The Schlafmutzes Mini MT’s design and progress

[JonW]

The rubber boot, is a K says, most likely from the car audio sector, where they are used to prevent the magnets from possibly getting bashed and chipping. In home audio they arent needed but could possibly dampen some very benign resonances in the basket/motor structure. In this case it will most likely be restricting the air flow around the driver. I mean I know the boot is probably 5mm at max, but if it's almost pressed into the cabinet edge, I'd be happy to take those 5mm!

So I had a look and was able to pull off the rubber boot. And put it back on. So playing here doesn’t cause any irreversible damage. I tried measuring the boot thickness and it’s difficult due to the presence of a lip and I lent out my calipers so they are not around. It looks to be about 2 mm thick. So it’s not going to save much space but it’s easy to take off and put back on should there be a need.

From my point of view FR measurements are the only way you will know that what you're doing is being in any way effective. If the size of the dip in the FR is proportional to the size of the impedance blip then you may have reduced its effect by a couple of dB.

Agreed.

One thing I will point out here is that stuffing usually results in an increase of effective box volume and should lower the system resonance. Your stuffing has apparently increased the resonance slightly and reduced the magnitude of the impedance peak, which I think would help support the theory that you are restricting the actual air flow from the rear of the driver into the effective cabinet volume.

Yes, I noticed that, too. Fix one problem, create another. I guess more measurements will tell us which way is the best to go.

This should be less of an issue with the ER15, but I think it helps point out that if you can remove the boot then it'd be a very good idea. I think it shows that you do need to balance the amount of stuffing that you are using so close to the opening of the cone at the rear of the driver. I mean its necessary in reducing the reflections (how effective this is in the FR domain we don't know yet), but at the same time, if you over do it you start affecting other things. You could perhaps remove the boot and then surround the magnet, inculding the top plate, with felt and this might help with the reflections.

Right. The ER15 has a much more open rear side. Although the ER15 speaker is more of a cheap option for other people. I’ll put more time into optimizing the W16 version.

Personally, I would opt for decreasing the resonances (if they do affect the FR) and accept the down sides of a bit of over stuffing. It might be worth running a nearfield FR measurement with the stuffing in place and compare it with the version without the stuffing, just so you can see how much it actually affects the box alignment. I will point out though that restricting the air flow round the back of the driver, via resistive means, is not a bad thing as per say (like the resonance are) it's just a different way of affecting the driver+box tuning, just like stuffing different densities of open cell foam into ports can be a very effective way at reducing their output.

I understand how the stuffing influences the location/frequency and the height of the big impedance peak. But does the blob of increased resistance at the lower frequencies (e.g., 30-60 Hz) bother you or signal anything too bad?

At this point, I think that I have a straightforward plan:

-Remove the rubber boot from a W16 woofer

-Add felt around the top side of the magnet, at least as much as I can fit on there.

-Keep the Whispermat in the cabinet

-Retake frequency and impedance measurements for the “revised” speaker in which the boot is removed and the Whispermat is present.

-Compare these new measurements with the old measurements for which the rubber boot is present and the Whispermat is absent.

-Pick whichever set of measurements look better and go to the crossover design from there.

-I will also retake measurements of the ER15 speaker with added felt on the magnet and Whispermat at the back of the cabinet.





For perspective, here are those photos. You can see how little space there is between the driver hole and the inside cabinet edge. Almost no clearance, just a few mm’s.


This photo is dark but you can see how wide the W16 driver is relative to the flange and such. A piece of Whispermat is also placed there showing that you can’t get the driver into the cabinet with anything added to the magnet.


I tried squeezing this arrangement in but no go.


Here you can see the Whispermat thickness relative to the cabinet and an ER15.


Here is a photo showing the Whispermat at the back of the cabinet.

[5th element]

I understand how the stuffing influences the location/frequency and the height of the big impedance peak. But does the blob of increased resistance at the lower frequencies (e.g., 30-60 Hz) bother you or signal anything too bad?

Are you talking about the blip at 60Hz in these measurements?

Image not available

If so I would expect them to be interference from the mains, that is if you're running 60Hz there in Indiana.

-Add felt around the top side of the magnet, at least as much as I can fit on there.

I would wrap the magnet in felt actually, or at least make sure that there is a strip of felt around the circumference of the ferrite.

-Keep the Whispermat in the cabinet

This looks like good stuff so I would consider it well worth keeping. Even though the magnet gets in the way of actually having whisper right up to the baffle, I would consider placing shorter strips of the stuff on the sides and bottom of the cabinet that meet up to the magnet. Although things do look rather tight just with the layer on the back.

-Retake frequency and impedance measurements for the “revised” speaker in which the boot is removed and the Whispermat is present.

I went and reread one of your previous posts and I didn't realise that the impedance 'improvements' were brought about solely by placing whisper mat on the back of the cabinet. That is indeed interesting as I wouldn't have expected it to affect the tuning of the cabinet in the way that it did. Perhaps the drive level was different as the T/S parameters do shift under different excursion levels. Either way, if you got that much of an improvement simply by adding the whisper to the back, I don't see why it shouldn't be possible to get things even better. You could try simply doubling up the thickness of the whisper on the back, or if the material is sealed on one side, as it's designed to affix directly to the wall, then perhaps a few layers of standard wadding on top of the whisper.

-Pick whichever set of measurements look better and go to the crossover design from there.

There is always the danger that even more measurements with trial and error might be necessary, because this is after all an iterative process.

For perspective, here are those photos. You can see how little space there is between the driver hole and the inside cabinet edge. Almost no clearance, just a few mm’s.

I think what I'd be looking at is rubber boot off, felt on the top plate and around the magnet, whisper on the back, as much else as you could put on the inside walls and then a bit of standard wadding in the middle, that's if it will all fit. If you over stuff you do start losing effective cabinet volume, but I cannot see that happening with only whisper affixed to as many side walls as is possible. Obviously this will require some experimentation as the whisper on the side walls might not be as effective as simply more on the back wall.

This build certainly makes you appreciate neo magnets on small drivers in small cabinets. I knew constrained internals affected things, but not quite as much as this.

[JonW]

I’m back from some travel. Normally work has me away about once a month but it’s been 3 times this past month. The travel should calm down some now and I might be around to work on the speakers, I hope.

I understand how the stuffing influences the location/frequency and the height of the big impedance peak. But does the blob of increased resistance at the lower frequencies (e.g., 30-60 Hz) bother you or signal anything too bad?

Are you talking about the blip at 60Hz in these measurements?

Here is the extra impedance I’m a little concerned about:



My graphing program and WooferTester don’t agree well so the plot is a little funny, but look at the extra resistance at the frequencies below the main peak. Should I be concerned about that? It came about from the stuffing placed at the rear.

I would wrap the magnet in felt actually, or at least make sure that there is a strip of felt around the circumference of the ferrite.

That’s the plan. Even with the boot removed, however, there will not be much space for felt to be wrapped around. I’ll put on as much as I can and just go with that.

This looks like good stuff so I would consider it well worth keeping. Even though the magnet gets in the way of actually having whisper right up to the baffle, I would consider placing shorter strips of the stuff on the sides and bottom of the cabinet that meet up to the magnet. Although things do look rather tight just with the layer on the back.

Yeah, it’s very tight in there. The front-to-rear inside of the cabinet is ~6.75”. Subtract 1.25” for the Whispermat and you’re left with 5.5”. Then subtract ~1.9” for the magnet and there is really very little space left in there. We can’t forget that there will be a crossover some day. So I’ve got to leave space for that. If I add the short strips on the side and bottom, all the real estate is pretty much covered. No place for the crossover.

I went and reread one of your previous posts and I didn't realise that the impedance 'improvements' were brought about solely by placing whisper mat on the back of the cabinet. That is indeed interesting as I wouldn't have expected it to affect the tuning of the cabinet in the way that it did. Perhaps the drive level was different as the T/S parameters do shift under different excursion levels. Either way, if you got that much of an improvement simply by adding the whisper to the back, I don't see why it shouldn't be possible to get things even better. You could try simply doubling up the thickness of the whisper on the back, or if the material is sealed on one side, as it's designed to affix directly to the wall, then perhaps a few layers of standard wadding on top of the whisper.

Right, in theory, those things should help. But I’m not willing to have an external crossover. If I double up the Whispermat at the back, I’d have less than 6 cm from the back of the magnet to the Whispermat! It’s tight in there, I’m tellin’ ya.

There is always the danger that even more measurements with trial and error might be necessary, because this is after all an iterative process.

I guess. I’m thinking now to put in as much Whispermat and felt as I can, while still leaving space for crossovers, take the measurements, and call it as good as it will get.

I think what I'd be looking at is rubber boot off, felt on the top plate and around the magnet, whisper on the back, as much else as you could put on the inside walls and then a bit of standard wadding in the middle, that's if it will all fit. If you over stuff you do start losing effective cabinet volume, but I cannot see that happening with only whisper affixed to as many side walls as is possible. Obviously this will require some experimentation as the whisper on the side walls might not be as effective as simply more on the back wall.

Right. Although I just don’t think that there is space for more on the side wall. The magnet almost touches the side wall. Less than 1 cm from the magnet to the side wall. I doubt that there is enough space on the bottom for the whole crossover. So some of the side wall needs to be reserved for that.

This build certainly makes you appreciate neo magnets on small drivers in small cabinets. I knew constrained internals affected things, but not quite as much as this.

Yup. These speakers are really small. That’s creating some minor troubles. But I think that it also means that the speakers will be somewhat unique. And a fun design challenge.

[JonW]

Here is the latest…

I tried wool and foam to see if the measurements would improve. Here is what I did:

The “Whispermat” that I have been using is, it turns out, the 1.25” thick Sonic Barrier from Parts Express. Might as well refer to it as what it really is from here on out. Here is a piece cut to fit at the back of the cabinet. With space removed for the binding posts and wires.



Here it is installed. Although I did not use the adhesive backing because I did not know if all of this would help or not.




Next up was to remove the rubber boot around the W16 woofer magnet.



It’s naked now. (Gentlemen, avert your eyes!)





Then it was time to add the felt. I’m using 0.25” thick, 2” wide strips of “Extra Soft Grey F13” felt from McMaster-Carr. Here is one piece placed on the bottom of the cabinet. It was OK to put it here because the magnet will hang over this part and we can’t put a crossover there anyways.



Another view.



Then I tried to wrap the whole magnet with a ring of felt but it was just a hair too tight to fit into the cabinet hole. So I put felt on the top half of the magnet. (The wires and leads will be toward the bottom of the cabinet.) Then I covered the back of the magnet with felt. It’s just held on with tape for now because I did not yet know if I will want to use the felt or not. If the felt stays, I’m not sure what kind of glue I’ll use. Maybe hot glue like will be used for the crossover? We’ll worry about that later.

[JonW]

OK, so did any of this help? Yeah, a tiny bit, I think. Here’s what we’ve got:

W16 woofer with (blue) and without (red) the wool and Sonic Barrier. Note that the ~700 Hz dip is present in both, but a little bit less pronounced with the wool and Sonic Barrier.




Here is the ER15 woofer with (yellow) and without (green) the felt and Sonic Barrier. Kind of the same story as with the W16. The dip at ~700 Hz is still there but a little less when stuffed.




The tweeter with (green) and without (blue) the felt and Sonic Barrier. Very similar. I’m trying to find a difference and if I squint enough, maybe there is a small dip just over ~1,000 Hz that is less with the felt. And maybe the magnitude of the trough at ~5,000 Hz to the peak at ~2,000 Hz is a little less with the felt.




Right, so where does this all leave us? I’d say that things are a tiny bit better, all around, with the felt and Sonic Barrier. Thanks to 5th element for the ideas. We're talking about a maximum of 1 db improvement in very narrow frequency ranges. It's looking a little better. But it's subtle.

We’ve got as much felt and Sonic Barrier as we can now fit in there, while still leaving a little space for the crossovers. Can’t really add in anything additional. These are cheap items that are pretty easy to install. So I’ll keep with the felt and Sonic Barrier. And now I’ve got good measurements to work with.

Next up is crossover design. Just for a quick check, I dropped these measurements into a slightly modified LR4 crossover that I posted earlier. Looks good. But I think that it can be better. Crossover design is where I go next.

[JonMarsh]

Progress!! :T

[JonW]

Progress!! :T

Yeah, it's progress, but shhhh, don't tell anyone.

By the way Jon, do you have any tips for how to merge the woofer near field measurements with the far field measurements? Maybe in Fuzzmeasure or in LspCAD? Thanks.

[5th element]

Progress indeed! Now for the fun stuff, xovers

[JonW]

Last night I got to spend a bunch of time working on a crossover. I’m starting with the deluxe (W16) speakers in room. The idea is to use this one as the model speaker. Get that crossover looking good. Then adapt it for the on wall and the basic (ER15) in room and on wall variants.

So far, it looks pretty good. Hitting a near ideal LR4. But I don’t want to post it yet. First, I want to look at it again a few times, with fresh eyes. Tweak it a little here and there. Make it as nice as I can. Then change all the component values to what can actually be purchased. Then I’ll show what I’ve got. Maybe even late tonight. And see if anyone has any thoughts.

Crossover design is kind of fun. Trying to get the frequency response, impedance, and phase all looking nice. Make one look nice but the others get worse. It’s kind of like a puzzle. Definitely more fun than taking measurements.

[JonW]

For your consideration… a crossover. I’d like to see if anyone has any thoughts on this before I move on to making up the other derivatives of the speaker. This one here is the deluxe (W16) speaker out in the middle of the room.


Here is the frequency response. Pretty nice and flat. No smoothing has been used at all. The slopes are right on LR4’s. The crossover point is around 1750 Hz. With the metal cone breakup, I thought that it was best to cross on the low side. And this tweeter can handle it.





The phase. I’ve seen better and I’ve seen a lot worse. The two drivers are running somewhat close to each other in the crossover region and then down by 30 db (e.g., 700-3500 Hz).





Impedance. It’s kind of flat down to ~150 Hz. Again, I’ve seen better and I’ve seen worse. There is an LCR filter on the tweeter to squash the tweeter impedance peak at ~800 Hz. Another LCR to eliminate the woofer peak at ~78 Hz became impractical (see explanation below).





Reverse tweeter null. Looks pretty good.

[JonW]

The circuit.



Let’s walk through the circuit. It’s not too complicated at all.

R2, R4, R8, and R9 are “dummy” resistors in place to account for the DC resistance of the inductors. These are not physical parts of the crossover.

Both the tweeter and woofer are third-order electrical and nearly ideal LR fourth-order acoustic.

R1 is a tweeter padding resistor. I like those a lot because they will make for an easy spot to try different resistor values when voicing the speaker. Dialing in more or less high end is easy and really alters the character of the speaker.

L1, C1, and R3 make up an LCR filter to flatten the ~800 Hz peak in the impedance plot. I think that most people do not use these just to flatten impedance peaks. But I found that doing so helped the sound of my last project so I have one here. Beyond that, here having this LCR actually helped make the frequency response and the phase work out to be just that final bit better.

I tried adding an LCR filter to flatten out that woofer impedance peak at ~78 Hz. Turned out to be impractical in that the inductor wanted to be 15+ mH and the cap was ~250 uF. We are space limited in these cabinets. And these speakers will probably be used with a sub by everyone (well, with me, at least). So speaker output at ~80 Hz will be reduced anyways. Let’s skip that second LCR filter. But keep the one that is shown in the circuit.

C5 is a cap run in parallel to the inductor. It provides a narrow notch to help kill the woofer metal cone breakup at ~5,000 Hz.

Dealing with that metal cone breakup is important to do. We do not want to hear it. And, possibly of equal or greater importance, is that we do not want to hear the distortion products from it. So I did three things to address it here. First, there are the steep LR4 slopes to stop using the woofer quickly. Second, the series cap notch filter. Third, the low ~1750 crossover point. Between those three things, I think that it should be handled OK.

For the on wall variant, I took a quick look and dropped the on wall measurements into this in room crossover. It’s not perfect and we may need to make some adjustments for placing the speaker right up against a wall. But it’s pretty darned close to what you can see here, above. Things did not look bad at all. So the crossover here could also work when the speaker is placed up to a wall. Although I will try to modify the crossover to be a little better for that situation.

Just for kicks, I looked to see and a pair of these crossovers would be about $140. So we’re in the neighborhood of $900 for drivers and crossovers for a pair of the deluxe speakers. Maybe the basics will come in around $525 or so. Many details there still to be worked out.

Anyways, that’s the first crossover. If anyone has any thoughts, I’d love to hear them. Thanks!

[---k---]

Pretty graphs!

From a nondesigners perspective, it looks good. Nice write up explaining everything. I think the prices are reasonable for what you're packing. I'm guessing the price could come down with less expensive parts, such as Erseaudio.com caps?

You mention you like the resistor in front of the tweeter for adjustablility - makes sense. But I'm guessing you also need it to make the driver sensitivities match. I'll mention that CJD dislikes resistors in that location because of what he thinks he hears (I shouldn't speak like I'm an expert in all things CJD). I think I may have heard it too once when we were experimenting. None of the Khan crossovers he did for me have them,because the drivers just worked out. But, if you have to have one, not much you can do about it. I guess this might be where those expensive resistors come into play.

In designing the crossover, did you just go for flat or did you include a little voicing that you learned in your last project? I seem to remember you liking a little bbc dip?

And did you mention a reason for all the different varieties you're designing? Is it because you can? Or do you have specific locations each design is going to live in? I know some are going to your work office. Where else?

Now go order some parts!


Sent from my DROID2 GLOBAL using Tapatalk 2

[cjd]

It's more serious L-pads that I've been less fond of, though I don't recall all the versions over time. Occasionally, the circuit only works well with the resistor in front of the network - it DOES shape things when placed after differently. In this case, that might not be a bad thing - it might help with that rising response on the tweeter.

Perhaps a split, with most of the impedance after the network, but some up front to allow wiggle room?

It looks solid either way. Why are the impedances so high on the inductors? Having to use 20ga?

C

[JonW]

Pretty graphs!

From a nondesigners perspective, it looks good. Nice write up explaining everything. I think the prices are reasonable for what you're packing. I'm guessing the price could come down with less expensive parts, such as Erseaudio.com caps?

Thanks! I did those quick cost estimates using Solen caps, Mills resistors, and 20 gauge air core inductors. Certainly, some money could be saved from there. I’m thinking that I’ll suggest more expensive parts for the deluxe speakers and cheaper parts for the basic speakers. We’ll see.

You mention you like the resistor in front of the tweeter for adjustablility - makes sense. But I'm guessing you also need it to make the driver sensitivities match. I'll mention that CJD dislikes resistors in that location because of what he thinks he hears (I shouldn't speak like I'm an expert in all things CJD). I think I may have heard it too once when we were experimenting. None of the Khan crossovers he did for me have them,because the drivers just worked out. But, if you have to have one, not much you can do about it. I guess this might be where those expensive resistors come into play.

Right. This driver pair needs a big resistor there to bring the tweeter down to the level of the woofer. So I put it where you can also adjust it to taste. Chris has some good ideas about it, which I will reply to in a moment.

“Expensive” for a Mills resistor is something like $4.25 whereas a “cheap” Dayton resistor is something like $1.50. With only 2 resistors in the crossover, the difference between cheap and expensive is, for a pair of speakers, maybe $11.

In designing the crossover, did you just go for flat or did you include a little voicing that you learned in your last project? I seem to remember you liking a little bbc dip?

Funny that you should ask. Because as I was working up the crossover, I was trying to decide what I like best and I have not yet settled on a preference. My favorite commercial speakers (Totem Forests) have a BBC dip. But my last DIY project ended up being pretty flat. Maybe even a teeny downward slope of 1-2 db in the 10,000 to 20,000 Hz range. So I was trying to figure out what to do here. I went for flat. The tweeter has a rise in the raw measurements that is hard to get rid of. (See the third image in post #185, above). In the end, I let it be what it wanted to be. Then I can tweak that padding resistor by ear and see where we end up. But Chris has an interesting idea of moving the resistor around by splitting it up. I’d be happy if that helps get rid of the raw rise in the tweeter.

And did you mention a reason for all the different varieties you're designing? Is it because you can? Or do you have specific locations each design is going to live in? I know some are going to your work office. Where else?

The speakers are going in my office, up against a wall. So I need a deluxe, on wall version. Might as well also make a more typical version in the middle of a room. In case I ever move the speakers. That’s the deluxe, in room. Then the expense of a pair of these speakers will probably keep most people away (~$900). In order to give back to this great community from which I have learned so much, I wanted to make a cheaper version, too. Hence the basics (~$500). Might as well make in room and on wall versions there, too. So we’re at 4 flavors of the speakers.

Now go order some parts!

Not yet! I want to work up some of the other crossovers first. Then order parts for them all. It will be cheaper to place one, big order with free shipping rather than make a whole bunch of little orders.

[JonW]

Hi Chris,

Thanks for the help!

It's more serious L-pads that I've been less fond of, though I don't recall all the versions over time. Occasionally, the circuit only works well with the resistor in front of the network - it DOES shape things when placed after differently. In this case, that might not be a bad thing - it might help with that rising response on the tweeter.

Perhaps a split, with most of the impedance after the network, but some up front to allow wiggle room?

Interesting. So if I place the resistor in different spots, I might be able to decrease the rising tweeter response? That would be super. I’ll definitely give it a shot. I’ll keep some resistance where it is now, say 3 Ohms to allow for adjusting to taste by ear. Then move the rest to other spots. It looks like there are 4 options between where it is and before hitting the tweeter (after the LCR, after the first cap, after the inductor, after the second cap). I’ll try them all.

It looks solid either way. Why are the impedances so high on the inductors? Having to use 20ga?

Thanks. It’s good to know that I’m on the right track and that I can keep going.

The cabinets are really small. Real estate will be very limited for where to mount the crossover inside. No idea where it will all fit. In another thread, BOBinGA pointed out that higher gauge inductors tend to be narrower and taller than their lower gauge counterparts, which are lower and broader. So this project seems to really call for 20 gauge inductors.

[JonW]

I just mucked around with the resistor placement a little. Kept a 3 Ohm padding resistor where R1 is in the scheme above. Then put another resistor in various places within the tweeter circuit. Such as after everything else but before the tweeter. Eh, some good, some bad. It’s not a big change by any means. After alterning various other component values I can get the frequency output to look flatter. Impressively flat, actually. But the crossover point drifts up from ~1800 to ~2100 Hz. Which is not a good idea with this woofer. We’d get a lot more distortion. Plus the impedance takes a dive down to 3 Ohms, swings wildly, and the phase is off a little. I could probably fix the phase. Probably not with the impedance.

Anyways, if I keep the crossover as shown above and increase the value of that 10 Ohm padding resistor, it drops the tweeter output. So the crossover can be anything from a slightly rising output in the higher frequencies (shown above) to a drooping output as the frequencies go up. At 12 Ohms, there is a mild BBC dip. At 15 Ohms the tweeter, overall, is running low compared to the woofer. You get a bit of the downward slope that Jon Marsh says he likes (but I’ve never heard). It’s an easy change to make and the only way to know what really is best is to try all versions and use the ears. We’ll see.

So the figures below are the original crossover topology as shown above but with the 12 Ohm padding resistor. And then some tweaked values for the other components. The frequency response is flatter than before, the phase is the same, the impedance is OK. I think that this version is a small improvement.

[Saurav]

Another option to tilt the top end down would be to split up that 12 ohm resistor, say into 10 and 2. Put then 10 where the 12 is currently, move the 2 to be just before the tweeter in series. Then put an inductor in parallel with that 2 ohm resistor. You'll have to play with the inductor value, but you want something that provides around 2 ohms impedance around 5kHz or so... something like 0.05mH. What this does is, as the frequency goes up, the combined impedance of the resistor+inductor goes up, and that reduces the signal going to the tweeter. So you actually get to tilt the tweeter response down, instead of just dropping the overall level (which may or may not be what you'd like to do).

Just a suggestion. I've seen many other ways of achieving this, but the inductor thing has worked well for me in the past, without messing up the rest of the network too much.

[cjd]

I know you'll love hearing this but - off-axis measurements + simulations may also help you determine what you care about, or if anything needs to change - even on that tweeter response. It's very probably flat even at 15° off axis. I'm in agreement, that change may not be worth it.

The slight bump at 1kHz might be tweakable if you had a zobel in the mix. I'd probably also be fiddling to see if the slight bump at 3k can be managed. This combination may make for a bit of harshness.

Of course, all this is nitpicking, trying to squeeze out every last bit. I've been fighting with the 4ohm version of the Pecorino and similar issues. Get it flat, and the crossover is (probably) too low for the tweeter. There's a very broad peak in the response, probably due to the tiny box. It makes for a nasty peaking response no matter what I do until I add a zobel.

If you can squeeze in even an 18ga on the main woofer inductor it will help. The difference in size isn't dramatic for the Jantzen: 1.3"x2" for the 20ga, 2.24"x1.18" for the 18ga, but 1.15Ω vs 2.04Ω is a lot. Erse goes to 3" dia, so flatter, and 1.09Ω.

C

[JonW]

Another option to tilt the top end down would be to split up that 12 ohm resistor, say into 10 and 2. Put then 10 where the 12 is currently, move the 2 to be just before the tweeter in series. Then put an inductor in parallel with that 2 ohm resistor. You'll have to play with the inductor value, but you want something that provides around 2 ohms impedance around 5kHz or so... something like 0.05mH. What this does is, as the frequency goes up, the combined impedance of the resistor+inductor goes up, and that reduces the signal going to the tweeter. So you actually get to tilt the tweeter response down, instead of just dropping the overall level (which may or may not be what you'd like to do).

Just a suggestion. I've seen many other ways of achieving this, but the inductor thing has worked well for me in the past, without messing up the rest of the network too much.

Suarav,

Thanks for the idea! It’s a great one. I appreciate your stopping by and providing this feedback. I tried that out and, sure enough, it helped a little. It’s not a big effect by any means. But it helps. The results are shown below.

[JonW]

Hi Chris,

Thanks so much for the feedback. This is really helpful! The crossover is looking a little better. I’ll post the latest in a moment, below.

I know you'll love hearing this but - off-axis measurements + simulations may also help you determine what you care about, or if anything needs to change - even on that tweeter response. It's very probably flat even at 15° off axis. I'm in agreement, that change may not be worth it.

Actually, I collected the off-axis measurements already. Now and then I pop them into the crossover model. They look a little different, but, ummm, I guess for now I am optimizing things for on-axis.

The slight bump at 1kHz might be tweakable if you had a zobel in the mix. I'd probably also be fiddling to see if the slight bump at 3k can be managed. This combination may make for a bit of harshness.

Great idea- thanks! I added a Zobel to the tweeter and the 3k bump is minimzed. See below. The 1k bump is being more difficult. I can squish it down. But then the crossover point wanders up to 2,000 Hz or a little higher. I don’t think that’s a good idea with this woofer. But my guess is that it’s OK as it now stands. The whole thing is +/- 2 db which, I think, is pretty good. This 3k bump is not that big.

Of course, all this is nitpicking, trying to squeeze out every last bit. I've been fighting with the 4ohm version of the Pecorino and similar issues. Get it flat, and the crossover is (probably) too low for the tweeter. There's a very broad peak in the response, probably due to the tiny box. It makes for a nasty peaking response no matter what I do until I add a zobel.

Yeah, some of us like to nitpick, I know. In my case, I’ll be whipping up 3 other crossovers for this series of speakers. So I think that it’s worth getting the main one as right as I can. Then use it to build the other 3 from. I am also dealing with tiny box issues. Speaking of which…

If you can squeeze in even an 18ga on the main woofer inductor it will help. The difference in size isn't dramatic for the Jantzen: 1.3"x2" for the 20ga, 2.24"x1.18" for the 18ga, but 1.15Ω vs 2.04Ω is a lot. Erse goes to 3" dia, so flatter, and 1.09Ω.

I know that 1” difference does not sound like much but… When I look at the bottom of the cabinet, where I would like to mount the crossover, it is very little space. Subtract out space for the woofer magnet which is almost touching the bottom. Then subtract out space for the 1.25” Sonic Barrier that we decided was needed to minimize the dip at 700 Hz. That only leaves 3.25” x 5” of cabinet bottom for the crossover. It’s teeny, tiny! I’ll have to use the sides and some Velcro to mound part of the crossover, I suspect. But you can see how an inductor with 1” smaller diameter is worth taking a 0.9 Ohm hit. Anyways, that inductor value changed to 4.5 mH in the latest version. There is no 20 gauge 4.5 mH at PE, so it’s now an 18 gauge inductor with DCR of “only” 1.25 Ohms.

[JonW]

Here is the latest… Let’s see what you think about this one…
Things have improved a little more. I definitely like this one better. Kind of excited, actually. The frequency response is pretty flat now. +/- 2 db! The crossover point is still around ~1800 Hz. Keeping a low crossover point was a priority for me, with this woofer.





Phase did not get any worse. I had to tweak various component values to keep it this OK.





Impedance is nothing too special. But I’m OK with it given that the other aspects are looking good now.





Reverse tweeter null. Here, too, I had to fiddle with various component values to keep the null so deep.

[JonW]

And here is the circuit. As suggested by Suarav, I added the parallel resistor + tiny inductor. And as suggested by Chris, I added a Zobel to the tweeter. Varied the values of those parts. When things looked good, I had to readjust all the other component values to get the FR flat, good phase, hit the LR4 targets closely, and get the deep reverse tweeter null.

There is a lot of variability in this circuit for voicing later.

10 Ohm padding resistor:
Change down to 8 Ohms and there is an upward slope for the FR.

Parallel resistor + tiny inductor:
Change resistor from 3 Ohms to 0.5 Ohms and there is an upward FR slope.
Change the inductor from 0.025 to 0.050 mH and there is a slight downward slope.

Zobel:
Change the cap from 1.5 to 8 uF and there is a significant downward slope
Change the resistor from 10 to 1 Ohms and there is an upward slope.

Lots of options for voicing. Something in there has got to sound good.

[cjd]

Jon, the 18ga inductor is almost exactly the same size other than which dimension its radius. Of course, you have to orient it differently, and perhaps that's a problem I can't see as easily.

Zobel on the mid-woofer doesn't help with the 1kHz bump? Even a non-flat one? (i.e. use it as a tuning mechanism, not just a way to flatten the impedance)

Every little tweak is something to learn if nothing else.

[Saurav]

Actually, I collected the off-axis measurements already. Now and then I pop them into the crossover model. They look a little different, but, ummm, I guess for now I am optimizing things for on-axis.

Listen to Chris I think you're picking your crossover point primarily to avoid the woofer's distortion peak? That's important, but it's also important to make sure that the directivity of the woofer and tweeter match. Otherwise, you'll get the typical polar response which narrows near the top of the woofer's range, and then flares out again in the tweeter's range. And that's not going to sound good in any normal room, regardless of how flat the on-axis response is.

Your XO point is low enough that this might not be an issue, but it's always a good idea to keep an eye on the combined off-axis response of the drivers + crossover. Not sure what software you're using, that looks like LspCAD to me. I think you should be able to add the off-axis response to the driver models, and then the software should be able to calculate the combined off-axis response for every XO simulation without you having to manually swap out FR data sets.

I tried that out and, sure enough, it helped a little. It’s not a big effect by any means. But it helps.

Just another tool in your toolbox. A bigger inductor will have a bigger effect, but tat will mess other things up too. I think you're in the "add salt to taste" region now, and you'll be tweaking these values by ear before you're done.

[JonW]

Jon, the 18ga inductor is almost exactly the same size other than which dimension its radius. Of course, you have to orient it differently, and perhaps that's a problem I can't see as easily.

Yes, it’s very tight in there. I can’t see the whole crossover fitting on the cabinet bottom. Will likely have to use the sides. I’d rather not because then those boards might fall down some years in the future. For maximizing parts on the bottom, a tall, narrow inductor is preferred over one of the same volume but wider and flatter. For now, I’m working on the crossover, optimizing it, and then adjusting component values to what you can buy. When that big inductor was 4.0 mH, a 20 gauge version was available. At the moment, it’s 4.5 mH for which 20 gauge is not available but 18 gauge is. I’ll make the gauge decisions after the component values are set. Which they are not yet.

Zobel on the mid-woofer doesn't help with the 1kHz bump? Even a non-flat one? (i.e. use it as a tuning mechanism, not just a way to flatten the impedance)

Nah, it didn’t help. But I’ll try it some more, on a different day with fresh perspective. The dip on the low side of the 1 k bump is, I think, from the raw measurement. See the first image in my post #185 on this page. With the woofer Zobel, I could get the 1 k bump flatter. But everything else went to hell in a hand basket to get there. Totally not worth it.

Let me give it another shot and see if I can improve it a little. Worst case is that it stays the same as shown above. Which is pretty good, I think, considering everything (FR, phase, low Xover point, LR4 target, etc.). The current Xover is the first one that I’ve actually been excited about.

The other thing that I might try is to see if I can eliminate the LCR network. It helps to flatten the impedance a little but not tons. Also, it helps keep the FR smoother, the phase matched better, and the LR4 targets better. But just a teeny bit. Then again, maybe there is not much reason to avoid these 3 extra parts.

Every little tweak is something to learn if nothing else.

Yes! You folks are helping me a lot- both with this crossover and teaching me many things. I’ve learned more on this speaker project than I expected to. It’s totally worthwhile to tweak now.

[JonW]

Listen to Chris I think you're picking your crossover point primarily to avoid the woofer's distortion peak?

Exactly. Everything I’ve read about such metal coned drivers says that it’s really important. In this particular case, it’s a relatively small woofer, at 5.5”. So a low crossover point is important. And I’m using a tweeter that can handle working low.

That's important, but it's also important to make sure that the directivity of the woofer and tweeter match. Otherwise, you'll get the typical polar response which narrows near the top of the woofer's range, and then flares out again in the tweeter's range. And that's not going to sound good in any normal room, regardless of how flat the on-axis response is.

Your XO point is low enough that this might not be an issue, but it's always a good idea to keep an eye on the combined off-axis response of the drivers + crossover. Not sure what software you're using, that looks like LspCAD to me. I think you should be able to add the off-axis response to the driver models, and then the software should be able to calculate the combined off-axis response for every XO simulation without you having to manually swap out FR data sets.

Good point- thanks. Yes, it’s LspCAD. If I do the off-axis simulation at 30 degrees, there is the typical drop off at the high frequencies. If I use the measurements I collected at 30 degrees, there is a general downward slope in everything, the FR looks a little more lumpy, plus a small peak at maybe 10,000 Hz or so. Let me tweak the crossover a little more, as I mentioned to Chris, above. When the crossover looks like it is done, I’ll post the off axis FR both simulated and using the real measurements.

Just another tool in your toolbox. A bigger inductor will have a bigger effect, but tat will mess other things up too. I think you're in the "add salt to taste" region now, and you'll be tweaking these values by ear before you're done.

Yes, I’ll tweak by ear. But first, I’d like to get the computer model looking as good as possible. I really do appreciate the help. :T

[Saurav]

If I use the measurements I collected at 30 degrees, there is a general downward slope in everything, the FR looks a little more lumpy, plus a small peak at maybe 10,000 Hz or so.

A general downward slope is fine. Very loosely speaking, you want something that looks like this:



Rather than this:

[JonW]

Hi Saurav,

Those are cool plots. I don’t think that LspCAD can make them.

So what is it about a crossover that generates one result versus the other? I’d think that low crossover points would help.

Last night got to be too busy and there was no chance to fire up LspCAD. I’m hoping that I can try to improve the crossover later tonight. It may already be as good as I can get it. But I’ll try some more and see what I can do. After getting the crossover as good as I can get it, I’ll post all the off-axis info.

[JonW]

You want flat? I got yer flat right here.

OK, it’s almost down to +/- 1 db flat. No smoothing here at all. But… everything else looks just horrible. The crossover point is too low. The phase and impedance are awful. And the reverse null is only 5 db deep. It’s kind of an LR4. I have not given up yet. Some of these issues can be improved, although I do not know if they can be solved. At the moment, I am favoring the last full crossover that I posted above, with the complete circuit and all the associated diagrams (posts 201 and 202). I’ll keep at it and see if a total package can yield anything better than that one.

[Saurav]

So what is it about a crossover that generates one result versus the other? I’d think that low crossover points would help.

You want to cross over in a region where the woofer and tweeter both have similar off-axis drop-off. That way, an XO that smoothly transitions between them on-axis, will also generally have a smooth transition off-axis. Lower is generally better with a dome tweeter, because both woofer and tweeter have fairly wide off-axis response. If you're using a horn tweeter which has a narrower directivity pattern, you'll often find people pairing that with a larger woofer, which is more directive at a given frequency than a smaller woofer, again, to get the directivities to match up at the XO point.

But that's just one consideration out of many, and design is all about trade-offs. Like you discovered, you can get very flat FR, at the cost of phase, amplitude, etc. Is it worth it? Trust your ears

And I should warn you, I only know just about enough to be dangerous to myself. So take everything I say with a big grain of salt

[JonW]

You want to cross over in a region where the woofer and tweeter both have similar off-axis drop-off. That way, an XO that smoothly transitions between them on-axis, will also generally have a smooth transition off-axis. Lower is generally better with a dome tweeter, because both woofer and tweeter have fairly wide off-axis response. If you're using a horn tweeter which has a narrower directivity pattern, you'll often find people pairing that with a larger woofer, which is more directive at a given frequency than a smaller woofer, again, to get the directivities to match up at the XO point.

OK, that makes sense. Adding this extra issue can only make speaker design more complicated. With my driver choice here it does not sound like it will be too big of a deal. We’ll see. Driver choice is part of the design effort. I go for low distortion drivers so that aspect is good up front.

But that's just one consideration out of many, and design is all about trade-offs. Like you discovered, you can get very flat FR, at the cost of phase, amplitude, etc. Is it worth it? Trust your ears

Indeed! Making one aspect look good is easy. Making them all look good... not so much. So the best speaker is probably the best middle ground between all the parameters. Which isn’t to say that all aspects are weighted equally, but that’s the general idea. The latest FR plot (post #209) is very flat. But everything else looks like total junk and it would not be a good sounding speaker at all. I already know that much. At this point, I do not know if I can improve on what I have but I am going to give it some more effort and see.

For perspective, I’ve been looking at the final FR plots of the various completed designs here in the Missions Accomplished section. My +/-2 db FR of the full crossover above in posts 201 and 202 seems to be flatter than typical. So if I can’t wring out any more improvements, I might be OK with what I’ve got.

And I should warn you, I only know just about enough to be dangerous to myself. So take everything I say with a big grain of salt

Heck, I can’t imagine that anyone reading this thread would think that I have any idea what I am doing, either. We all learn from each other.

[technodanvan]

So what's the word Jon? Make any progress on this recently?

[JonW]

Hi Technodanvan!

Oh, I can’t tell you how busy life has been these past few months (work, family, etc.). I haven’t even been reading things here at htguide. So there is no progress to report on the speakers, themselves.

However, these little speakers will need a sub to play along with. And I have made some progress there. It is shaping up to be a coffee table-type of sub. Waaaay bigger than the speakers. Weather is getting colder here. So I had the sub out in the garage doing the finish work while I still can. That’s now complete. Just brought it back into the house last night. Still need to add the plate amp, driver, cabling, etc. No photos yet with the finish coats on but this old photo will give you an idea what it is looking like. I may do a build thread writeup when the sub is done.

Tomorrow I leave for a 10 day work trip so no speaker progress planned there. But when it gets cold, I will have to spend more time indoors. So my precious little free time will shift back to the speakers. The current commercial speakers in my office are dying from years of use. I really need to get this speaker project up and running. Soon, I hope.

Take care!


Image not available

[technodanvan]

Glad to see progress being made despite life getting in the way - it's just one of those things, right?

I'm looking forward to the cold too, though in Vegas that's pretty relative.

[JonW]

Chipped away at the speakers a little bit...

In room, deluxe (W16) speakers:
Played around with the crossover. I think that I’m about at the best combination of flat FR, phase, impedance, and LR4 that I’m going to get. One component changed a tad but it’s basically the setup shown on page 6, posts 200 and 201. I do want to try a few other things in LspCAD, in the name of seeking perfection. But I suspect that it’s about as good as it will get prior to buying parts and voicing. Keeping parts values to what can be purchased. I’m pretty happy with how it all looks.

On wall, deluxe (W16) speakers:
I started from the crossover mentioned above and used on wall measurements taken for 9 ms after the pulse. Can see the wall bounce nulls and such. FR looks a little different. Measurements are much noisier than what I have posted above. Unlike the in room measurements, I am having to add some smoothing to the plots here. Adjusting the crossover to flatten things out. Looks pretty good but I want to tweak it some more.

One more design constraint I’m adding in here is that I am trying to make a switchable crossover such that it can flip between being in room and on wall. I built the cabinets with 2 pairs of binding posts. So I’m trying to have one crossover for one binding post pair. Then have only the extra parts needed to change from on wall to in room (or vice versa) connected to the other binding post pair. It’s making things a little trickier. One cap value change helped the in room -> on wall change. But I can’t just add a cap in series to a circuit without throwing everything out of whack. Inductors and resistors, however, can be added in series easily. So it’s a further challenge here but I think that it will both work and be worth it.

Will post plots when I’m farther along.

[JonW]

Here are some fun results to report. I’ve got one crossover that is looking really good. And another that has left me scratching my head. I could use some help on the second one. But first, the good news.

In room, deluxe crossover:
It’s better now. I tweaked it more and got results that I am quite happy with. One of the changes that helped was Chris’ suggestion of adding a woofer zobel- thanks! I was shooting for an LR4 crossover at about 1700 Hz, given the drivers being used here.

Frequency response:
Ahhh, very flat, almost +/- 1 db. No smoothing was used here. The crossover point is where I want it (1700 Hz). The most recent tweaks helped get this pretty graph.



Phase:
I’ve seen better with other speakers but this plot here is one of the best I’ve been able to get with these drivers/measurements. I’m happy with it.



Impedance:
No big peaks or swings above 100 Hz. Just what I like to see.



Reversed tweeter null:
A 30 db null looks good to me.



Off axis data:
I took measurements 30 degrees off axis (to the side) and plugged those data into the crossover. Honestly, I did not expect anything to look as good as this one does.



Circuit:
It’s got a lot of parts but, eh, the graphs look good. Note that resistors 2, 4, 8, and 9 are not real. They are there to account for the inductor DCR’s. Using 20 gauge inductors to save space in the tiny cabinets. All the parts are at standard values that can be purchased or made by combining parts.




OK, I am declaring this crossover to be done for this stage. Next up is voicing. Then the crossover will probably change a bit.

Below is the on wall deluxe crossover. I am having some trouble there and would welcome any input.

[JonW]

On wall deluxe crossover:
Here is where I am less happy with the results and I feel like I have less understanding of what is going on. I took measurements with the speakers up against a wall and recorded for a while to be able to see the bounce back from the wall and the resulting cancelations/nulls. I think it was 9 ms of time. The data are noisier, which is to be expected. Then I tweaked the crossover a bit. More on the circuit below.


Frequency response:
It’s OK. Not pretty like the one up above for in room. I applied a wee bit of smoothing here (1/24) to make the plots workable. The dip at ~5,000 Hz seems to be in the measurement, not the circuit. The null at ~275 Hz is from the cancelation wave between the front of the cabinet and the wall. Nothing that can be done about it, really. The nulls at ~450 and ~750 Hz are probably also from the wall. For an explanation of that, see Figure 2 in this excellent link:

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

Phase:
This part looks good and I like it.



Reversed tweeter null:
It is a null. But only ~17 db deep. OK, I guess, but certainly not great.



Off axis FR data:
Like with the in room example above, these measurements were taken 30 degrees off axis (to the side) and placed into the crossover circuit. Ummm, I don’t know what to think. Ugly? Acceptable? Any thoughts?



Off axis phase:
Really ugly. I probably could not have made it worse if I tried. But thinking about it, I would expect the phase to turn bad as we get off axis, where the drivers are doing their thing more differently than on axis. So is this one a problem or no?


Circuit:
What I would like to do, ideally, is to make a crossover that is switchable between in room and on wall. I built the cabinets with 2 pairs of binding posts. So maybe I can have one crossover connected to one binding post pair. Then have just a few extra parts connected to the other binding post pair, then run into the main crossover. That way, one binding post pair is for in room and the other for on wall. If that makes sense. This crossover circuit is, basically, the in room circuit with the values of two parts changed (R1 and L4). The increased/decreased values of these parts could be changed by adding a resistor and an inductor in series, on one of the binding post pairs.


So this on wall crossover has minimal changes to the in room crossover. But will allow easy switching between the two, depending upon which binding post pair the cables from the amp is hooked up to. Easy.

But are the results acceptable? I don’t know. I’ve never worked with on wall, long timed measurements. The data are noisier given that I wanted to collect for long enough time to see the wall effect.

Options for improving this circuit include
A) Tweaking this crossover a bit, which I will certainly try.
B) Adding some on/off switches with added capacitors, to improve the look of things. Then I could open up the speaker and flip switches when going between in room and on wall. Which won’t happen often.
C) Just make a fresh crossover for this situation. I don’t like this option because I may be moving in the coming months and I have no idea what the next room will be like- whether the speakers will have to be placed up against a wall (like in the current room) or they could be placed away from a wall. So I’d like some flexibility.

Anyways, I’d love to hear some feedback for this on wall crossover. Is it good or bad, given the on wall placement and long timed measurements? And what suggestions are there to improve it? Thanks!

[JonW]

I played around with the on wall crossover a bit and things look a little better now. Although I’d be curious to hear if anyone out there has perspective on these plots. Good or bad? Given that it’s on wall data collected long enough to see the nulls resulting from the wall bounce and cancellations?

Phase and impedance were similar to my last post, above. The LR4 target and 1700 Hz crossover point both look good.

The current frequency response. Flatter than before. The big nulls are from the wall cancellation. I kind of like this plot, all things considered.




Reverse tweeter null. A little better at -20 db. But I’ve seen better. Although I guess not with an on wall setup using data collected for this long. I could get the null to be much deeper (>-30 db) but the phase did not match up as well.




Off axis response, 30 degrees to the side. Eh. I’m not impressed. But maybe it’s as good as I can hope for, given the circumstances?

[JonMarsh]

You highlight a very specific and necessary point- "under the circumstances".

Usually I would only try a flush to wall approach- this I've done in recording studio monitor builds in the 70's and 80's. that likely will help the off axis stuff, too. You've got a very discontinuous baffle surface with wall mount as I understand what you're doing. Considering those circumstances, I'd say you're doing pretty well...

[JonW]

You highlight a very specific and necessary point- "under the circumstances".

Usually I would only try a flush to wall approach- this I've done in recording studio monitor builds in the 70's and 80's. that likely will help the off axis stuff, too. You've got a very discontinuous baffle surface with wall mount as I understand what you're doing. Considering those circumstances, I'd say you're doing pretty well...

Hi Jon,

Thanks for the much needed perspective and encouragement.

Yes, I’m making the speakers in their cabinets and then one crossover for normal, in room use. And then another crossover for when the speakers are pushed up against a wall. If things go well what I’ll really do is make the crossover switchable between in room and on wall. I suspect that a lot of people have room constraints and need their speakers pushed to the wall. So maybe these speakers will help out someone else.

Up against the wall is how I need to have them in my smallish office. I do realize that flush mounting speakers into a wall avoids a lot of design issues, like the typical wall bounce cancellation nulls I’m getting. But if I ripped open the office walls like that I would get fired from my job. In which case I would certainly eliminate the speaker-up-against-the-office-wall design constraint.

[JonW]

The wallet is a little lighter now. I just ordered all the parts that I'll need for voicing the crossovers. I could use a bunch o' caps, inductors, and resistors to supplement the in-house stash. Plus some other things like wire to make up the speaker cables, switches for the switchable crossover, and more alligator clip leads. Also I ordered a replacement tweeter for the one that looks like it's bad. Fun toys are on the way.

[JonW]

It’s alive!




I have begun the voicing and, what do you know, the speakers actually work. And they sound really good, too.

I’ll post pictures and details in the future, when I get a chance. For now, I’m spending as much time as I can on the actual speakers. These are the in room deluxe flavored speakers. The crossovers are just held together with alligator clips and wires, so I can change each component to higher and lower values, to see what sounds best.

First impressions are that they are really clear and crisp. No distortion. A nice high-to-mid balance. Little bass. The numbers are changing a bit but the general crossover is the same as posted above (post #216). Actually, I started out by changing some component values to bring out more air and highs. But after longer listening, backed down to the original values, which I think will be better for long-term listening.

The only tricky part here is deciding on the mids-versus-highs balance. I’m using ~10 different music CD’s for the voicing. You can really tell how some are heavier in the highs and some are not. Voicing to one CD may get it perfect but then it does not sound great with another CD. So I’m trying to find a balanced crossover that makes all the music and recordings sound their best.

Still more tweaking to do. It’s fun to have working speakers to listen to.

[JonMarsh]

Still more tweaking to do. It’s fun to have working speakers to listen to.

Yeah, it's good to finally get to that stage, isn't it?

[JonW]

Yeah, it's good to finally get to that stage, isn't it?

Indeed it is, indeed it is. Work on this project has been spread out over so much time that it’s easy to forget the whole idea is to make little boxes that actually produce sound.

[JonW]

Been doing more listening and crossover tweaking. These drivers are really clear. They don’t seem to distort, no matter how loud I play them. And for teeny little speakers, they project a nice wall o’ sound. Bass is lacking and so a sub is essential. But the bass that is there is quite clean. Everything that comes out of these little guys is crisp.

On the very off chance that anyone out there is playing along at home and building these speakers... The deluxe (W16 woofer) in room crossover is pretty much done. There may be more tweaks but I’ve been listening for days and I think that the crossover is set. Use the crossover circuit shown above in post #216 but change C2 to 32 uF and R10 to 3 Ohms and that should be good. Details on all of the specifics and impressions to follow some day.

I was curious, so just to get an idea of costs, a pair of crossover components is running in the range of $230, roughly. Add in drivers and the deluxe speakers may be around $1,000 per pair and the basics could be in the ballpark of $650. I’ll post specifics when I have them.

So one crossover (deluxe, in room) is done and there are 3 more to make: One is the deluxe (W16 woofer), on wall. It’s done on the computer. Still need to voice it. Then there are the basics (ER15 woofer) both in room and on wall. I have measurements and need to start the crossover design. Kind of from scratch. But the tweeter circuit from the deluxe flavors will be a good starting point.