Distortion Measurements (Part 1)- Seas W15CY vs Accuton C79

[Jed]

Seas W15CY measured in box at 90DB/1meter with mic 4" from cone.

Image not available

Accuton C79 measured in box at 90DB/1meter with mic 4" from cone.

Image not available

Some of you may know that I am working a 3-way using the Aurasound NS10 driver, an Accuton C79, and Vifa D26 tweeter. I've been looking for a midrange that can outperform the C79 or possibly offer an alternative midrange bandpass section. However, the Accuton C79 seems to do things that no other stiff cone driver can do, such as maintain low odd order harmonics beyond 1.5k. Notice how low the 5th order harmonic distortion is in comparison to the W15CY.

The W15CY seems to be the champion below 200hz though, and it is very very close up to 1.5k. The question then becomes, what are my design goals and if I want to stick with the Vifa D26, will that really work with a W15CY?

I think the answer is no, since the Vifa D26 distortion really starts to rise below 2k.

Image not available

That makes a very robust tweeter, like the Seas 27tdfc-g, an absolute mandate. The W15CY distortion profile suggests to me that a crossover of 1500-1700hz would be ideal.

The C79 should be crossed over ideally at 2k, since the 3rd harmonic distortion starts to get elevated by 3k, but it still isn't nearly as high as the W15CY 3rd order spike at nearly the same frequency. I would bet that the C79 can be used with a ribbon tweeter, such as the Fountek CD3.0 crossed around 3-3.5K. However, currently, I'm using the D26 with the C79 crossed at 2.25k and the sound is superb.

So, I just thought I would share some of my thoughts as I look more closely at these drivers. The W15CY looks very good indeed, just seems a bit more limiting as a midrange.

[Johnloudb]

I would bet that the C79 can be used with a ribbon tweeter, such as the Fountek CD3.0 crossed around 3-3.5K.

The C79 really does look good. You know you want a ribbon oke: Jed, give in already.

[capslock]

Try simulating current drive by using something like a 30 to 100 Ohm resistor to drive the W15. This should have a huge effect on distortion, but you'll have to deal with the fundamental resonance which will have higher Q.

[jkrutke]

Hi Jed,

Our W15 plots look exactly the same. Here's mine:



This is a brand new one with a rubber boot and 8.2 kHz breakup. You have Arta set to percentage, but the response curve of the W15 is a straight line in it's useful range so it looks like my level based plot.

I've done 2 systems using the W15, soon to be 3. The first one used an older one with a 6.5kHz breakup, HD sweep:



Here's the system HD plot of the first one, which had some research done to see how low I could cross. (note: further mic distance to get proper summing between drivers, but more room is in the plot)

Image not available

This is a 1900 Hz LR4 crossover to a T25 Excel tweeter. The newer W15 seemed to work with 1900 Hz LR4 too. It completely controls the 3rd order peak and takes down a good portion of the 5th.

You could probably make a good guess at the best crossover point but to be sure, you could put a filter on both the woofer and tweeter and run an HD sweep. I used the D26 down to 2kHz once. You'd be treading new ground to push it lower. (let us know)

For the record, the W15 is currently my favorite 5" driver and my current reference system has 2 of them per side with a Peerless 810921, soon to be replaced with a DXT waveguide.

Is that Accuton this one? "C79-6 3" dome, 88.5dB (C90-6-79)" Looks good. Their descriptions are confusing and at Madisound I think it might link to the wrong PDF file. Looks good though. How is the response curve? The only Accuton I've ever tested was the C23 tweeter and it was only good. (for that price I want it to wash the dishes too)

[Jed]

Is that Accuton this one? "C79-6 3" dome, 88.5dB (C90-6-79)" Looks good. Their descriptions are confusing and at Madisound I think it might link to the wrong PDF file. Looks good though. How is the response curve? The only Accuton I've ever tested was the C23 tweeter and it was only good. (for that price I want it to wash the dishes too)

Hi John,

The C79 I have is the "3" inch dome with the black ears. It's not really a 3" dome though. More like a 4.5" dome.

The FR is not nearly as flat as the W15CY and that makes it difficult to work with, but I've finally nailed the crossover for a flat response.

Here is the C79 response in a box (Note the quick rolloff below 200hz which is in line with the increase in distortion). This makes it difficult to hit a 2nd order acoustic slope, so I had to use a 1st order electrical solution for that, even when crossing at 400hz. I'm not thrilled about that because now the C79 has to work harder in a bandwidth of increased distortion. I could use 3rd order or 4th order slopes, but I like the sound of 2nd order.

Image not available

This is the response in room of my 3-way with the Aurasound NS10, C79, and D26 with a bit of smoothing:

Image not available

A bit of a dip at 175hz, but I hope that is room related because I can't get rid of it and it disappears in the nearfield when I measure the NS10 with crossover in place. I wonder if it has something to do with the woofer positioned rather low on the baffle, which is centered at about 18" and crossed over LR2 with a Fc at 400hz.

[Jed]

..... soon to be replaced with a DXT waveguide.

I emailed madisound to see when they will get those in and I didn't hear back from them. They probably get a million emails like that every day so I can understand, but do you know an ETA for them?

By the way, my test of the W15CY is the current version with the boot as well.

[jkrutke]

I heard from Seas that there was a delay because one of their part suppliers messed something up. Madisound recently mentioned that they heard from Seas that they are now ready to start production. It should be soon but I don't have a real ETA unfortunately.

I'm guessing that 175 Hz dip is a reflection, more likely a side wall than the floor if the woofer is mounted low. (4-5 feet away?) The dip at 400 might be the floor.

[Rick Craig]

I think the distortion measurements have been given too high of a priority in selecting drivers. If you look at the research done by Sean Olive and Earl Geddes you'll find that there are other more important factors to consider. That's not to say that distortion is something to totally ignore but it certainly should not be at the top of the list.

[jkrutke]

I think the distortion measurements have been given too high of a priority in selecting drivers. If you look at the research done by Sean Olive and Earl Geddes you'll find that there are other more important factors to consider. That's not to say that distortion is something to totally ignore but it certainly should not be at the top of the list.

When it comes to selecting drivers, IMHO distortion is the absolutely most important thing. Equal parts linear and non-linear of all types are all I look for. People tend to emphasize non-linear distortion simply because it can't be fixed in the crossover. There's all the basic measurements, and there's also Klippel results and level dependent results that I always look at when available. All that's left is looks and cost. Important to some, but pretty minor to most.

Now I'm going to assume you meant to say "system design" and not "selecting drivers." Big difference, and individual types of distortion, Like harmonic or IMD may not be the most important item, but they are near the top of the list. Of course there's lots of other things to consider under the umbrella of system design. Geddes is always hyping up directivity... just a form of off axis linear distortion that's a consideration in system design.

All IMHO. Your specific comment is going to be a tough sell with this crowd. Feels like the start of a flame war, so I'm going to flee like the animals run from the burning forest.

[Jed]

My thread is purely objective in nature and I'd like to keep it that way if possible.

Yes there are other very important factors to consider in speaker design, but I'll consider them when the next part of the design process calls for it.

I just thought people might find it interesting to see some of the decisions and comparisons that are made that go into designing a loudspeaker. I will be designing a "no compromise" 3 way speaker based on my goal of using a driver in its optimum passband with the lowest system distortion as possible.

The W15CY is a definite contender for use in my system. I like what I see from this driver.

The next phase is to take more measurements to determine how it will work with my selected woofer and tweeter. I take one variable at a time and distortion measurements help determine the outcome of many decisions that follow.

[Rick Craig]

My thread is purely objective in nature and I'd like to keep it that way if possible.

Yes there are other very important factors to consider in speaker design, but I'll consider them when the next part of the design process calls for it.

I just thought people might find it interesting to see some of the decisions and comparisons that are made that go into designing a loudspeaker. I will be designing a "no compromise" 3 way speaker based on my goal of using a driver in its optimum passband with the lowest system distortion as possible.

The W15CY is a definite contender for use in my system. I like what I see from this driver.

The next phase is to take more measurements to determine how it will work with my selected woofer and tweeter. I take one variable at a time and distortion measurements help determine the outcome of many decisions that follow.

Jed - Sorry, I didn't mean to derail the thread. Both are fine drivers and in my opinion you can't lose with either one. The C79's main limitation is excursion so you have to pay attention to the high pass filter and slopes but I'm sure you already realize that. I've used both drivers with a ribbon tweeter and had good results. With the W15 it becomes a little more tricky because of the strong peak in the upper end so the crossover points available with the ribbon become critical. If that's an issue then the C79 will cross a little higher than the W15.

[Jed]

I was playing around with some topologies to hit an LR2 slopes on the HP AND LP sections of the crossover for the C79 midrange and I had some success.

Image not available

Considering my box for this upcoming design will use a sloped baffle,

Image not available

I don't think it will be much of an issue to get the driver centers aligned to better integrate the phase to avoid cancellations in the frequency response. Since the C79 seems somewhat happy playing up top, I figured I'd give it a try. A lot of people really like LR2 slopes, and I must say I like it too, but I haven't really heard a fully optimized design using solely LR2 slopes.

One issue with moving on to this next design phase is my original selected tweeter (Vifa D26) won't work very well if crossing this low and if I try to hit an LR2 slope at 2.4k. So I'll need to look into another tweeter.

Augerpro inspired me with his measurements of the Scan Speak 9800 tweeter. It has a Very extended response and should be able to hit an LR2 response easily down to 2.5k. Maybe I could go lower but I'll have to test that out for myself when my SS9800 tweeters arrive this Friday. The nice thing is if the SS9800 doesn't work out with the C79, it'll also do fine LR4 @1800 with the Seas W15CY.

[Dennis H]

Heck Jed, if it's that easy to hit a 2nd order bandpass with the mid, it looks like a natural for a Duelund XO.

[JonW]

I just thought people might find it interesting to see some of the decisions and comparisons that are made that go into designing a loudspeaker.

Indeed! :T Thanks for posting the thought process. So I was wondering about a few things...

I see that you’re using the Aurasound NS10. Why not go for the larger NS12? Won’t that give a little more extension?

From your plot it looks like you might be able to get down almost flat to 20 Hz. Do you think that’s possible? I thought that was pretty tough to do with a single 10”. (It’s a TMW, right?)

Are there any other bass/sub drivers you considered for the low end? There may be a 3 way in my future and it’s not as easy to find third part measurements on the bass drivers as you can with tweeters and mid. So it’s not so easy to pick the best drivers for the bottom.

Why did you start with the D26 as your tweeter. If you’re going for the ‘no compromise’ approach, aren’t there better ones out there?

Just curious about your thinking. Thanks.

-Jon

[Paul Ebert]

I would bet that the C79 can be used with a ribbon tweeter, such as the Fountek CD3.0 crossed around 3-3.5K.

Jed,

Have you tried the C79 with the Visaton MHT (if you still have 'em)? Might be a nice combo.

Paul

[Jed]

Jed,

Have you tried the C79 with the Visaton MHT (if you still have 'em)? Might be a nice combo.

Paul

Yup, tried it but the MHT12 drops over really fast around 3.5k and I didn't have much success getting a flat response between 2-4k as the result. Some might have liked the BBC dip but I prefer a flatter response. The Founteks have more low end than the MHT12. MHT12s work well with the Dayton RS52 and W4 1337.

Image not available

Heck Jed, if it's that easy to hit a 2nd order bandpass with the mid, it looks like a natural for a Duelund XO.

Hi Dennis. I played around with a crossover Duelund style using the D26, C79, and NS10s. I'm not sure if it is exact but the slopes on the mid are flat delay 2nd order and on the tweeter/woofer are flat delay 4th order. Only 24 components in the crossover to do it. And that's without a zobel on the tweeter.

Image not available

[Jed]

....

I see that you’re using the Aurasound NS10. Why not go for the larger NS12? Won’t that give a little more extension?

From your plot it looks like you might be able to get down almost flat to 20 Hz. Do you think that’s possible? I thought that was pretty tough to do with a single 10”. (It’s a TMW, right?)

Are there any other bass/sub drivers you considered for the low end? There may be a 3 way in my future and it’s not as easy to find third part measurements on the bass drivers as you can with tweeters and mid. So it’s not so easy to pick the best drivers for the bottom.

Why did you start with the D26 as your tweeter. If you’re going for the ‘no compromise’ approach, aren’t there better ones out there?

Just curious about your thinking. Thanks.

-Jon

Hi Jon,

The NS12 looks really good but it also needs a much larger box. So you might get that little extra on the bottom end but you don't get something for nothing. I've got the single NS10 in about 60L and the bottom end is pretty amazing. I'd say yeah, in room you'll get bass in the 20s as you see in my measurements. I'm using a bottom mounted port JonMarsh style.

Other 8-10" bass drivers I've used or heard are the Seas W22EX, Dayton RS225, Seas L22, Usher 8995a, and now I'm using the Aurasound NS10. I bet you can bet which one I like best, though the 10"er has an unfair advantage.

As for the Vifa D26, it is pretty hard to beat at any price. You have to spend a lot to get better performance, and above 2.5k there are very few tweeters that do.

Visaton Ke25sc at 90db

Image not available

Versus Vifa D26 at same level and same conditions:

Image not available

[Johnloudb]

Hi Dennis. I played around with a crossover Duelund style using the D26, C79, and NS10s. I'm not sure if it is exact but the slopes on the mid are flat delay 2nd order and on the tweeter/woofer are flat delay 4th order. Only 24 components in the crossover to do it. And that's without a zobel on the tweeter.

That response looks great, are you still using that crossover? I guess, if you're going to try the W15 you'll be starting over?

Edit: Ok, I get what you mean now. That's a the simulated response, not the real crossover yet.

[Jed]

Edit: Ok, I get what you mean now. That's a the simulated response, not the real crossover yet.

I'll have to go and read more about the Duelund crossover advantages as posted by JonMarsh. I wonder if it is better sounding than 2nd order all around.

And you are correct about the Duelund not getting constructed yet in my design. Dennis mentioned the concept and I thought I'd give it a try to see how many parts it would take with a circuit model.

Jed

[Dennis H]

I'll have to go and read more about the Duelund crossover advantages as posted by JonMarsh. I wonder if it is better sounding than 2nd order all around.

Probably pretty similar. Both have a total of 360 degrees of system phase wrap from low to high. The Duelund keeps the woofer and tweeter in phase at all frequencies for a more 'perfect' theoretical solution but one or the other is so many dB down where they overlap that it probably doesn't mean much in the real world.

[Rick Craig]

When it comes to selecting drivers, IMHO distortion is the absolutely most important thing. Equal parts linear and non-linear of all types are all I look for. People tend to emphasize non-linear distortion simply because it can't be fixed in the crossover. There's all the basic measurements, and there's also Klippel results and level dependent results that I always look at when available. All that's left is looks and cost. Important to some, but pretty minor to most.

Now I'm going to assume you meant to say "system design" and not "selecting drivers." Big difference, and individual types of distortion, Like harmonic or IMD may not be the most important item, but they are near the top of the list. Of course there's lots of other things to consider under the umbrella of system design. Geddes is always hyping up directivity... just a form of off axis linear distortion that's a consideration in system design.

All IMHO. Your specific comment is going to be a tough sell with this crowd. Feels like the start of a flame war, so I'm going to flee like the animals run from the burning forest.

No intent to start a flame war - is it possible to have a civil discussion? I would hope that different opinions are tolerated here.

[Jed]

No intent to start a flame war - is it possible to have a civil discussion? I would hope that different opinions are tolerated here.

Rick, of course your comments are welcome, but it might be more fitting if you posted a new thread about "why distortion measurements don't matter as much as . . . " or a thread about room interaction, directivity, or xyz as it connects to your theory or opinions about that specific topic.

This thread is about driver measurements and how design decisions can be made as the result of them. Specifically the C79 versus the W15CY.

That said, and since you have worked with the C79, I have a few specific questions.

What crossover frequencies did you select and why?
Do you see the rising distortion of the C79 below 200hz as a potential limiting factor?
What tweeters have you used with the C79 and which did you like best?
Do you have a specific crossover slope you liked best with the C79?

I have the same questions for the W15CY.

Regards,

Jed

[Nemophyle]

btw i think i never seen/read a flame war on htguide ? i'm not old around here though

[jkrutke]

btw i think i never seen/read a flame war on htguide ? i'm not old around here though

Nah, that's actually why I come here. Thomas and crew keep it under control, and a lot of the stuff that goes on with other forums doesn't happen here. I also like it here because it's a more objective, scientific crowd.

Rick: no worries, your opinion is fine. If you start another thread, I'll participate if I have time. Please be verbose, as just saying "distortion measurements" is pretty broad. For example, I could list 20+ types of distortion measurements (some of which I don't have the equipment to do) and list how much I value each one and why. Some are more important than others.

Jed: you're rekindling my interest in Accuton drivers. That does look good, and 2nd order could be sweet. Madisound links this from the C79-6 3" dome, which looks wrong. Interestingly, the C23 I tested had no tall order breakup peaks in the HD sweeps even though there was a pretty strong breakup at 30kHz.

[Jed]

Jed: you're rekindling my interest in Accuton drivers. That does look good, and 2nd order could be sweet. Madisound links this from the C79-6 3" dome, which looks wrong. Interestingly, the C23 I tested had no tall order breakup peaks in the HD sweeps even though there was a pretty strong breakup at 30kHz.

Wow, just checked madisound and they raised the pricing of the C79 3.5" dome unit. It's now $340!!! Yikes. I don't think a lot of the Accuton drivers test as well as the C79 though. I've seen some average/poor harmonic distortion of the C82 and C92 in Klang and Ton or Linkwitz's site.

Anyway, here is a real pdf of the C79: http://www.clofis.nl/nl/thiel/C79-6_kompleet2.pdf

I've got some ScanSpeak 9800 tweeters coming that should hit a nice LR2 response with the C79. I'll post more when I test them and try them out.

[JonW]

Hi Jed,

Thanks for the thoughts. That all makes sense. Good luck with the project- it looks like a fun one.

Hi Jon,

The NS12 looks really good but it also needs a much larger box. So you might get that little extra on the bottom end but you don't get something for nothing. I've got the single NS10 in about 60L and the bottom end is pretty amazing. I'd say yeah, in room you'll get bass in the 20s as you see in my measurements. I'm using a bottom mounted port JonMarsh style.

Other 8-10" bass drivers I've used or heard are the Seas W22EX, Dayton RS225, Seas L22, Usher 8995a, and now I'm using the Aurasound NS10. I bet you can bet which one I like best, though the 10"er has an unfair advantage.

As for the Vifa D26, it is pretty hard to beat at any price. You have to spend a lot to get better performance, and above 2.5k there are very few tweeters that do.

[jkrutke]

Wow, just checked madisound and they raised the pricing of the C79 3.5" dome unit. It's now $340!!!

I suppose cutting one in half so I can see the motor is out of the question. :B

[Jonasz]

For the record, the W15 is currently my favorite 5" driver and my current reference system has 2 of them per side with a Peerless 810921, soon to be replaced with a DXT waveguide.

OT, John, would you like to elaborate a little bit about your reference design? I bet there's a lot of people interested in your designchoices, how you select the drivers and so on. Even if noone builds it it would be a great learning experience for a lot of people like myself. :B Maybe a topic for your blog? :W

[Jed]

I suppose cutting one in half so I can see the motor is out of the question. :B

I'd feel sorry for even a little NT1.

[JonW]

Jed,

I’m back to wondering about the thought process behind your driver choices. I was thinking a little... and I know I’m not supposed to do that.

About your woofer... I’m also planning a 3 way project (once my MT’s are done) and trying to pick the best woofer I can get my hands on. It’s tough to find distortion measurements comparing various woofers. I can’t find measurements out there like, say, Zaph comparing all the tweeters and mids. (Thanks Zaph!) Jon Marsh did compare a few here:

https://www.htguide.com/forum/showthread.php?p=242114&highlight=Aurasound#post2 42114

It looks as though the Dayton RS315HF tends to have generally lower distortion than the Aurasound 12”. So that would make me lean toward the Dayton over the Aurasound. I’m not sure if the same correlation would hold for the 10” analogues you’re more interested in- the Aurasound NS10 and the Dayton RS265.

It looks like all the Aurasounds are 4 ohms, as is the Dayton RS265. But the Dayton RS315 is available in 8 ohms. Perhaps that’s useful, although I have not yet modeled a 3 way speaker. For my MT’s, not dipping to 4 ohms has been a nice thing.

I see that Aurasound has 2 versions of their 10” and also 2 versions of their 12” drivers. There are the cheap (~$150) and the expensive (~$500) versions. No idea about what the differences are, especially with regards to distortion. The more expensive versions have distortion plots in the .pdf’s at Madisound. But conveniently, the cheaper versions do not have the distortion data provided. So there’s no way to compare.

I see the benefits of the D26 tweeter. It looks good in many regards. And I’ve got a project in mind for a high end MT set of desktop speakers. Right now, the D26 would be my top choice, both because it’s a good tweeter and because it’s got a nice, small flange that would be super for a very small speaker. (Although flush mounting looks to be a little difficult.)

Interesting that this Acuton does look more promising than many of their other drivers.

[Jed]

Hey Jon.

After testing various woofers, it should be noted that JonMarsh will use the NS12 for his upcoming speakers. Probably because it has an amazing flat FR and remains pistonic so high in frequency with decreasing harmonic distortion. I noticed the same trend in harmonic distortion tests with my NS10s. The Dayton does look good too, but I'll have to take a closer look at those plots.

I've seen some plots of the Dayton RS270 versus the NS10 and the NS10 looked like the better choice to me. Not to mention the box volume needed for the NS10 is much smaller than the Dayton RS270. Haven't seen much about the RSS10" woofers though.

[Johnloudb]

I moved this to a new thread, Here ... this could get out of hand. :B

Jed, I don't mean to hijack your thread here. But you're discussing woofers. I'm building a sub for my dad. He's thinking of getting a single $2400 Earthquake sub. Our room has bad room modes in one direction and needs two subs IMO. He likes detailed, fast bass.

I have these two sub enclosures, 4.5cf (127L) each. I'm modifying a stereo pair of old Nelson/Reed subs. They were dipole and have four 12" woofers each. I'm going to make them sealed. I'm just trying to decide on a the woofers. Two woofers per box. Any opinion on these drivers, for low distortion and clean bass??

Madisound touts it's own 12" driver as more musical. Could only use one per box though.

MAD1259

Both of these 12" would work, does 12" epic have lower distortion? Worth $1200 for 4?

Epic 12" (dual 4ohm version)

Dayton 12" HF 4ohm

I can do better than one Earthquake sub, right???

[Dennis H]

I can do better than one Earthquake sub, right???

Uh, for $2400, you betcha.

[Rick Craig]

Jed,

I’m back to wondering about the thought process behind your driver choices. I was thinking a little... and I know I’m not supposed to do that.

About your woofer... I’m also planning a 3 way project (once my MT’s are done) and trying to pick the best woofer I can get my hands on. It’s tough to find distortion measurements comparing various woofers. I can’t find measurements out there like, say, Zaph comparing all the tweeters and mids. (Thanks Zaph!) Jon Marsh did compare a few here:

https://www.htguide.com/forum/showthread.php?p=242114&highlight=Aurasound#post2 42114

It looks as though the Dayton RS315HF tends to have generally lower distortion than the Aurasound 12”. So that would make me lean toward the Dayton over the Aurasound. I’m not sure if the same correlation would hold for the 10” analogues you’re more interested in- the Aurasound NS10 and the Dayton RS265.

It looks like all the Aurasounds are 4 ohms, as is the Dayton RS265. But the Dayton RS315 is available in 8 ohms. Perhaps that’s useful, although I have not yet modeled a 3 way speaker. For my MT’s, not dipping to 4 ohms has been a nice thing.

I see that Aurasound has 2 versions of their 10” and also 2 versions of their 12” drivers. There are the cheap (~$150) and the expensive (~$500) versions. No idea about what the differences are, especially with regards to distortion. The more expensive versions have distortion plots in the .pdf’s at Madisound. But conveniently, the cheaper versions do not have the distortion data provided. So there’s no way to compare.

I see the benefits of the D26 tweeter. It looks good in many regards. And I’ve got a project in mind for a high end MT set of desktop speakers. Right now, the D26 would be my top choice, both because it’s a good tweeter and because it’s got a nice, small flange that would be super for a very small speaker. (Although flush mounting looks to be a little difficult.)

Interesting that this Acuton does look more promising than many of their other drivers.

One thing to consider is that the Dayton drivers tend to be suspension-limited on the Klippel tests so in that respect I think the Aura woofers are better. The RS180's were also missing the shorting rings in a recent test so quality control is also something to keep in mind.

The more expensive Aura versions have different motors and greater excursion capability. I recently built some subwoofers with the 15" and am currently working on a design with two of the ultra 10" drivers per side in a floorstander.

[jkrutke]

The RS180's were also missing the shorting rings in a recent test so quality control is also something to keep in mind.

I yanked this out of an old blog entry for you:

Here's a long blog entry about a small type of measurement: the Le(X) curve. In Voice Coil magazine, which is a periodical for the loudspeaker industry, the Dayton RS180 and RS125 were tested. The tests primarily consisted of impedance, frequency response curves and a bunch of Klippel plots. The response curves were done in boxes rather than on an infinite baffle, and important non-linear tests such as harmonic distortion and IMD were nowhere to be found. As such, these tests are of limited use. The Klippel plots however are great information. The plots were relatively decent compared to other Klippel plots I've seen, with the exception of one particular measurement of the RS180: the Le(X) curve. This curve is an inductance vs excursion chart. An ideal plot would be a flat line, from it's -Xmax position to it's resting point to it's +Xmax position. The RS180 varied from .45mH with the coil in to .20mH with the coil out. This caused Vance Dickason to comment "Both drivers are supposed to have shorting rings, but the RS180S Le(X) curve clearly shows this not to be the case."

I am correcting Vance. The RS180 does have a shorting ring. But the presence of a shorting ring only plays a small part in the Le(X) curve. From the Le(X) curve alone, I don't think you can definitively say whether a copper ring exists. The ring's location, proximity to the voice coil and the shape of the pole piece play bigger parts.

Vance was wrong, and improperly reported the lack of a shorting ring by reading the Klippel curves wrong. In some cases, you can't tell if there's a ring there unless the driver is torn apart.

The Dayton RS series woofers have some of the best quality control I've come across.

[Jed]

Thanks for clearing that up John. I was going to say something like, "can you back that up with evidence...." but now we have it.

Vance was wrong, and improperly reported the lack of a shorting ring by reading the Klippel curves wrong. In some cases, you can't tell if there's a ring there unless the driver is torn apart.

The Dayton RS series woofers have some of the best quality control I've come across.

Regards,

Jed

[jkrutke]

You can kinda see what makes the RS180 Klippel plots look like they do. The phase plug, being aluminum and above the gap, works as a short circuit far better than the lower ring, which is not quite as close to the voice coil. As such, inductance still lowers when the coil is in the outward position... much like a driver with a basic motor and no copper at all. The impedance curve and low Le numbers are a much better indicator that there is some copper placed in effective locations. My manual Le(x) tests (shown in the tidbits section under Le(X)...) show impedance curves that look pretty close to the the .20 and .45 Le numbers represented by the Klippel tests at Voice Coil. So, there's a little more here than just me saying Vance was wrong.

Interestingly, the RS180 would have a flatter Le(x) curve without a phase plug. But, it would also have higher non-linear distortion, higher overall Le, and possibly more motor noise due to dustcap pressures.

[Rick Craig]

I yanked this out of an old blog entry for you:



Vance was wrong, and improperly reported the lack of a shorting ring by reading the Klippel curves wrong. In some cases, you can't tell if there's a ring there unless the driver is torn apart.

The Dayton RS series woofers have some of the best quality control I've come across.

To my knowledge he never retracted the statement; furthermore, I believe he has quite a bit of experience interpreting Klippel results. Not that I think the RS180 is a bad driver - I still use it myself.

[Rick Craig]

Rick, of course your comments are welcome, but it might be more fitting if you posted a new thread about "why distortion measurements don't matter as much as . . . " or a thread about room interaction, directivity, or xyz as it connects to your theory or opinions about that specific topic.

This thread is about driver measurements and how design decisions can be made as the result of them. Specifically the C79 versus the W15CY.

That said, and since you have worked with the C79, I have a few specific questions.

What crossover frequencies did you select and why?
Do you see the rising distortion of the C79 below 200hz as a potential limiting factor?
What tweeters have you used with the C79 and which did you like best?
Do you have a specific crossover slope you liked best with the C79?

I have the same questions for the W15CY.

Regards,

Jed

I've used the C79 in a couple different designs, crossover points typically 400-500hz and 3-4K. The tweeters have been ribbons or the Accuton C12. I prefer ribbons myself but a good dome also works well. L/R 24db acoustic slopes were implemented. I didn't find distortion to be a problem with the box Q ,slopes, and frequencies I used. In that regard the W15CY will be better if you need a lower crossover point of say 250-300hz. The size of the subenclosure is critical with the C79 due to the minimal excursion available.

With the W15 a highpass of 250hz or higher @12db+ slopes is acceptable. Most any dome can integrate well with the W15 as well as a few ribbon and planar tweeters. The upper end can be handled with high order slopes or a combination using a parallel notch filter. I've read where some people have said that it should be crossed at 2K or below due to distortion tests; however, I've not found higher crossover points (2-3K) to be a problem.

The main area of differing opinions is how much the upper end peaks should be reduced. Sean Olive of Harman International did some controlled listening tests focusing on this issue for their Infinity and Revel speakers. If I can find the exact paper I'll post it here because I think it will give some insight about this issue and maybe the audibility of driver distortions in general.

[Jed]

Good info Rick and I'd love to see that paper if you can find it. Thanks,

Jed

[Jed]

I think I'm going to next test the C79 harmonic distortion at 90-95DBS from 100hz-300hz and see how much sound makes it through the 2nd order acoustic filter I've implemented.

I've modeled both LR4 and LR2 crossovers and the LR2 sure looks nice, but if the LR4 is more pratical or a necessity to limit excursion of the C79 I'll probably end up with that. I'll report back in the post in a bit with my findings.

Edit:

Image not available

This is with the woofer disconnected and the midrange and tweeter playing the test tones at about 90db @ 1 meter.

My passive crossover contains a series cap to hit an LR2 acoustic slope on the C79. The manufacturer says that it can do LR2 down to 200hz, but I really don't think that is the case at all. This distortion measurement reflects a 400hz Fc LR2 acoustic slope, and you can see that harmonic distortion rises steadily below the Fc. This suggests to me that I should investigate LR4 slopes if I want to crossover at 400hz to the Aurasound and minimize any colorations to the sound or investigate the Seas W15CY. For now I'm going to keep trying to use the C79 and hit LR4 targets. However I don't think I'm going to go as high in FR as Rick did, because of the C79's rise in 3rd order harmonic distortion at 2.5k.

When my ScanSpeak 9800s get here on Wednesday. I'll do some comparisons of it versus the D26 with LR4 Fcs of 2k-2.5k. My feeling right now is an LR4 bandpass of 400hz and 2K is going to be "it."

We'll see what developes. Thanks all for your input.

[augerpro]

Hi Jed. Anything keeping you from staying with LR2 and just raising Fc?

BTW I bet you'll really like the SS9800. Very nice tweeter, just wish it didn't cost so much. I'll be interested to hear your impressions versus some of the other tweeters you've had.

[Jonasz]

BTW I bet you'll really like the SS9800. Very nice tweeter, just wish it didn't cost so much. I'll be interested to hear your impressions versus some of the other tweeters you've had.

Is the SS9800 considered to be a better tweeter than the RS28? Looking at the price it is but in reality?

[Jed]

Is the SS9800 considered to be a better tweeter than the RS28? Looking at the price it is but in reality?

The 9800 has lower overall distortion from the tests I've seen from augerpro and JohnK, and it has a much more extended response. For example the RS28a is flat to 1k, where the 9800 is flat to 600hz. Below 2k, 3rd order harmonics rise quickly for the RS28a, where the 9800 only has 2nd order harmonics rise and very satisfactory 3rd, 4th and 5th order harmonics from 1k on up. I suppose I should buy a pair of RS28a but I'm really not interested in it given the observations above. Not to mention the RS28a has a 5db dip at 14-16k and the 9800 does not.

From a subjective standpoint, TacoD here on the board has heard both the 9800 and RS28a, but prefers the 9800.

So is the 9800 worth an extra $100? I guess it depends on whether the above "advantages" can be worked into your system's design/goals.

Jed

[Jed]

Hi Jed. Anything keeping you from staying with LR2 and just raising Fc?

BTW I bet you'll really like the SS9800. Very nice tweeter, just wish it didn't cost so much. I'll be interested to hear your impressions versus some of the other tweeters you've had.

I have a funny feeling I screwed something up with that harmonic distortion measurement I posted above because the distortion is expressed as a percentage. So, for example, although the distortion may be higher below 100hz, the output is low enough that it would be off the radar. LR2 at 400 hz might still be ok. So I'll check the total system harmonic distortion with the woofer playing to compare to the distortion of just the mid and tweeter hooked up that I posted above. I may see that the percentage shifts to an acceptable level with the summed NS10 response in the system.

So I haven't given up on LR2 :-)