX-V659, 120 Watts 2 channels driven at 8 Ohms and 200 Watts 2 channels driven at 4 Ohms. They play fine until you crank up the volume. Then the receiver shuts down, in protection mode. This never happened with the deluxes.



Reverse tweeter null:
OK, this one is interesting. When I had the computer model, before voicing the speakers, the frequency response was flat. And the tweeter reverse null was -25 db. It looked good on the computer screen. But didn’t sound great. So I changed a lot of component values, as described above. And guess what? By ear, this reverse null went from -25 db to almost -40 db. Who would have thunk it? I guess my ears are worth something.



Off-axis response:
Using data measured 30 degrees horizontally off-axis.

[JonW]

Some thoughts on the deluxe (W16 woofer) versus the basic (ER15 speakers). Everyone views such things differently. But here is where I am at, for the moment.

-I really do prefer the deluxes. The difference is not subtle. But both speakers came out well. Both sound very good. For me, I like the extra clarity, low distortion, and tighter bass that the W16 woofer offers. It’s a noticable difference. Although the ER15 is supposed to be pretty good. At ~$82 each it’s well into the high value range. I probably should have listened to the basics before the deluxes. So as not to be so spoiled. If you don’t hear the deluxes first, you might be OK. When I first heard the basics, I was underwhelmed. But then with the voicing, I brought them to life. They do sing well now.

-The deluxes are going to be maybe $300-350 more per pair. (Specifics to follow, in the future.) That’s a lot. So don’t build the deluxes if it’s going to mean that your kids go hungry or you can’t make the rent.

-The ER15 and the W16 woofers fit into the exact same cabinet hole. So you could start with the ER15’s and, some day, upgrade to the W16’s with the same cabinets. But the crossover changes, too.

-The deluxes are easier for an amp to run, I found. When playing through a receiver, the deluxes played fine. As did the basics. But when the volume was turned up high, the basics shut down the receiver, putting it into protection mode. The deluxes did not do that. You might want to consider what you are using to drive the speakers before deciding on basics versus deluxe.

-Bass extension. For these cabinets, the W16 modeled to reach deeper, solid to 80 Hz. The ER15 modeled to not go as deep, maybe only to 95 Hz or so. But I think that the near field measurements I took in the cabinets look as though both woofers hit 80 Hz prior to any roll off. I’ll post those plots overlayed with typical sub crossovers some day. For the moment, I would not let bass extension factor in to a basics versus deluxe decision.

[technodanvan]

Jon,

Not to open a whole new can of worms for you, but I just reviewed this thread (yeah, the whole thing!) with increasing attention as the relevance to my own project is becoming more pertinent.

First off, I just wanted to thank you (perhaps again?) for having such great documentation. I don't know if I will be able to do that, but when the time comes I ask for help I hope I have the patience to take all input - be it critical or otherwise - the way you did.

Would you have done anything differently if your speakers were not only 'near wall' but also on a desk? You would get plenty of low-end reinforcement...but outside of an EQ (and trying to isolate the speakers from the desk) is there anything else that would need doing?

Also, you mentioned the lower budget design were tough to run. I double checked the impedance profile and I don't think I saw anything to suggest they would dip below ~6 ohms...am I missing something when I read it? It just doesn't look like an amp rated for 4 ohm impedance should have an issue. Admittedly, I'm not very smart about such things so I look forward to being educated!

By the way, if a Padawan is to a Jedi, what would you be called when compared to the Dark Lord?

[Evil Twin]

Jon,



By the way, if a Padawan is to a Jedi, what would you be called when compared to the Dark Lord?

That would simply be a Sith apprentice...

[technodanvan]

That would simply be a Sith apprentice...

Touche!

[JonW]

Jon,

Not to open a whole new can of worms for you, but I just reviewed this thread (yeah, the whole thing!) with increasing attention as the relevance to my own project is becoming more pertinent.

First off, I just wanted to thank you (perhaps again?) for having such great documentation. I don't know if I will be able to do that, but when the time comes I ask for help I hope I have the patience to take all input - be it critical or otherwise - the way you did.

Hi technodanvan!

Thanks for the kind words. I hope that you learned something from all of this. Man, I have still not finished this project, as much as I’d really like to. Don’t ask about how busy life is here these days... work is crazy, with more travel this year than I’ve ever had. And things at home are also hectic. As you might be able to tell from the posts above, the in room deluxe and basic crossovers are voiced and finalized. The on wall crossovers are modeled but not yet voiced. After finalizing those, I’ll need to actually solder them up. Then I’ll be done. It’s not too much more work, but finding the time has not been easy. OK to answer your questions...

Would you have done anything differently if your speakers were not only 'near wall' but also on a desk? You would get plenty of low-end reinforcement...but outside of an EQ (and trying to isolate the speakers from the desk) is there anything else that would need doing?

For an on desk situation, yes, I would do one thing differently: Take raw measurements of the speakers on a desk. Then use those data for a crossover. With the bouncing that you get off the desk, I’d expect the crossover to change a bit. I’m not sure if some of the bounce will be at frequencies below the crossover region. In which case you might need to apply the EQ. But with the measurements, you’d at least know how much and where to apply the EQing.

I thought about taking those on desk measurements. But this project already has 4 different flavors so I didn’t want to make it 6. Plus I don’t think that I’d ever place my speakers on a desk. I’d want them to be ear height. So if I end up with desk placement, some day, I’d put the speakers up on some tiny stands. In which case, the standard on wall crossover will probably be appropriate or close. Also, if you put the speakers up on, say, 12” stands, you can gain back a little desk space for papers and such, which is always precious. You might want to think about that one.

It sounds as though you’re cooking up your own project. If, perchance, you’re making these speakers here, you could take the on desk measurements and I could throw them into the crossover model here and then suggest component changes for on desk.

Also, you mentioned the lower budget design were tough to run. I double checked the impedance profile and I don't think I saw anything to suggest they would dip below ~6 ohms...am I missing something when I read it? It just doesn't look like an amp rated for 4 ohm impedance should have an issue. Admittedly, I'm not very smart about such things so I look forward to being educated!

Right. I’m not sure what’s going on there. The deluxes played very loudly just fine. The basics shut down the receiver whenever I cranked the volume up, putting it in protection mode. Maybe something is going on that I don’t understand here. The impedance phase goes up more sharply from ~110-250 Hz with the basics than the deluxes, which take ~90-500 for the same swing. But they both have a sharp change below ~100 Hz, so I doubt it’s that. One possibility that comes to mind is that I could have some wires crossed or shorted somewhere, even though I checked things over. It really is a mess of wires at the moment. See post 228, above. When time comes to voice the on wall crossovers, I’ll probably rip everything apart and clip everything together again. Then I’ll see if the deluxes versus basics amp problem goes away. That said, this issue does not bother me too much in that I really prefer the sound of the deluxes to the basics. The difference is not subtle. So I’ll keep with the deluxes, anyways. The basics are just being put out there for others.

By the way, if a Padawan is to a Jedi, what would you be called when compared to the Dark Lord?

That would simply be a Sith apprentice...

Wow! Some day... some day... I can dream of being a Sith apprentice!

[---k---]

I'm a happy storm trooper.

[technodanvan]

Thanks for the long reply Jon! I might follow your high-end design for myself sometime, but I have my own project in mind for my wife. However, the reason I've been following your design is that it seems so very few projects have approached this form factor. Even fewer have used anything but relatively inexpensive drivers, and fewer still have done it without simply using a full-range driver and calling it a day! I think the placement of such small speakers makes them a unique challenge compared to their larger siblings.

This project will use a pair of Tang Band W4-1879 drivers and the SB Acoustics SB26 small frame tweeter. It should end up being smaller than yours! The driver is capable of a lot more than I'll get, but size is important to my wife (ha!) so less bass it'll be. I'm toying with some 3 1/2" Peerless PRs right now to see if they'd be of benefit...not quite sold on them yet, but they're cute. I'm hoping to do more small-to-tiny speakers in the future so if they aren't used now I'm sure I can use them later.

Of course all of this depends on whether the wife changes her mind again.

I wish I hadn't sold my other set of these...I really think they would have made a midrange midrange in a 3-way. Now that their price has doubled I'm not sure the value is there any longer.

Once I get DATS loaded back up and the drivers measured I'll reopen my old thread. Hoping to get that and a test box together in the coming days.

I'm a happy storm trooper.

I don't know about you, but sometimes I'm pretty happy just being told to "shut up and color"!

[JonW]

I'm a happy storm trooper.

You get the Little Storm Trooper Urban Achiever award.

[JonW]

Thanks for the long reply Jon! I might follow your high-end design for myself sometime, but I have my own project in mind for my wife. However, the reason I've been following your design is that it seems so very few projects have approached this form factor. Even fewer have used anything but relatively inexpensive drivers, and fewer still have done it without simply using a full-range driver and calling it a day! I think the placement of such small speakers makes them a unique challenge compared to their larger siblings.

Yup, there’s not much out there with a small form factor, low distortion/top shelf drivers, and bass to reach a sub at 80 Hz. Hence this project.

This project will use a pair of Tang Band W4-1879 drivers and the SB Acoustics SB26 small frame tweeter. It should end up being smaller than yours! The driver is capable of a lot more than I'll get, but size is important to my wife (ha!) so less bass it'll be. I'm toying with some 3 1/2" Peerless PRs right now to see if they'd be of benefit...not quite sold on them yet, but they're cute. I'm hoping to do more small-to-tiny speakers in the future so if they aren't used now I'm sure I can use them later.

Of course all of this depends on whether the wife changes her mind again.

I wish I hadn't sold my other set of these...I really think they would have made a midrange midrange in a 3-way. Now that their price has doubled I'm not sure the value is there any longer.

Once I get DATS loaded back up and the drivers measured I'll reopen my old thread. Hoping to get that and a test box together in the coming days.

That W4 looks interesting. Not cheap, at ~$148, it seems. I know that the cheaper W4-1337 is much loved. And pretty to look at. Although I don’t think that I’ve ever heard either flavor of the W4. Are you going with two W4’s per side? Why? I might think that a single, larger driver might get you more bass in a smaller cabinet? But I have not modeled the W4.

Small frame tweeters that are low distortion are not too plentiful. I’m not familiar with that SB26. It looks to be smaller than average. But still not as small as the Scanspeak tweeter here. But the SB26 is muuuuch cheaper.

I think that it was Jed that had a very small speaker project using the W4-1337 and the Vifa D26 small form tweeter (I think it was). That tweeter had very good looking measurements (low distortion) and was cheap. But is no longer produced, as far as I know. Right, I just found a thread on that project:

https://www.htguide.com/forum/showthread.php?27583-Lineup-Series-Build-Thread-(F4-D4-D44-R4-R44)

Looks like his D4 flavor is the smallest cabinet with both a tweeter and woofer. My project here is coming in a little smaller, for whatever that’s worth. Jeds’, however, would cost much less and be simpler cabinetry.

Best of luck with your project! :T

I was thinking a little bit more on your question of “on desk” versus “on wall.” In addition to the points mentioned earlier, I will further recommend trying to put the speakers up on little stands. Of course, you’ll avoid some of the desk bounce issues. And you might get a little desk space back, which will come in handy when doing real work. But maybe there is an even more practical matter: Get the speakers off the desk surface and you decrease the odds of a stray pen piercing a driver. Or a cup of spilled coffee shorting out a crossover. Just something to think about.

[JonW]

Oh, one other thing about those Lineups. I think that Ryan (-k-), our Little Storm Trooper Urban Achiever, made a pair. He might have some thoughts.

[5th element]

The real DIY stormtrooper mask!

[---k---]

Yep. I have a pair sitting of D4's on a shelf at the office. I listen to them daily. Unfortunately it is usually at very low volume, often FM radio, and sometimes AM radio.
But, I do think they sound great when allowed to. At one point I had them up at CJD's and we listened them against his MTM and a few other speakers. I think you Jon might have even been there. I seem to recall that the W4 was slightly clearer, more resolving and airy than the RS150. But the speakers were voiced similar and the differences could have been due to lots of things, including the tweeter I had wired out of phase. ops: I'm not sure at this point my memory can be trusted. But definitely a driver worthy of respect and it's price point.

[technodanvan]

YThat W4 looks interesting. Not cheap, at ~$148, it seems. I know that the cheaper W4-1337 is much loved. And pretty to look at. Although I don’t think that I’ve ever heard either flavor of the W4. Are you going with two W4’s per side?

"snip"

I was thinking a little bit more on your question of “on desk” versus “on wall.” In addition to the points mentioned earlier, I will further recommend trying to put the speakers up on little stands. Of course, you’ll avoid some of the desk bounce issues. And you might get a little desk space back, which will come in handy when doing real work. But maybe there is an even more practical matter: Get the speakers off the desk surface and you decrease the odds of a stray pen piercing a driver. Or a cup of spilled coffee shorting out a crossover. Just something to think about.

I apologize, I worded that poorly. The goal will be a single W4 per side. The 1879s are quite expensive now, however when I bought them they were $60 apiece. PE has since upped the price considerably! I had initially bought two pair but ended up selling a set before the price hike. Probably a good thing, but I still would have liked to have them around. I keep thinking about using a larger driver instead but I'm fighting that urge - in the end the boxes will end up being built very much like your own (facets and all!) in that the baffle dimensions will largely be determined by the size needed for the drivers.

Of course I got sidetracked on speaker stuff this week and started working on a tool cabinet instead. Once materials funds for that run out I'll switch back to the speakers until next paycheck.

I'm hoping that choosing a driver designed for use full range and a tweeter with moderately low frequency extension will make it easier to approach an LR2 crossover. If not possible with the SB26 I can always grab a set of SB29s as they dig a bit deeper. Given how the Dark Lord is so enamored with the SB26 I'm sure it wouldn't go unused!

Good call on the potential damage to speakers on a desk. If they were on my own it wouldn't be a problem (it's massive) and on my wife's...well she rarely has more than her laptop there. What I may do is purchase some foam to mimic something like in the link below, though in my case they either wouldn't be angled (or at least they would be angled 'up').

http://www.amazon.com/Auralex-Acoustics-Mopad-XL-MOPAD-XL/dp/B00266HBSU/ref=sr_1_3?s=musical-instruments&ie=UTF8&qid=1396621621&sr=1-3

Thoughts?

[JonW]

Hi technodanvan,

Sorry for the slow response- quite busy here. Sounds like you got those woofers *really* cheap compared to what they’re costing now. Yes, for projects like these, you need to always fight the urge to let the cabinets get bigger. Mission creep. For me, I said “flat to 80 Hz” for mating with a sub. And then the smallest possible from there.

I also tried hitting LR2 with this project. But LR4 just worked better, once I had the measurements in the cabinets. In my case, that’s probably due to the majority of the design work revolving around a metal coned driver (W16).

For your project, you have a tweeter and a woofer. In that case, I would think that there is no need for the woofer to be capable of running full range. But if you already have that woofer, there is probably no harm in making the woofer work less. It should turn out better, I’d think.

Sure, putting the speakers on a foam stand should be fine, I might imagine. Just an inch up will minimize damage from spilled drinks. With this arrangement, you’re back to the “on desk, on wall” situation. Not just “on wall.” So I’d take the measurements accordingly.

Check out this link for on wall placement:

http://www.genelec.com/learning-center/presentations-tutorials/placingloudspeakers/wallcancellation/

It can be an audio mess. Adding on desk will probably make it even worse. If you can get the speakers up higher off the desk, it might help. But if it’s only a foot or so up, I will guess that the nulls issue will still be present. So it might not help. Make sure that the speaker height from the desk and the depth from the wall are not the same. Otherwise, the nulls could get even worse.

The next step in my speaker project, here, is to voice them up against a wall, with the on wall crossover than I have already modeled. I’m curious to see how different it sounds versus in room. I’m also curious to see when I actually have time to do that.

[JonW]

I'm trying to get back to this project. It's time to voice the on wall crossovers. Due to various things going on in the house, I am relegated to the basement. So I made an on wall set up of sorts and clipped together the deluxe on wall crossover. Hopefully, I'll have a chance to listen a lot and get to a good crossover. I'm in the midst of a lot of work travel. Still jet lagged from one trip and have another overseas visit next week. But I'll see what I can get done while home. Here is a grainy phone photo of the current setup

[cjd]

Now those walls have some character! Looks like a nice quiet place to tinker.

C

[JonW]

Now those walls have some character! Looks like a nice quiet place to tinker.

C

Yeah, they kind of tie the room together. With a little kid running around the house and now the in-laws visiting for 6 weeks, this corner of the basement is all I can get. Regarding the in-laws visit, I am referring to it as "The Great German Invasion of 2014."

I got in a little voicing. Details to follow after I can make up and post all the relevant plots. A quick preview is that I have improved the basics crossover a little. Same general topology but some part values have changed. Details to come in the vaguely defined "near future."

[5th element]

I would imagine that all you'd need to do is reduce the amount of baffle step compensation and rebalance the tweeter level, but that can easier said than done sometimes depending on how the design comes together.

[JonW]

I would imagine that all you'd need to do is reduce the amount of baffle step compensation and rebalance the tweeter level, but that can easier said than done sometimes depending on how the design comes together.

Hi Matt,

Thanks. Yeah, that's kind of what I was thinking. At least, that's the general idea. But it does not look to be working out quite that way. I took a set of on wall measurements with a longer gate time, in order to see the nulls from bouncing back off the wall. I put those measurements into the regular, in room crossover and then changed things to make all the graphs look pretty. Removed some parts, added some in elsewhere, changed some values. Clipped it together and... it did not sound great at all. I went back to the in room crossover and it sounded great again even when placed on the wall. Then I looked at all the plots with the in room crossover and the on wall measurements... and they actually looked quite good, in retrospect. So the in room and on wall crossovers may turn out to be exactly the same. My guess is the tiny cabinets here are why. They both sound great and all the plots look very good. (OK, except the wall bounce nulls when on wall.)

I still want to listen to them a bit more. And then I'll make up all the plots to post them here. I guess you learn something new with each project. Go figure.

[5th element]

What you must not do, and it does seem somewhat counter intuitive, is actually measure with a long gate time. I know you want to see how the room reflections are specifically affecting the sound but there is nothing you can do about it. Nulls, such as floor bounce/wall bounce, well anything bounce, will show up, you cannot however correct for these.

The most important part of the loudspeaker design is the direct sound. Yes, the reflected sound also matters a lot when it comes to recreating spaciousness and in the way that the reflections will impinge on the way the loudspeakers image, but you cannot affect these in any way. The only things you can do here would be room treatment, repositioning the loudpeakers or changing them into a horn type design. Here the loudspeaker design and loudspeaker positions are fixed, some room treatment could be added, but this is besides the point.

You must design the loudspeakers with the usual gating in place. If you open up the gate you will let in reflections that you do not want and this will compromise the accuracy of the design with regards to the direct sound. Yes you're going to get more serious reflections with the on wall set up, but you must not compromise the on axis/direct sound, to try and correct for these. We, as humans, have been listening to sounds produced in rooms for a very long time, we've even evolved that way, so we are pretty used to processing sounds produced within them. We don't really hear comb filtering from reflections for example, it looks terrible in measurements, but sounds nowhere near as bad as you would expect it to. If the loudspeakers direct sound actually measured as bad as the comb filtered response did though it would sound awful.

The only areas that you really want to adjust the loudspeaker, vs what the rooms doing, is with the amount of baffle step compensation you're using and how much you might tilt the treble up or down to get the top end balance right.

You can use EQ with subs to complement the mains in an effect to reduce the effects of room modes, but those are an entirely different thing.

[JonW]

Hi Matt,

Very interesting. At first, yes, it is counterintuitive. But when you think about it briefly, everything that you said makes perfect sense. Your explanation is very clear. Thanks!

It so happens that I took measurements with both short/typical and also long gate times. I’m glad that I have the short gate measurements to look at them in this context. But I also did want to see the effects of the wall bounce and such, hence the longer gate time measurements. It’s the first time I’ve considered an on wall situation, so actually getting to see the effects with the longer gate times was quite useful for me, even if it does not influence the final speaker design.

OK. so here is what I did… I put the short gate time, on wall measurements into the in room crossovers. The plots still look very good. The treble is a little high, but not by much. Adding a mere 1 Ohm to the tweeter padding resistor brought things back to a flat frequency response. Then I listened to the speakers, both on wall and in room, both with the added 1 Ohm and without. The result? Everything sounds very good to me. There are differences but we are in the subtle changes area now. It kind of depends on what song is playing, some sounding better with the added resistance or different placement, some not. I don’t think that you can go wrong any way here. It’s up to personal preference. I kind of prefer the sound without the added resistance. But I am also listening for shorter periods of time. If, say, I had music playing for a few hours, I might want the tweeter backed off just that touch.

And this brings me to what I think is a good plan. I have two pairs of binding posts in the cabinets. (I put them in to hook up the tweeter and woofer separately for measurements and voicing.) So I think that I will have the regular crossover hooked up to one pair of binding posts. To the second pair of binding posts, I will just add 1 Ohm of resistance for the tweeter, but otherwise then run it through the rest of the regular crossover. That way, I have a speaker that can be switched between on wall and in room configurations, depending upon which binding post pair the cables are connected too. What this also allows me to do is raise or lower the treble level a touch in the future, should I decide that I prefer it one way or another. I like this plan.

I will post all the relevant plots some day, after I make them all pretty, to show what I am talking about. Thanks again for the help.

[JonW]

I just ordered all the parts to make up four of the deluxe crossovers. Amusingly enough, everything was in stock except for one single inductor, backordered for a few weeks. Not inductors of a single value, but one individual inductor. PE had 3 of the 4 I need in stock.

No matter. I have plenty of other things to keep me busy for this project- make cables, finish the accompanying sub, make up all the crossover plots to post here, etc.

That and I have a lot of work travel that will keep me away from the speakers. In the next 4 weeks, I will be in Ireland, California, and Washington DC. Ireland is tomorrow, just as I got adjusted back to an Indiana sleep schedule after Singapore. You know it’s a bit too much travel when you arrive at the airport to check in and the agent says “You look very familiar. You come here a lot, don’t you?”

[JonMarsh]

You know it’s a bit too much travel when you arrive at the airport to check in and the agent says “You look very familiar. You come here a lot, don’t you?”

That's priceless! :rofl:

I can empathize from past years, but this last 12 months it's just 50 miles and back to Cupertino! I could use a little variety myself!

Amusingly enough, everything was in stock except for one single inductor, backordered for a few weeks. Not inductors of a single value, but one individual inductor. PE had 3 of the 4 I need in stock.

That's fairly typical in my experience with PE, but then, you could get at least 3 crossovers built and checked out? Though not with your travel plans!

We're heading out a week and a half from now for GF's niece's wedding in Mt Rainier area- it's just dawned on me that there's a heck of a lot of stuff to get together for that trip- I just fixed the window regulator on the Element, but we've got a lot of gear check out and packing to do, and some more vehicle prep before we leave on the 30th. Going to camp at a few spots after the wedding on the way back. It's been a weirdly busy summer for being mostly tied up at work- the only reason I'm getting the time off is that the new spin on the silicon won't be ready to test until the 2nd week of August, leaving things a bit more slack the last week in July and first week in August.

Good luck making progress on your build- I'm setting up now after lunch for Ardent measurements!

[JonW]

Jon-

Have fun with the Ardent measurements. I'm already thinking about my next speaker project and the measurements. But I need to first finish this current project. And given how long this one has taken, I may not want to start the next one for a while. At least until my son goes off to college in about 13 years.

Enjoy Washington state. It's gorgeous up there. If you have time and are looking for somewhere to go, try Olympic National Park. Mt. Rainier is very, very nice. Olympic, however, is on another level, IMHO. Rialto Beach is quite a sight. Regardless of where you go, have fun!

[---k---]

Jon,
I'm dissapointed. I would think your 5-year old would be capable of taking the measurements for you. Another year or two and he can be building the boxes.

[cjd]

What kind of SPL change is that 1ohm? I found tweaking padding changed the crossover no matter what I did in the Nebbiolo, which wasn't expected. I assume it's the lower order slopes. Do you have the data to identify how bad wall bounce is for you? I've found on-wall oddly frustrating.

Regardless, enjoy the slow builds if you can - I don't know if you're more about process or product, but especially with such manageable size... a hobby is a hobby, meant to be there when you need a break and not an imposition. IMO.

I packed/stacked all the raw drivers I have yesterday evening - the 4xMaelstrom 18's is only half! I think I'll be working on a few still. But I think I'll actually have a room for the build that uses those M18's... Still, a depressingly large stash.

The Olympics have to be one of my favorite places anywhere. The drive up to Hurricane ridge can be daunting, especially in fog (like we hit last time) but get above and despite the tourists and the usual groups that don't belong (hiking up to the peak in flip flops?! and your feet hurt? gee...) Plus there's the variety in the rainforests. I haven't even been to them all. And La Push can be an amazing view as the sun sets on the right day, if you can get there...

If you like flying kites, hit the coast as you approach Oregon and all the way through. Last trip out there I packed my flexi (2-line airfoil with a leading edge spar - old school, but still the world record holder for speed, and oh so much fun! plus I could fit it in my bag...)

I did the math, I've spent over 250 days worth of my life commuting to my current job from my current house... that's over almost 8 years, but is somewhat depressing. I haven't put half that many days into vacation over the same 8 years, and I work for a travel company!

Carry on... enough tangents from me!

C

[JonW]

Hi Chris,

Changing the padding resistor by 1 Ohm has little effect on the crossover. Other than, of course, changing the tweeter level. It’s not a perfectly flat influence, but it’s not much more than that. That’s the main reason I like having padding resistors there, early in the tweeter crossover. They make for very easy changes to the level of highs. I wonder why you are seeing otherwise. My slopes hit LR4 acoustic, so maybe that is part of it. I’ll post all the plots, to show what I’m talking about, when I get back home. Including the wall bounce data, which I think are cool to look at.

This build has been so slow due to things with my busy job (7 days/week), having a kid, and some other family things. I’d prefer speaker fun and progress to be more rapid, but, you know, there is the rest of life. Speaking of which... 250 days of commuting? Ugh! More than anything else in life, time is the most precious. It’s not easy to always keep that in mind, but we need to try.

I agree that the Olympics are one of the prettiest spots that I’ve ever seen. Especially Rialto Beach. I just landed in Ireland earlier today. It’s a work trip and I’ll only be here for 50 hours between flights (eek!) but will try to see a few things. As the plane was descending, the landscape looked... exactly as I had imagined a typical Irish countryside that probably only exists in my head. Nope, it was right there. It’s real.

[JonW]

Jon,
I'm dissapointed. I would think your 5-year old would be capable of taking the measurements for you. Another year or two and he can be building the boxes.

Hi Ryan,

Yeah, I should put the little guy to work. I don’t think that he has the manual dexterity to build the cabinets at this stage. But I do think that he has the intellectual heft to make a better crossover than I can. The other night, I was struggling to get him into bed and asleep, which is always the case. And our exchange went:
Dad: “OK, buddy, you really need to get into bed now. *Everybody* else is already asleep.”
Kid: “Are the owls asleep?”
Dad: “Yes, even the owls are asleep.”
Kid: “But owls are nocturnal.”

[---k---]

That is funny. I'm expecting much of the same here in the near future.

[JonW]

That is funny. I'm expecting much of the same here in the near future.

You will be exhausted, broke, and have no time for building speakers. And yet it's totally worth it. Sappy, I know. But that's my experience. And so I wish you very good luck!

[JonW]

Deluxe on wall crossover finalized

Deluxe on wall crossover

OK, it’s time to post all the crossover plots. No smoothing has been applied to any graph. Here we go…

The following plots are for the deluxe speakers, in room versus on wall.

Here is the frequency response for the deluxe speakers, in room crossover and in room measurements. This is from the final crossover, as posted earlier and our starting point for this post:



Then we have the on wall measurements placed into the in room crossover. It looks pretty good, but the top end is a tiny bit high so we will deal with that in a moment:



On wall measurements taken long enough to see the effects of the wall bounce cancellations. Placed into the in room crossover:



So how so we adapt the crossover to these on wall measurements? Well, the first thing that I did was make a pretty much new crossover to get the plots looking as close to perfect as I could. Then I clipped it together and took a listen. And it was not too good. It was fine, but it did not sound as good as the final in room crossover played in room. Then I just remade the in room crossover and had a listen with the on wall positioning of the speaker. Guess what? It sounded really good!

I added one Ohm to the tweeter padding resistor to deal with that rising top end and it sounded a tiny bit better. Or not. I went back and forth with that minor change. It sounds great either way and kind of depends on the song that you’re listening to as to whether or not we need that little bit of extra tweeter padding. I was listening for short times. Let’s go with the extra padding to prevent listener fatigue after long periods of the speakers playing.

Here is the final crossover for the deluxe speakers with on wall placement:


Note that the only difference from the in room crossover is that the tweeter padding resistor, R1, has changed from 7 to 8 Ohms.

Also note that resistors R2, R4, R8, and R9 are not physical parts of the crossover. I just added them into the circuit diagram to account for the DC resistance of the inductors.

So we will have a super easy way to switch the crossover between in room and on wall. It’s just a resistor change. We could put a switch on the back of the speakers to have it run through that extra 1 Ohm or not. Alternatively, we could have two pairs of binding posts, one with the extra 1 Ohm and the other without. Both run through the complete crossover. For different placements, just change which pair of binding posts you hook your speaker cables up to. I have two bonding post pairs so that’s what I will do. In the future, I will make up diagrams showing these options, to make it easy for potential builders.


Final frequency response of the on wall measurements in the on wall crossover. Almost +/-2 dB:



The on wall measurements taken long enough to show the effects of wall bounce, placed in the on wall crossover:



Phase of the on wall crossover with the on wall measurements:



Impedance of the on wall deluxe crossover:



Reverse tweeter null of the on wall measurements in the on wall crossover:



30 degrees off axis FR of the on wall measurements in the on wall crossover. Real 30 degree data were used (not a simulation):