Javier's RS180/RS28A MTM & mic cal thread....

[fjhuerta]

Hi all!

I've been measuring and re-measuring things to get a clue as to why my Natalies don't sound good at all, despite following the original designs as closely as I could.

Some of you know my speakers have a sharp peak at 1700 Hz, almost 1KHz wide.

I started by finding out that the tweeters were to blame, so I chamfered the baffles (they had 90° angles). The FR varied a bit, but the peak was still there.

I started by creating a felt mask around the tweeter, the size of the baffle. Things changed a bit, but the peak was still there.

After a week or so of thinking about it, I covered the woofers with pillows and re-measured.

Bingo (?). The peak was gone, and FR was as flat as I could possibly expect.

I have included three measurements (without any diffraction treatment, with felt on the baffle, and with woofers covered).

I now know what's wrong, but have no clue as to how to fix the problem (if it can be fixed). This is the first time I've had this problem. I'd suppose the phase plug and the size of the drivers are part of the problem, but, AFAIK, there's no solution to my problem other than a very complex crossover. I already tried that, and I thought the sound was dull and dead (although FR was completely flat).

What can I do? The only thing I can think of is sink the woofers even deeper on the baffle (they may be 0.5 ~ 1mm ahead of the tweeter), and that's it...



Thanks for any help!

[Paul H]

I'm not sure you're seeing what you think you're seeing ...

The graph you posted could be read as showing a dip in response centred around 2300, rather than a peak at 1700 Hz.

If you're covering the woofers with pillows, you're measuring the response of the tweeter with pillows interfering with accurate measurements of the real tweeter response.

You might want to post woofer measurements also, let us know your testing conditions, and maybe snap a pic of your crossover. I'd venture to say that if a Nat design sounds actually bad there's a crossover construction problem or a faulty driver.

[JonMarsh]

You've correctly identified (for the most part) a problem which plagues to one extent or another all small speakers with drivers packed close together.

The effects can be lessed by using well rounded baffle edges (3/4" radius roundover), and a sheet diffraction control with a small diamond shaped cutout for the tweeter. In principle, you could also include an LCR notch filter in the treble high pass, but we're trying to get away from more crossover parts with the Natlie P design.

This is the measured repsonse of a completed Modula MTM, which uses the same drivers and cabinet layout- with a rounded baffle and diffraction control felt.

In this speaker there may be a little compensation for the tweeter baffle and driver reflections in the crossover, but not much- there's no separate notch filter.



The roll off above 16 kHz is due to a microphone cable with too high capacitance that was used at that time becuase of the difficulty of getting a replacement cable for my B&K 4133 mic body.

Personnaly, I don't care for drivers with phase plugs, like the Seas or Dayton, but the value proposition of the Dayton's is quite high in terms of general performance versus cost. My own opinion is that concave diaphragms give better cone control, higher breakup frequencies, and less diffraction issues, but not that many vendors use them- Accuton, Aurasound, HiVi, and Revel being among the few that offer or build quite a few drivers with this type of cone.

Wider driver spacing can be part of the cure, too, but then this necessiates changes in the crossover frequency (much lower) and more expensive crossover approaches.

Another technique to consider for new designs is a waveguide, but this does introduce other complications.

You do have the offset tweeter placement on the baffle, right? A centered tweeter will also lead to problems.

~Jon

[fjhuerta]

You might want to post woofer measurements also, let us know your testing conditions, and maybe snap a pic of your crossover. I'd venture to say that if a Nat design sounds actually bad there's a crossover construction problem or a faulty driver.

Hi Paul!

The woofer measurements are as close as possible to the stock PE measurements, so I'm not worried there.

The measurements I took are of the "raw" RS28A, without any crossover on the signal path.

Jon, thanks a lot for your answer! for the record, the roundover is 3/4", indeed, all throughout the baffle. The baffle itself is identical to the reference design, with the tweeter offset by the same amount. It doesn't have any felt right now (I removed it). I'm left scratching my head as to why nobody has had the same diffraction problem I have.

I'll try to take some pics of my baffle and post them before New Years Eve (party at home!).

[fjhuerta]

Photos of my NatP's

Here are the pics of my Nat-P's. I hadn't noticed the woofers are not exactly flush-mounted; I suppose this has a lot to do with the foam gasket I made for them. Could this be the culprit?

[Dennis H]

Edit: I started to say yes but Zaph's measurements don't show it. Thick (1/2") wool felt (or an old wool blanket) should help a bunch with any diffraction problems.

http://www.zaphaudio.com/mtg-surface.html

[dlr]

The only fix for adjacent driver diffraction is felt

Here are the pics of my Nat-P's. I hadn't noticed the woofers are not exactly flush-mounted; I suppose this has a lot to do with the foam gasket I made for them. Could this be the culprit?

I've found there to be very little diffraction impact from the rim of a surface-mounted driver adjacent to the tweeter. The distance is highly distributed, it's at an angle that has much less impact on most listening axes and it's not large in surface area.

You'll have two main sources to consider. These are the baffle edges in the short dimension and the doubled amount from two drivers in the MTM vs. an MT. The larger the adjacent unit, the more impact due to larger included angle and a larger surface area to create the diffraction. Mid/woofers have much more impact than a small midrange unit.

Try a test using a 1" wide by 6" long by 1/2" or even 1" depth on each side of the tweeter between it and the other drivers. The extra thickness of a 1" piece is usually very helpful for adjacent drivers.

Strips such as these aren't too aesthetically displeasing. Systems such as the old NTH 3.3 used a small strip on one side that seemed to blend in, but they were shallower and rounded on the ends.

I suspect that there's little need for a strip on the near edge. The far edge of the width dimension will likely still have some diffraction, even with the roundover as it's probably too small in radius to help a lot for the frequencies involved with this edge. The near edge probably does benefit from the roundover.

Another factor is the hard diaphragm of the mid/woofers. I've found that hard materials have a bigger diffraction signature as they are much more reflective than paper or poly materials.

[oxcartdriver]

This is the measured repsonse of a completed Modula MTM, which uses the same drivers and cabinet layout- with a rounded baffle and diffraction control felt.

I'll have to try the diffraction control felt on my modula MTM's. I built my own cabinets with a 3/4 round over and matched the dimensions of the PE cabinet front facia.

[jkrutke]

Hello,

I think what I'd do at this point is take a very near field measurement of the tweeter. About 1cm from the dome should do, and this should remove most of the baffle from the response curve giving you a smooth low end rolloff.

Your low end should then look close to this infinite baffle plot:

http://www.zaphaudio.com/oldversions/

If it doesn't, you likely have a defective tweeter.

[fjhuerta]

Hi John!

I was pretty sure the driver was defective, and took that measurement yesterday as close as I could take it to the tweeter. The measurements look exactly like your reference plot. At around 15 cms the diffraction starts being obvious, and at 30 cms the 1700 Hz peak is quite prominent.

[jkrutke]

Well that settles that about the response curve then. The frequency irregularities are definitely diffraction related. Note that some of it is baffle diffraction, but some of it is likely woofer cavity effect, an issue (to some extent) in every large MTM. It's usually a dip with a peak just above it somewhere between 3 and 4kHz.

The next thing I'd do is make sure your problems are not non-linear distortion related. (rub and buzz)

Good:



Defective:

RS28A4-HD-sample1.gif Image not available


Defective:
RS28A4-spectral1Khz-sample1.gif



Note that you don't have to do these plots to see the problems, you can hear rub and buzz with a simple signal generator sweep. Start at 800 and go up.

Assuming no problems with rub and buzz, it pretty much has to be frequency response and/or power response that you are hearing. I would not rule out rub and buzz until it's been proven however. It's very hard to specifically detect with music but a sine wave can easily make it audible. This has been an issue with RS28's, both early and recent.

[fjhuerta]

John, you've got me thinking...

I've simulated most of the RS180 / RS28 designs (the Modula, Natalie, RJB's and even the Dr. K's) and I've found out either my measuring equipment is completely wrong, or something else doesn't make sense.

The woofers' simulations always are correct. I always get the roll-off rate with every design I model on LSPCad, and the break-up modes are properly supressed.

But the tweeters, besides the 1.7 KHz peak, have something else. They are way, waaaay louder than the woofers. The reference designs' attenuation make the tweeter louder by about 2 dB or more.

Just for kicks, I took the Madisound MB Quart tweeter, installed it on the Natalies, and took some quick measurements. There's no 1.7 KHz peak with it, and I got a flat design within 5 minutes with LSPCad.

At this point, I've tried everything - I even added 1" of sound deadening material around the woofer cone, without any good luck - but substituting the RS28A's got rid of the peak...

I think I'll be buying a pair of SEAS tweeters (the 27TBFC/G), which, I understand, are a drop-in substitution for the RS28A, and hope that indeed, the RS28's are damaged.

[Dennis H]

I think I'll be buying a pair of SEAS tweeters (the 27TBFC/G), which, I understand, are a drop-in substitution for the RS28A, and hope that indeed, the RS28's are damaged.

They are not a drop-in. You'll need to change the crossover.

[fjhuerta]

They are not a drop-in. You'll need to change the crossover.

Oh, yes, sorry about that I know the crossover needs to be re-done. I'm not that worried about it - it's the shape of the baffle cut-outs I'm concerned about.

[fjhuerta]

Back to square one...

Darn...

I just re-checked the MB Quart FR charts.

The peak is still there. I missed it because I had smoothed the graphs. When I saw the original measurement, it was quite clear the problem hadn't gone away with a new tweeter. :cry: :evil: :roll:

I removed the woofers from the enclosure and re-measured.

The peak is *still there*. The diffraction measurement is a bit different, but it's obviously not that much better. So flush-mounting the woofers will not help that much.

I checked the flush mounting on the tweeters and fixed it a bit so that it was as close to the baffle as it could be.

No luck.

I think I'm going to destroy the cabinets in a fit of rage... no... wait... count to ten... and breathe slowly.

What else is there for me to check?

Could a cabinet resonance mode develop that high (1700 Hz). Considering my speakers are almost 1.30 meters high, I'm beginning to think the bracing / baffle are simply not up to the task.

At this point, I'm willing to cut the cabinet and make a pair of big bookshelf speakers, just to see if, indeed, the baffle or cabinet are resonating...

Am I going crazy? I had never had such a strange problem (I've built about 6 pairs of speakers, and by far, this has been the most troublesome of them all).

[Evil Twin]

Do your enclosures have a loose roll of batting behind the midwoofers, dacron or polyester? Long fiber wool works well, too, but is more expensive.

[fjhuerta]

Do your enclosures have a loose roll of batting behind the midwoofers, dacron or polyester? Long fiber wool works well, too, but is more expensive.

No, they have 1.5" of foam, or so... still I'm not sure this would help, since my last measurement was done without the woofers, and the peak was still there.

[speedle]

Has anybody else had this problem with Natalie P's that can be documented? Maybe a consistent change to the sticky design? Just asking before I build... ;x(

[fjhuerta]

Speedle, as far as I know, I'm the only one having this particular issue, so you should not have any problems at all.

I'm not sure everyone has taken measurements of their own Natalies, though. So maybe the problem is a bit more common than I think.

I've gone through 10 crossover iterations by now, and right now I think I may have diffraction under control with a 6th order acoustical LR filter. I may build it this week-end and post some results.

I'm now wondering whether I'm the only one with this particular problem, or if everyone else is just living with it...

[fjhuerta]

Has it been more than a year already? Wow.

I decided to take another shot at these speakers.

After listening to them for this long, I found the following:

1) Their top-end is suppressed.
2) They have "shouty" mids.

The tweeters aren't damaged. Yesterday, I dropped in a new RS-28, just to be on the safe side. FR is basically identical.

I also put a new 27TBFC. Although FR drops faster, there's still a peak at that frequency.

I'm still scratching my head, wondering why noone else had this problem while building their Nat-P's or their Modulas. This is a pretty obvious problem for me. What did I do wrong? The only thing I did differently was building a big tower instead of a bookshelf.

Anyway, those packages from McMaster Carr arrived last week, so now I'll see what some real wool felt (F-13, 1/4", 2" width) will do to this issue.

Although I suspect felt will help only in the highest frequencies, I still have hope...

[dlr]

Double up or triple the thickness of the felt

Has it been more than a year already? Wow.

I decided to take another shot at these speakers.

After listening to them for this long, I found the following:

1) Their top-end is suppressed.
2) They have "shouty" mids.

The tweeters aren't damaged. Yesterday, I dropped in a new RS-28, just to be on the safe side. FR is basically identical.

I also put a new 27TBFC. Although FR drops faster, there's still a peak at that frequency.

I'm still scratching my head, wondering why noone else had this problem while building their Nat-P's or their Modulas. This is a pretty obvious problem for me. What did I do wrong? The only thing I did differently was building a big tower instead of a bookshelf.

Anyway, those packages from McMaster Carr arrived last week, so now I'll see what some real wool felt (F-13, 1/4", 2" width) will do to this issue.

Although I suspect felt will help only in the highest frequencies, I still have hope...

I use 3/4" felt typically, but found that doubling that to 1-1/2" between drivers was much more effective. You should have 3/4" at a minimum IMO.

If you can put a built-up amount that is 2" wide x 6" long x 3/4" (or more) thick between the drivers, do so. You can overlap the woofer up to just short of the surround, just ensure it doesn't actually touch. Overlap the tweeter a bit as well. Rectangular pieces have a nice distributed distance from the tweeter cone, so it's not a big issue when using only pieces of felt to separate drivers acoustically. Round openings should always be avoided.

After testing this, try adding some to fill in the area to the side farther away. Don't put any on the short side, it won't do much with the roundover there.

Woofers and midwoofers that are deep with hard cones exhibit a lot more diffraction influence. Doped paper cones seem to have internal damping that seems to reduce their influence on diffraction.

You'll probably also see some change in response from the midwoofers as well. There is some diffraction (aside from step) related to them, but it shouldn't be dramatic. It will be for the tweeter.

Dave

[fjhuerta]

Thanks for your help, Dave.

Can I just pile up self-adhesive wool felt? I only got 1/4" thick, so I'd have to double / triple the layers... do you think it would help?

[dlr]

Thanks for your help, Dave.

Can I just pile up self-adhesive wool felt? I only got 1/4" thick, so I'd have to double / triple the layers... do you think it would help?

Yes, you can pile it up and it will help. The only problem is adding/removing it. The self-stick backing on F13 loses tackiness quickly. When experimenting, I use two-sided tape at times. It doesn't adhere well to the felt after the first time it's applied.

I also test with two-sided tape on the sticky side cover at times. It's easy to attach temporarily for testing, but has a tendency to stay on the baffle when removing the felt. Experiment a bit, see what works best.

Dave

[jdybnis]

A couple of tips on felt:

- Get it from Sutherland Felt. I bought a few samples from McMaster to experiment with at first, but they are way too expensive. A 72"x72" square of 1/4" F-11 cost me less than $75 from Sutherland, including shipping to California.

- Those big cans of canned fruit or vegetables you can buy at Costco are exactly the right size for cutting RS180 sized holes.

Dave,

What's the effect of going from 1/4" to 3/4"? Do you worry about horn loading effects on the tweeter?

[dlr]

A couple of tips on felt:

- Get it from Sutherland Felt. I bought a few samples from McMaster to experiment with at first, but they are way too expensive. A 72"x72" square of 1/4" F-11 cost me less than $75 from Sutherland, including shipping to California.

- Those big cans of canned fruit or vegetables you can buy at Costco are exactly the right size for cutting RS180 sized holes.

Dave,

What's the effect of going from 1/4" to 3/4"? Do you worry about horn loading effects on the tweeter?

Thanks for the link, M-C also dropped some of the dimensions in their lineup a while back.

Felt that is too tall and too close can have a horn effect, though it's not predictable at all. A circular opening, especially small diameter, is a concern. Height as an issue diminishes rapidly with distance, especially with non-circular openings. Almost anything not round works well.

For rectangular or other sided shapes, a single piece very close is not much of an issue, either. I use a piece between tweeter and adjacent driver, even if the spacing is narrow. The improvement is worth any minor reflection issues.

One of these days I need to add my audioXpress article on stepped baffles and diffraction to my site. It's far more detailed than what is there now.

Dave

[Dave Bullet]

Probably too late now... but when I find a diffraction anomaly, I convert the frequency dip (or peak) to a measurement (based on wavelength at that frequency) then measure out from tweeter centre until I find the culprit.

In my 8"/2 way design - the dip was bang on 2250Hz, followed by the peak (as John K described). And funnily enough the frequency exactly matched the distance between the tweeter centre and woofer centre (and no other points on the baffle).

If baffle treatment doesn't sort it, then a notch filter could be added to tame the peak. I'm not too clued up on whether a series or parallel notch would be better.

Cheers,
David.

[dlr]

Probably too late now... but when I find a diffraction anomaly, I convert the frequency dip (or peak) to a measurement (based on wavelength at that frequency) then measure out from tweeter centre until I find the culprit.

In my 8"/2 way design - the dip was bang on 2250Hz, followed by the peak (as John K described). And funnily enough the frequency exactly matched the distance between the tweeter centre and woofer centre (and no other points on the baffle).

This is a good method to see if anomolies are diffraction related. In testing the new Seas DXT recently, I was curious to see what impact there was from each of the "diffraction expansion" rings in the driver. The fact that the rings are concentric accentuates their impact, as can be seen in the measurements and tests I posted.

What surprised me was the impact of diffraction from the motor mount screw openings. Granted these are so high in frequency that they are likely inaudible, especially off-axis, but there is quite a bit of impact in the measured response. With woofers, it's got to be audible if not controlled given where the diffraction is seen in the frequency domain.

Dave

[fjhuerta]

Felt is ready. As it is right now, I have 3/8" of felt all around the tweeter, overlapping the woofers, etc. I'll add a pair of 6"x2"x1/4" strips below and on top of the tweeter, so I'll end up with 5/8" in the most problematic area (I think).

One silly question, though - how do you deal with the uglyness of the felt (IMHO, of course)? The baffles looked very pretty until I tacked it on top. They now look, according to my wife, as if someone had dumped some trash in it (I cut the felt pretty nicely, if I say so myself - but it does look ugly).

I believe the only real answer will be to add a grill afterwards.

[fjhuerta]

This is interesting, too. If I take the distance between the center of the tweeter and the center of the woofer, and use it to calculate the frequency in Hertz of the wavelength, I get 2,300 Hz. Which is *very* close to the minima of the dip I see in my graphs.

So, I'm not saying I found the problem... but it very well could be it.

[fjhuerta]

The short answer...

Overall FR changed a bit. The peak flattened a bit. The speaker looks pretty ugly

I'm tired of trying to make this work. I'd love to know if anyone else found the same issues with the Modula / Natalie. As for now, I'll disassemble both speakers and build a three-way based on the RS-90, RS180 and RS28. I'm pretty sure this one won't have as many issues as my old one.

This is the only speaker that has left me completely confused for close to two years...

[Dave Bullet]

Hi Javier,

A peak is more audible than a dip. Have you tried notching out the peak? I would personally prefer that than putting "trash" over the front of the speaker (Isn't it funny how our better female halves might be from different sides of the planet, but they all think the same.... I could just see my partner saying exactly the same if I smothered the baffle with felt).

[ThomasW]

I'd love to know if anyone else found the same issues with the Modula / Natalie.....This is the only speaker that has left me completely confused for close to two years...

In the more than 3 years since Jon first posted this design no one has described a similar problem. So it would appear the issue for whatever reason is unique to your build.

[fjhuerta]

In the more than 3 years since Jon first posted this design no one has described a similar problem. So it would appear the issue for whatever reason is unique to your build.

I have to agree with you, since I have never seen such a strange tweeter response in any other design I've built. I'm completely baffled.

I think I'll play with the baffle for a while (change drivers, add felt, whatever) just to really learn what went wrong with this build. So far... well, I really have no clue, other than (apparently) the woofers are diffracting the tweeter response in a nasty way.

[subynube]

Well, thought I would add my .02 cents. I have built the Nat P's, using PE cabinets. I have it ported as it should be in the write up, and properly damped. It is powered by emotiva amp and processor.
I thought I did have a problem with these speakers (never measured, just subjective opinion). They were quite shouty mids, with what I thought was a forward sound. I also have hardwood floors all over. I recently tamed my environment by putting in huge area rugs. Now it is a year later, and I have gotten used to the sound. I still believe there might be some forward sound there, but it is much less. However, possibly because I am quite used to them now. I also thought I was the only one, because everyone else had very high praise for these. I still have high praise, but always wondered if it was just me hearing "forward" mids.

Just my .02.

[ThomasW]

They were quite shouty mids, with what I thought was a forward sound.

Did you put the wool felt on the front baffle?

[subynube]

Did you put the wool felt on the front baffle?

I have not.

[ThomasW]

Adding the wool felt to the baffle in addition to treating the room's first reflection points goes a long way toward taming the in-room performance of any speaker.

[fjhuerta]

subynube,

Your problems sound exactly like mine. Shouty mids, forward sound, got used to them - until I built another pair of speakers which didn't have so many problems.

Do you have access to measuring equipment? I'd be interested in looking at the overall FR you have.

I decided to re-work (yet again) the crossover on my pair, and I added a tiny bit of a "BBC Dip" from 1 KHz to 2.5 KHz, or so. A very gentle -2 dB dip (I measured the speakers, and I found out their measurements weren't similar to the ones I took 1 1/2 years ago - perhaps the drivers finally broke in?).

The speakers sound better than ever. I'm working now on eliminating as much as possible the dip without getting that "forward" or "aggressive" sound quality, one step at a time.

I know far too many people have built this speaker for it to be badly designed. It's 99% more probable I did something wrong. I just can't see what. So far, I've ruled out the tweeter, woofers, the crossover, the baffle layout... basically everything.

[Dave Bullet]

Hi Javier,

It's probably you've listened to / built enough speakers to discover your voicing preference. I suppose many people new to speaker DIY build the most talked about / popular speaker first - designed with the voicing preference of the designer. You can see similarities between different designs by the same designers (say for example Troel, who tapers off the FR - by about 2dB per decade up the spectrum).

Nothing wrong with having a different voicing preference.

Cheers,
David

[Dennis H]

So far, I've ruled out the tweeter, woofers, the crossover, the baffle layout... basically everything.

Maybe the room? Lots of hard surfaces emphasize the upper mids' power response and SL said something about how our ears don't hear diffuse sounds in the BBC dip range the same as they do the direct sound.

[fbov]

... well, I really have no clue, other than (apparently) the woofers are diffracting the tweeter response in a nasty way.

Javier,
One way to get at the apparent woofer correlation is to measure the tweeter mounted in a baffle plate without woofer holes (made of scrap?), or mounted on an even larger surface. That would change the diffraction variable significantly, and if the peak follows, you'd know you're on the right track. If the peak is unchanged, you know you need something more than diffraction fixes.

And thanks in advance for the Microlith XOs.

Frank

[subynube]

subynube,

Your problems sound exactly like mine. Shouty mids, forward sound, got used to them - until I built another pair of speakers which didn't have so many problems.

Do you have access to measuring equipment? I'd be interested in looking at the overall FR you have.

I decided to re-work (yet again) the crossover on my pair, and I added a tiny bit of a "BBC Dip" from 1 KHz to 2.5 KHz, or so. A very gentle -2 dB dip (I measured the speakers, and I found out their measurements weren't similar to the ones I took 1 1/2 years ago - perhaps the drivers finally broke in?).

The speakers sound better than ever. I'm working now on eliminating as much as possible the dip without getting that "forward" or "aggressive" sound quality, one step at a time.

I know far too many people have built this speaker for it to be badly designed. It's 99% more probable I did something wrong. I just can't see what. So far, I've ruled out the tweeter, woofers, the crossover, the baffle layout... basically everything.

Sorry it has been a while since I have gotten back to this. I do not have any measuring equipment. I wish I did, but dont for now. Everything I have noted was just my subjective opinion. I would love to know if anyone else has experienced this problem.

[fjhuerta]

Javier,
One way to get at the apparent woofer correlation is to measure the tweeter mounted in a baffle plate without woofer holes (made of scrap?), or mounted on an even larger surface. That would change the diffraction variable significantly, and if the peak follows, you'd know you're on the right track. If the peak is unchanged, you know you need something more than diffraction fixes.

And thanks in advance for the Microlith XOs.

Frank

I (finally!) posted the updated crossover.

[fjhuerta]

Sorry it has been a while since I have gotten back to this. I do not have any measuring equipment. I wish I did, but dont for now. Everything I have noted was just my subjective opinion. I would love to know if anyone else has experienced this problem.

Well, in my case, I *finally* found what the d**ned problem was.

I always focused on the tweeters. Surprise, the problem was a peak in the 1-1.6 KHz region on the woofers. I was so focused on the upper end I forgot the midrange.

I shaped that peak, eliminated almost all of the "BBC dip", and the speakers sound great now. NO midrange glare whatsoever. The RS-180's can really sing!

I still wonder why the measured FR of my design is so different from the reference one. I guess that's something I'll never know.

[Dave Bullet]

Hi Javier,

It reminds me of a post Zaph made some time ago, that he often looks to woofer behaviour / performance first before tweeter behaviour when there is forward anomaly in the upper midrange or treble (assuming your usual 2 way designs).

Can you post your crossover schematic here please? I'd be interested in how much further down in SPL the 1-1.6KHz range is on the woofer.

Can you take distortion measurements? Would be interested in how low the F2 - F5 are in that range.

Cheers,
David.