Why not lower RS180/225 RS28F XOs?

[Wushuliu]

Out of curiosity, since the 28F is able to go as low as 1khz, how come no one has attempted lower XO designs with the venerable RS drivers (given the breakup issues)? Would there be no benefit? The lowest design I have seen is Paul's Core 2-way with 1400, but why not go even lower?

[Mark65]

I'm with ya, if I had (or could afford) a pair of 28F's I'd absolutely be looking to plumb those depths.


Someone surely should look into it.:)



Mark

[Jay1]

I would like to see more designs that skip the RS mid.

[Deward Hastings]

Out of curiosity, since the 28F is able to go as low as 1khz, how come no one has attempted lower XO designsPhysics works against it. There are a number of instruments with a potentially loud fundamental down there . . . trumpet and saxaphone for example . . . and for all its charms the RS28 motor does have its limits. Granted with a LR4 electric crossover you'd be 6 dB down on the tweeter at the crossover frequency, but you'd still be pushing its excursion limits to get 100 dB of even remotely clean output at 1000 Hz.

http://www.baudline.com/erik/bass/xmaxer.html

[GranteedEV]

Out of curiosity, since the 28F is able to go as low as 1khz,

At what SPL and distance?

[Wushuliu]

I'm thinking of the Aviatrix which crosses at 1k, correct? If you mean my listening environment I listen at low-moderate levels 60-75db (if that) at 9ft.

[cjd]

With a waveguide, sure. There are a couple. I think MarkK's TM with the RS225 goes to 1200Hz. Well, it used the RS28A since that's what was available back then.

With the RS180, it's probably going to hurt more than help if you don't skimp on the crossover. 1500Hz is great, and a whole lot less stress on the tweeter, leaving you more headroom.

If you're not listening at high levels, sure. But that's a risky design to put out there. Too many people won't pay attention to that part.

The other thing is: why? Most of us design because something interests us. I know I have no interest in trying to re-do any existing designs just to try to push the tweeter a little lower. That alone simply isn't worth the investment.

You'll see some. But it won't be most.

C

[Wushuliu]

Thanks. Was just curious.

[Paul Carmody]

For me, it was a matter of how much I notched out the breakup. The RS180 needs a hefty notch filter. Unless I got the breakup down 50+ dB, I was still hearing it, hissing at me. However, once it was down that low, it just wasn't an issue anymore.

The Core and Amiga both use about the same amount of suppression on the breakup, even though one has a crossover point at 1400 Hz and the other is more like 2000 Hz.

[Pete Schumacher ®]

While we won't make InDIYana this year with such a design, we are planning on being at Dayton this year with a Dual RS225 / RS28 setup using an 8.5" waveguide on the RS28, which will allow a crossover around 900Hz. The guide boosts the RS28 output at 1000Hz over 10dB compared to the infinite baffle response. Once the XO is applied, contouring the response to LR4, the transfer function is over 16dB down at 1000Hz relative to the level at 14KHz. That will allow the system to play at pretty high SPL levels, cleanly, as verified by measurements.

Here's a plot of a two way using an 8" woofer and the RS28 in that waveguide, along with the THD. This was taken at 10W RMS. The THD is hovering well below 1% at 1000Hz, right where the crossover point is. The sound is very "unstrained" to say the least.

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Byzy%20Mini/latestxo.png

[Jim Holtz]

FWIW, Curt had his Avatrix playing in my living room the night before the last Iowa DIY event most of the evening. My 1st reaction to the design when Curt was telling me how low he'd crossed the RS28F was no way they'd stand any kind of volume. So, we cranked them up to very high (not concert) levels with out a whimper from the RS28F. They sounded very smooth and sweet all the way up, IMHO.

That listening session also convinced me that the RS28F had what I was looking for in a dome tweeter. Great sound quality and super flexibility. It could very well have been Curt's crossover expertise that made the difference but, IMHO, it certainly works well in that design.

My $.02 worth...

Jim

[Deward Hastings]

For me, it was a matter of how much I notched out the breakup. The RS180 needs a hefty notch filter.Yes . . . and a 4th order electric crossover. As do pretty much all metal cones. But it has an excellent motor, and if you take the necessary care to avoid the cone issues it is a world-class driver, and works beautifully with the RS28 in a two-way crossed at 1200-1400 Hz. I've compared that side-by-side with a W22/Millennium (both boxes active), and it really comes down only to which tweeter you prefer (with maybe a tiny nod to the bigger cone at the very bottom). I'd pick the RS box and pocket the price difference.

[fjhuerta]

Hey PE...

You should start selling a nice waveguide that fits the RS28s/A/F.

I'm expecting a big check for my awesome idea in the mail. ;)

[Deward Hastings]

a Dual RS225 / RS28 setup using an 8.5" waveguide on the RS28,Is that with the "a" or "f" dome? I've wondered how the domes would hold up to the increased pressure from horn loading . . .

[Pete Schumacher ®]

Is that with the "a" or "f" dome? I've wondered how the domes would hold up to the increased pressure from horn loading . . .

That plot was with the aluminum dome. But the F performed equally well once the XO was modified for the differences in response at the top end.

[Pete Schumacher ®]

Hey PE...

You should start selling a nice waveguide that fits the RS28s/A/F.

I'm expecting a big check for my awesome idea in the mail. ;)

Contact Ed Rosenquist (D, Rose here at PE). He'll cut you a pair of custom baffles in MDF with that WG built in, with whatever woofer hole you decide on. Here's the response of the RS28A on one of Ed's waveguide 10" x 17" baffles (pictured below).

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Project%20LOKI/RS28AonaxisrawWGfromEd.png

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Waveguide%20Mania/waveguidebaffles.jpg

[Wushuliu]

Thanks for the responses guys. Lots of interesting stuff. Here another question:

I've been wondering: if listening at low to moderate SPL (55-75db max), is it better to use woofers with lower moving mass (for a 2way)? Do heavier cones need more power and therefore perform better at higher volumes?

Basically I am trying to decide between Shawn A's Arpeggio's (Usher 8948a, 9950), Jay's W6-789e TM (which has the 6.5" 7g moving mass), and potentially Paul's Core 2-way (RS180). With my low listening levels I don't want to have to crank up the volume to get the best out of the drivers.

[Deward Hastings]

Do heavier cones need more power and therefore perform better at higher volumes?All else being equal a heavier (more mass) cone will result in lower sensitivity and thus "more power" required for the same output SPL. That has nothing to do with "performing better at higher volumes", though, and all else is never equal anyway. Cone mass is just one of the many variables one must weigh (har, har, har) in driver design, and it can pretty much never be singled out as a/the "critical factor" in final performance.

[MSaturn]

if you're not going to crank anything up, using a waveguide is probably more trouble than it's worth.

[GranteedEV]

if you're not going to crank anything up, using a waveguide is probably more trouble than it's worth.

Not sure how controlled polar response is "more trouble than it's worth" especially if we can get the waveguide pre-made. Even at 1khz an 8" woofer is getting directive.

[Wushuliu]

All else being equal a heavier (more mass) cone will result in lower sensitivity and thus "more power" required for the same output SPL. That has nothing to do with "performing better at higher volumes", though, and all else is never equal anyway. Cone mass is just one of the many variables one must weigh (har, har, har) in driver design, and it can pretty much never be singled out as a/the "critical factor" in final performance.

M'kay. Thanks Deward.

[Face]

if you're not going to crank anything up, using a waveguide is probably more trouble than it's worth.
That's the worst thing you've ever said...

[Pete Schumacher ®]

That's the worst thing you've ever said...

Yes indeed. The benefits of a WG are far more than simply a lower XO point, but the very uniform off-axis response it provides. Unless you're sitting in an anechoic chamber, optimizing off axis is always worth the extra trouble.

Here's the result of the system at 20, 45 and 60 degrees. Notice the complete lack of "bloom" in the off axis response above the crossover, typical of a 7" TM. Definitely "worth the trouble" even if you never turn it up. That's something even a totally active system cannot fix.

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Byzy%20Mini/20-45-60.png

[Wushuliu]

I keep waiting for an easy to build waveguide-based design (other than the sort-of ER18DXT and er, expensive Geddes) for those of us who do not have the tools to make our own (both design and waveguide itself). Jon Marsh's ER18/HDS comes closest but even that requires more tools than I have - not to mention I am still traumatized from building the Modula MTM crossover...:eek:

[Pete Schumacher ®]

I keep waiting for an easy to build waveguide-based design (other than the sort-of ER18DXT and er, expensive Geddes) for those of us who do not have the tools to make our own (both design and waveguide itself). Jon Marsh's ER18/HDS comes closest but even that requires more tools than I have - not to mention I am still traumatized from building the Modula MTM crossover...:eek:

Hopefully, your wait will be over before the summer. Ed Rosenquist will be building the L.O.K.I. cabinet after AK Fest, and we will post the design as soon as the XO voicing is complete. I know he wants to make the baffles available for DIY'ers. Everything else will be available from PE, the RS225-8 and the RS28. And because it'll be a dual 8" woofer design, house shaking bass performance should be available.

And I anticipate a fairly simple XO by comparison with the Modula.

[MSaturn]

eh, I don't see how it's drastically superior considering you're only using one characteristic of it.

[Pete Schumacher ®]

eh, I don't see how it's drastically superior considering you're only using one characteristic of it.

Please explain, as your statement above is almost as off base as the one you made earlier.

It's not just off axis performance that improves. A waveguide loaded version will perform with even lower distortion, measurably, than the same tweeter on a flat baffle. Not only are you improving the usable SPL, but you're improving in-room performance. Both are audible improvements, with superior in-room performance. Then, there's the near total elimination of diffraction ripple compared with the version on a flat baffle.

If you don't want to use a waveguide, then you live with the shortcomings. There's everything to gain using one, from any objective standpoint. The only thing you lose, is a little time crafting the guide.

[MSaturn]

Distortion at 75dB with decent drivers isn't even going to factor in.

Also, you increase CtC distance a great deal.

I'm tired of all the bandwagoning around here - RAAL, waveguide, whatever. Everything has tradeoffs, and your average user can't mill a waveguide, or measure to set up the complicated crossover.

[Whitneyville1]

Pete, can I make a wild leap of a conclusion? The Peerless 8" Nomex mid-woofer and a top-notch low distortion/low FS tweeter (probably in a waveguide) should make an excellent two-way speaker, all other things being equal. I'm not certain if the RS28F or A really fits as a "top-notch" tweeter or what does, in fact. I'm fairly sure it's something I can't afford. The break-up of the RS 180/225 up top doesn't make them look to me, to be the best options out there, without alot of "taming". Is this true of all metal coned mid-bass drivers, or is this a function of the cost compromises Dayton had to make?

[isaeagle4031]

ctc doesn't mean nearly as much when you can cross at 800Hz. Saturn have you ever heard a waveguide? The biggest drawback I can say is that you aren't going to be using them as a near field monitor of sorts do to size, but the imaging with a waveguide just beats the pants off a surface mounted tweet IMO

[isaeagle4031]

Any metal coned driver is going to ring at some point.

[MSaturn]

I'd argue an omni is way better.

[Wushuliu]

I'd argue an omni is way better.

But there are no omni designs around here either, except I think a complicated one one by Dan N. from a few years back and the GR Research kit.

@Pete: looking forward to that MTM!

[Wushuliu]

Pete, can I make a wild leap of a conclusion? The Peerless 8" Nomex mid-woofer and a top-notch low distortion/low FS tweeter (probably in a waveguide) should make an excellent two-way speaker, all other things being equal. I'm not certain if the RS28F or A really fits as a "top-notch" tweeter or what does, in fact. I'm fairly sure it's something I can't afford. The break-up of the RS 180/225 up top doesn't make them look to me, to be the best options out there, without alot of "taming". Is this true of all metal coned mid-bass drivers, or is this a function of the cost compromises Dayton had to make?

Consider the RS28 resale value. Whenever you're ready to get rid of them, there'll be a line waiting...

[Pete Schumacher ®]

Distortion at 75dB with decent drivers isn't even going to factor in.

Also, you increase CtC distance a great deal.

I'm tired of all the bandwagoning around here - RAAL, waveguide, whatever. Everything has tradeoffs, and your average user can't mill a waveguide, or measure to set up the complicated crossover.

Well, if you're sitting 1m from your speakers, then 75dB isn't tough.

If you're sitting 4m from your speakers (far more realistic), then CTC won't matter, but you won't be getting 75dB from your system at 1m either. Now, you'll need an extra pile of SPL to hit your 75dB at the listening position.

The "average guy" can acquire the WG from other enterprising individuals who CAN do it.

Now, let's talk about band wagons shall we? Active ain't all it's cracked up to be . . . the expense, the complexity, the extra amps, talk about something that the average guy won't be doing.

If you're tired, take a nap.

Your serve. :p

[Wushuliu]

Active ain't all it's cracked up to be . . . the expense, the complexity, the extra amps, talk about something that the average guy won't be doing.

If you're tired, take a nap.

Your serve. :p

Well you just woke up Deward, I'm sure! Incoming!

[Pete Schumacher ®]

Pete, can I make a wild leap of a conclusion? The Peerless 8" Nomex mid-woofer and a top-notch low distortion/low FS tweeter (probably in a waveguide) should make an excellent two-way speaker, all other things being equal. I'm not certain if the RS28F or A really fits as a "top-notch" tweeter or what does, in fact. I'm fairly sure it's something I can't afford. The break-up of the RS 180/225 up top doesn't make them look to me, to be the best options out there, without alot of "taming". Is this true of all metal coned mid-bass drivers, or is this a function of the cost compromises Dayton had to make?

The RS28, A or F, is objectively a top-notch tweeter. Smooth, extended response, very low distortion, excellent construction. There are a LOT of tweeters worse, and some of them even cost more. There are better dome tweeters, but they are more expensive. I doubt you could find a tweeter at the price point of the Dayton that would measure better. I'd consider a top shelf dome one that has a bit more sensitivity, like the Scan 6600 or their ring radiators.

Aluminum is a very inexpensive material to make a cone from. The process to make a stiff paper variant, with a curvilinear profile, requires more expensive machines. The problems with aluminum up high in frequency, also makes it a great material to use for bass reproduction; stiff and lightweight. They definitely take more work to tame the problems, but when the proper attention is paid to the shortcomings, the results are quite good.

I've not seen any metal cone drivers, other than the DA175, that don't have horrible breakup modes. The SEAS Excel have some really nasty resonances, but if tamed, also some of the best midrange available regardless of material. There's a reason the Orion uses the 8" Excel as a mid. But even that design could greatly benefit from a waveguide on the tweeter, to better match the pattern produced by the OB mid, as well as drive the XO point down half an octave or more, further from the breakup of the mid.

A guide gives you options that not using one doesn't. It allows the same tweeter to cross lower without strain, lowers distortion, allows higher peak SPL, and opens up the possibility of using much larger woofers. I'd have no qualms using the Eminence 3012LF with the waveguide loaded RS28 crossed at 1000Hz. I dont' think the Dayton would have a problem keeping up, but would fall apart well before the Eminence if the WG were not present.

[Pete Schumacher ®]

Well you just woke up Deward, I'm sure! Incoming!

Remember, we're talking "average guy" here as put forth by MSaturn. I wouldn't consider the "Orion" a kit for the average guy here at PE. In fact, a WG 2-way is far less complex, and far more achievable, by the average guy, than an active 3-way. Without measurements, both are a crap shoot.

That said, an active system using a WG would still be superior, to one without.

[Whitneyville1]

Thank you Pete. You told me a lot more than you thought you might have. I thought I understood the advantages to aluminum cones (or magnesium or even titanium) and thanks for confirming their drawbacks. It's plain that there's much more to making a musical speaker than simming drivers until the graphs look good. They help, but they aren't the only thing that matter, I guess.

[MSaturn]

well, I DO buy almost everything used to save money. for a "main" setup, I think active is worth the investment.

in that vein, perhaps waveguides are as well - certainly, an MTM with an RS28F and two RS225s would be pretty sweet, especially active (you could even try getting a "transient perfect" fifth-order crossover going in this case - if my assertion in another thread is correct ...)

[benchtester]

...
That said, an active system using a WG would still be superior, to one without.

I should have an active one this year (probably the HDS tweeter in the MCM waveguide). My experience and preference is for minimalist analog active circuits. I hope a WG will allow a simple first order crossover with gain adjustment. So another feature of the WG is that it can simplify an active circuit.

[Pete Schumacher ®]

I should have an active one this year (probably the HDS tweeter in the MCM waveguide). My experience and preference is for minimalist analog active circuits. I hope a WG will allow a simple first order crossover with gain adjustment. So another feature of the WG is that it can simplify an active circuit.

Depends on the guide and tweeter. Check Dave Pellegrine's WG thread. While a 1st order HP can be enough to produce a workable response, usually, you'll need a bit of EQ along with it.

[Pete Schumacher ®]

Thank you Pete. You told me a lot more than you thought you might have. I thought I understood the advantages to aluminum cones (or magnesium or even titanium) and thanks for confirming their drawbacks. It's plain that there's much more to making a musical speaker than simming drivers until the graphs look good. They help, but they aren't the only thing that matter, I guess.

Sims are valuable at getting a starting point. But in the end, the ears have to be happy with the result, despite what pretty pictures the mic is sending you. Linear and non linear distortion can tell you where a driver will be awful and to avoid using in that range, but response and distortion plots alone won't tell you how they actually sound in the final system. Ears are pretty good at telling you when something just ain't right once you've taken all the measurements and done your sims, and built a crossover. Voicing is that last critical step.

[Face]

Distortion at 75dB with decent drivers isn't even going to factor in.
It doesn't matter if the drivers are cheap or expensive, regardless there will be an improvement in low end extension and directivity if done right.

[Pete Schumacher ®]

well, I DO buy almost everything used to save money. for a "main" setup, I think active is worth the investment.

in that vein, perhaps waveguides are as well - certainly, an MTM with an RS28F and two RS225s would be pretty sweet, especially active (you could even try getting a "transient perfect" fifth-order crossover going in this case - if my assertion in another thread is correct ...)

I too am curious about Jeff's TP design. A waveguide practically forces a 4th order or higher slope. His comment about terrible lobing does raise the issue of power response, which is something to seriously consider when designing a speaker.

The Dual RS225 system is going to be a TWW 2.5-way. Ed Rosenquist is building the first one for his son I believe, and he prefers the look with the WG on top. And we all know, looks do affect the sound . . . ;)

[isaeagle4031]

Let's see 28a wave guided, dual 225s TLed, JBs TP x-over at 800Hz. Possible? Could be interesting.

[Pete Schumacher ®]

Let's see 28a wave guided, dual 225s TLed, JBs TP x-over at 800Hz. Possible? Could be interesting.

Not sure about the TL aspect, as it depends on the cabinet Ed is constructing. He's going for a curved sided style, coming to a point in the back. That might make it a bit challenging to get a line installed via internal baffle/divider.

[ryanbouma]

I keep waiting for an easy to build waveguide-based design (other than the sort-of ER18DXT and er, expensive Geddes) for those of us who do not have the tools to make our own (both design and waveguide itself).

http://techtalk.parts-express.com/showthread.php?t=229015

I'm not done documenting it, and I've since changed the cross over to cross at 950hz since I last posted in that thread. I took measurements and vertical nulls were at 15 degrees, so I lowered it. That and I did compression testing that said that tweeter had no problems crossing at 950hz. 4 components in the cross over. A single 1.2mH inductor. :)

I also have an MT version in the works.

That waveguide costs $10 :D

[GranteedEV]

I'd argue an omni is way better.

Omnis have too many early reflections. They're too room sensitive.

[Deward Hastings]

There's a reason the Orion uses the 8" Excel as a mid. But even that design could greatly benefit from a waveguide on the tweeter, to better match the pattern produced by the OB mid, as well as drive the XO point down half an octave or more, further from the breakup of the mid.ORION, in its original (single tweeter) incarnation, actually used the “tweeter bloom” to smooth power response. The abrupt (4th order) transition from dipole to uni-directional monopole was balanced by the tweeter’s side radiation in a reasonably “live” room, and the power response curve transitioned from “flat” to following that of the tweeter’s beaming. Adding a waveguide to the (single) tweeter would undo that, and be a step backward.

With the addition of the rear tweeter (found desirable for “imaging” and for power response balance in a more “dead” room) the tweeter “bloom” again became a problem (rather than a solution to a problem) . . . waveguides would be a solution, but their depth, and the resulting need to adjust for offset in all dimensions, adds a different set of complications (although a dipole radiator with matched waveguides on each side is an intriguing possibility). The M/T transition remains an “issue” for ORION . . . although there is some “dipole cancellation” between the tweeters it is less-than-complete, and they are more “omni” than dipole (or dipole like) in the tweeter’s lower range. This has been addressed to some extent in the revised ASP, but that isn’t an ideal solution. There’s work yet to be done . . .


I've not seen any metal cone drivers, other than the DA175, that don't have horrible breakup modes.The DA175 has “horrible breakup modes” too, as do “paper” and “plastic” cone drivers (typically at even lower frequencies), it’s just masked (if you look only at frequency response). The secondary effects (intermodulation and “spectral contamination”) are still there.


A guide gives you options that not using one doesn't. It allows the same tweeter to cross lower without strainSort of. It is easy to accept the name ("waveguide") as defining what's happening, but there is in fact very little pattern control from the "waveguide" at low frequency (it's typically just too small). In any case you don't get 10 dB of "gain" from the (relatively modest) pattern control. A better understanding comes from seeing them for what they are . . . horns . . . and recognizing that the "gain" comes from acoustic impedance transformation and better loading of the driver at the throat (and, perhaps, a slight improvement at the mouth). It's an interesting exercise in complementary tradeoffs and compromises (the reduced acoustic offset typical with a waveguide is another advantage) . . . and so far (when it's not pushed too far, and the throat is properly designed) the benefits seem to predominate.

[MSaturn]

My apparent hate for waveguides is driven more by a "what a listener will notice for their time invested" angle - while I agree lowering distortion and controlling power response are very important, a lot of people still go back and listen to their Advents and think it's more pleasant. I guess I'm just playing devil's advocate.

In a perfect world I'd probably have full-range dipoles (ribbons down to 2Khz or so, planar mid to 250, 10" SS drivers to 80Hz, 18" megadrivers to 16Hz...) and a whole load of room treatments. Controlled cardioid response is awesome.

[Deward Hastings]

Omnis have too many early reflections. They're too room sensitive.Many of us (self included) used to believe that, but a more complete understanding of what the ear is doing with those reflections (and how omni speakers actually sound in a real room) suggests otherwise. Much depends on the source material . . . what works best for close miced and console mixed "studio" recordings may not work (generally doesn't work) for ORTF capture in a large hall. And vice versa. While there are certainly exceptions, for many of the recordings in my collection "omni" is actually preferred, and it's an accpeptable compromise for most of the rest. I can't, actually, think of any recordings that I own where it actually sounds bad . . . (that don't sound bad with other polars as well ;)).

[Deward Hastings]

Controlled cardioid response is awesome.Difficult to implement full range . . . can't say I've ever heard it done. It would appear to be a good solution for small rooms, or placement on the long wall of larger rooms. Dipole gives you front wall reflections, cardioid (in theory) side wall reflections and omni all-wall reflections . . . what "works" is circumstantial, and a matter of taste.

[Pete Schumacher ®]

The DA175 has “horrible breakup modes” too, as do “paper” and “plastic” cone drivers (typically at even lower frequencies), it’s just masked (if you look only at frequency response). The secondary effects (intermodulation and “spectral contamination”) are still there.

While they are still there, the relative levels compared to other drivers are significantly less. My personal feeling is that the dustcap offers some attenuation of cone ringing.

http://www.zaphaudio.com/6.5test/DA175-8-FR.gif

Compared to the RS180, the difference is quite noticable.

http://www.zaphaudio.com/6.5test/RS180-FR.gif

And the SEAS L18 for some real contrast.

http://www.zaphaudio.com/6.5test/L18-FR.gif




Sort of. It is easy to accept the name ("waveguide") as defining what's happening, but there is in fact very little pattern control from the "waveguide" at low frequency (it's typically just too small). In any case you don't get 10 dB of "gain" from the (relatively modest) pattern control. A better understanding comes from seeing them for what they are . . . horns . . . and recognizing that the "gain" comes from acoustic impedance transformation and better loading of the driver at the throat (and, perhaps, a slight improvement at the mouth). It's an interesting exercise in complementary tradeoffs and compromises (the reduced acoustic offset typical with a waveguide is another advantage) . . . and so far (when it's not pushed too far, and the throat is properly designed) the benefits seem to predominate.

Toe may toe, Toe mah toe. Waveguide, horn, whatever. The fact remains that compared to a flat baffle implementation, the pattern control of the 8.5" guide I'm talking about offers over 10dB higher SPL for a given voltage at 1000Hz with the RS28. If there was significant acoustic loading acting as a compression chamber, it would be visible in impedance plots, and it's simply not there. What ever change in response is not due to acoustic loading at the diaphragm, but purely pattern modification in these guides. This is not a compression chamber/horn but simply a modified baffle . . . a waveguide.

Good call on the acoustic offset correction. One more in the plus column for waveguides.

[Deward Hastings]

The fact remains that compared to a flat baffle implementation, the pattern control of the 8.5" guide I'm talking about offers over 10dB higher SPL for a given voltage at 1000Hz Catagorically wrong, Pete . . . pattern gain is in direct proportion to pattern narrowing, and there is no 10dB worth of narrowing happening here. You need to re-evaluate your understanding of the mechanism(s) at work.

[Pete Schumacher ®]

Catagorically wrong, Pete . . . pattern gain is in direct proportion to pattern narrowing, and there is no 10dB worth of narrowing happening here. You need to re-evaluate your understanding of the mechanism(s) at work.

I'll stick to what the mic tells me, over what you're telling me.

There is pattern control to below 1000Hz on this guide Deward, verified by actual measurements on the baffle.

If you contend that the "gain" is not due to the pattern control, but due to acoustic loading, why do I not see the impact on the impedance plots?

[Deward Hastings]

I'll stick to what the mic tells me, The problem is not with (what little) "the mic tells you", it is with your mistaken interpretation and understanding of it.

[Face]

If you contend that the "gain" is not due to the pattern control, but due to acoustic loading, why do I not see the impact on the impedance plots?
You should see a change in impedance with a waveguide that size. Even one as small as the MCM makes a very noticeable difference.

[Pete Schumacher ®]

You should see a change in impedance with a waveguide that size. Even one as small as the MCM makes a very noticeable difference.

What would you consider "noticeable?" What percentage?

Below are with, and without WG, of the RS28.

At 2KHz, the impedance changes from 3.16, to 3.15 Ohm.

[Deward Hastings]

What would you consider "noticeable?" The terminal (electrical) impedance will be dominated by the static impedance (resistance and inductance) of the voice coil, followed by the dynamic impedance from the mass of the moving system and the compliance of the suspension and air cavity. The contribution of the change in radiation resistance will be very small (probably within the margin of error of your measurement) and easy to miss if your intention is to ignore it anyway . . .

[Pete Schumacher ®]

The terminal (electrical) impedance will be dominated by the static impedance (resistance and inductance) of the voice coil, followed by the dynamic impedance from the mass of the moving system and the compliance of the suspension and air cavity. The contribution of the change in radiation resistance will be very small (probably within the margin of error of your measurement) and easy to miss if your intention is to ignore it anyway . . .

Oh, so it went from "noticeable" to "not noticeable" when actual data is presented.

Sorry Deward. Noticeable should be more than .01 Ohm change out of 3.1 Ohm. Or do you consider a .3% change in impedance responsible for the 10dB higher output at 2KHz?

There's simply not that much extra acoustic loading presented by the guide, at least, not that can be seen in the impedance plot. Compared to horn loading on a woofer, which is OBVIOUS in the impedance plots, this is simply unnoticeable.

I still maintain that the guide is doing just that, "guiding" the side firing radiation from the dome, forward, instead of to the side and what would be wrapping around the tweeter. It's changed from radiating into 4pi, into a much smaller forward lobe.

[Face]

What would you consider "noticeable?" What percentage?

Below are with, and without WG, of the RS28.

At 2KHz, the impedance changes from 3.16, to 3.15 Ohm.
http://i.imgur.com/Uqzch.jpg

http://i.imgur.com/4Uc0F.jpg

[Pete Schumacher ®]

Well, there you go. Those are definitely noticeable changes. And similar changes are simply absent in the guide I'm using on the RS28.

To what would you attribute that lack of difference in my plots, Face?

[Face]

To what would you attribute that lack of difference in my plots, Face?
I have no idea Pete.

[Pete Schumacher ®]

I have no idea Pete.

At least that's an honest answer.

That's why I contend that the acoustic loading presented by most "horn" implementations simply isn't present with this particular guide. It's basically a 4" roundover, rotated around the tweeter center. Is that the real reason?? I'll answer the same . . . I don't know. All I do know is, what you'd expect to see with significant acoustic loading just doesn't show up in the impedance plots.

[Deward Hastings]

Oh, so it went from "noticeable" to "not noticeable" when actual data is presented.Now you're just making things up . . . show me where I said anything about "noticeable" or "not noticeable"?

We're talking simple loudspeaker acoustics here, and you're simply wrong in your interpretation of what's happening with your "waveguides".

[Pete Schumacher ®]

Now you're just making things up . . . show me where I said anything about "noticeable" or "not noticeable"?

We're talking simple loudspeaker acoustics here, and you're simply wrong in your interpretation of what's happening with your "waveguides".

See Face's plots for noticeable examples.

[Deward Hastings]

I have no idea Pete.It looks like you're posting measurements and Pete's posting "simulations", for a start. There may be differences in the throats, too . . . I'm a bit surprised at the magnitude of the difference you're measuring, it looks like a narrow throat horn.

[ryanbouma]

I'm with Pete on this. My impedance plots didn't show a change either. I haven't measured it with the larger Dayton 8" I'm using though, only smaller ones.

[Pete Schumacher ®]

It looks like you're posting measurements and Pete's posting "simulations", for a start. There may be differences in the throats, too . . . I'm a bit surprised at the magnitude of the difference you're measuring, it looks like a narrow throat horn.

Now, you're the one making things up Deward.

That data I posted was from a REAL RS28A that I have here, and a REAL waveguide, that also happens to be here. I just chose PCD to create the GIF to upload from the ACTUAL ZMA file created by DATS.

And HERE is the actual measured off axis response showing pattern control. XO point is 1KHz.

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Byzy%20Mini/20-45-60.png

[Pete Schumacher ®]

I'm with Pete on this. My impedance plots didn't show a change either. I haven't measured it with the larger Dayton 8" I'm using though, only smaller ones.

:)

[Deward Hastings]

Now, you're the one making things up Deward.You're such a lying sack, Pete . . . :p . . . the "question" wasn't of your frequency response measurements, but your impedance measurements. Re-posting a frequency response chart doesn't say anything to that.

And none of it goes to the question of whether the response changes are from radiation pattern (NOT) or acoustic impedance (loading) of the driver at the horn throat (probably, per Face). In fact polars actually showing the alleged radiation pattern "improvement" (flat baffle to same sized "waveguide" baffle) are conspicuously absent from your posts. All we get is "a miricle happens at 1000 Hz." that somehow gives a 10dB boost.

Not good enough . . .

[Pete Schumacher ®]

You're such a lying sack, Pete . . . :p . . . the "question" wasn't of your frequency response measurements, but your impedance measurements. Re-posting a frequency response chart doesn't say anything to that.

And none of it goes to the question of whether the response changes are from radiation pattern (NOT) or acoustic impedance (loading) of the driver at the horn throat (probably, per Face). In fact polars actually showing the alleged radiation pattern "improvement" (flat baffle to same sized "waveguide" baffle) are conspicuously absent from your posts. All we get is "a miricle happens at 1000 Hz." that somehow gives a 10dB boost.

Not good enough . . .

Deward, YOU were the one claiming that I "simulated" the impedance plots. You also stated that I wasn't getting pattern control.

Both your statements are demonstrably false.

Here's the raw response of the RS28A on the guide.

The flat baffle response meets this one at 14KHz.

The "throat" in this setup is the RS28 faceplate mated to the guide.

You can see the interface yourself in the picture at the bottom. Notice the RS28 faceplate clearly visible at the center. There is no throat to speak of, just a smooth transition to the baffle (guide).

Explain it then Deward. Why does Face's impedance plots show such obvious acoustic loading, and this does not? I'm not making this up Deward, just because you can't explain it.

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Project%20LOKI/RS28AonaxisrawWGfromEd.png

http://i12.photobucket.com/albums/a203/pete_schumacher/Speaker%20projects/Byzy%20Mini/Aurora2.jpg

[Pete Schumacher ®]

One more point to ponder, especially Deward.

Even a flat baffle creates diffraction with a tweeter. We can see large fluctuations in the response especially on the low end as wavelength approaches the baffle dimensions. This guide is using that same diffraction to steer the response, smooth it, and direct it. It's not relying on a compression chamber to increase acoustic loading, or we'd see it in the impedance plots.

I really think this is pure diffraction control.

[Pete Schumacher ®]

Dave Pellegrine was nice enough to post his measured impedance files of the OX20, with and without his guides. He found a 2 Ohm change on the impedance peak, so I opened up my ZMA files and looked for something similar . . .

Fs stayed the same at 645Hz, both before and after the guide. And impedance changed from 7.58 Ohm without waveguide, to 7.52 with it.

What you make of that Dewie?

[GranteedEV]

Even a flat baffle creates diffraction with a tweeter.

Does an infinite flat baffle?

[Deward Hastings]

Does an infinite flat baffle?The real "apples to apples" comparisons (and measurements) are not (if they are being made at all) being posted here. What we're getting are exagerations, misrepresentation and mis-direction worthy of a marketing brochure for some "super speaker-cable".

At least the RAAL fad seems to be over . . .

[Pete Schumacher ®]

Does an infinite flat baffle?

Nope. But that's only because the edges are well beyond the tweeter's low frequency operation.

But the WG is not infinite, nor is an 8.5" wide baffle. Both have an effect on the tweeter response based on the geometry of the baffle. In the case of this guide, with no throat to speak of, it's basically a large roundover, but moving forward of the dome plane instead of back from it. The smooth contour smooths response ripples, just like a large radius roundover that moves away from the tweeter.

At least that's my theory, unless someone comes up with something better.

[MSaturn]

raal in a waveguide, anyone?

[GranteedEV]

Nope. But that's only because the edges are well beyond the tweeter's low frequency operation.

But the WG is not infinite, nor is an 8.5" wide baffle. Both have an effect on the tweeter response based on the geometry of the baffle. In the case of this guide, with no throat to speak of, it's basically a large roundover, but moving forward of the dome plane instead of back from it. The smooth contour smooths response ripples, just like a large radius roundover that moves away from the tweeter.

At least that's my theory, unless someone comes up with something better.

I guess I just don't follow... even with a waveguide, a baffle still has edges that create diffraction...my.. er..."contention?" if-you-will, is that Geddes stresses the value of larger edge roundovers, even in cases where he uses a 15" WG:

http://www.audiocircle.com/circles/galleryimage.php/albuu56/white_sealed.sized.jpg

Is he just... wasting his time..? His protege, Duke Lejeune, also seems to go for huge bevels despite 12" WGs:

http://gallery.audioasylum.com/cgi/gi.mpl?u=2112&f=PlanetariumBeta.jpg

Sorry if what I'm trying to say isn't clear, (at least it feels that way to me). It just seems that you feel waveguide = no need to worry about baffle edge diffraction (of course, others feel that there's no need to worry about it, period IE Linkwitz i think). I'm just trying to wrap my head around it. Either it matters or it doesn't, which is fine, but "it dissapears if you use a waveguide" is something I find hard to follow.. I guess to put it "concisely" - is ALL diffraction really visible on the frequency response chart?

Or is it because a dome tweeter, for example, is already going to be a good ~16db down in "flat baffle level" at these frequencies that it won't excite the baffle diffraction after a crossover, relative to the listening level? One thing I've read Geddes say, is that his speakers minimize diffraction, and because of this, they're also very sensitive to diffraction from nearby objects. If the waveguide totally removed the possibility of diffraction, I can't imagine this to be a case.

Or is it just that we don't care about diffraction at lower frequencies? Then why does Geddes, whose every philosophy is rooted in some AES research or another.

I'm not trying be argumentative, so don't take my post as being agressive or anything.

[GranteedEV]

raal in a waveguide, anyone?

Only if you can pony up the money for the long ribbon (http://www.diysoundgroup.com/waveguides/seos/seosraal.html). The short ribbon is really best used with a small format mid driver that can play high, and the OEM ribbon is really hard to get as it is.

[Face]

Only if you can pony up the money for the long ribbon (http://www.diysoundgroup.com/waveguides/seos/seosraal.html). The short ribbon is really best used with a small format mid driver that can play high, and the OEM ribbon is really hard to get as it is.
If someone could cut me a break on a pair, I can take measurements. Pretty hard to justify that expense without any measurements to back it up.

[Pete Schumacher ®]

I guess I just don't follow... even with a waveguide, a baffle still has edges that create diffraction...my.. er..."contention?" if-you-will, is that Geddes stresses the value of larger edge roundovers, even in cases where he uses a 15" WG:



Is he just... wasting his time..? His protege, Duke Lejeune, also seems to go for huge bevels despite 12" WGs:



Sorry if what I'm trying to say isn't clear, (at least it feels that way to me). It just seems that you feel waveguide = no need to worry about baffle edge diffraction (of course, others feel that there's no need to worry about it, period IE Linkwitz i think). I'm just trying to wrap my head around it. Either it matters or it doesn't, which is fine, but "it dissapears if you use a waveguide" is something I find hard to follow.. I guess to put it "concisely" - is ALL diffraction really visible on the frequency response chart?

Or is it because a dome tweeter, for example, is already going to be a good ~16db down in "flat baffle level" at these frequencies that it won't excite the baffle diffraction after a crossover, relative to the listening level? One thing I've read Geddes say, is that his speakers minimize diffraction, and because of this, they're also very sensitive to diffraction from nearby objects. If the waveguide totally removed the possibility of diffraction, I can't imagine this to be a case.

Or is it just that we don't care about diffraction at lower frequencies? Then why does Geddes, whose every philosophy is rooted in some AES research or another.

I'm not trying be argumentative, so don't take my post as being agressive or anything.

You're missing the point I was trying to make. I'm not saying you don't need roundovers on the baffle edges. I'm saying that the response modification produced by the guide is a kind of diffraction control, not acoustic loading like you see with a compression driver. In the case of this guide and tweeter combo, we don't see the impedance change that acoustic loading produces. What I'm saying is that the response shaping is because of controlled diffraction.

Look at what an 8" wide baffle does to a dome tweeter response. There's a big peak and dip in the 800-2000Hz range, caused by the geometry of the baffle. Change that geometry, and the response also changes. Start flaring that flat baffle forward, and you start redirecting what energy was moving along the baffle forward as well. That can't help but modify the response compared to the flat baffle. Is that changing the acoustic load on the dome? You'd think it would show up in the impedance plot if that were the case. Maybe it is. But does holding a reflector up next to the baffle change the acoustic load appreciably? It would have a major impact on the response, just like you mention above with Geddes keeping things away from the guide.

[Pete Schumacher ®]

The real "apples to apples" comparisons (and measurements) are not (if they are being made at all) being posted here. What we're getting are exagerations, misrepresentation and mis-direction worthy of a marketing brochure for some "super speaker-cable".

You don't seem to know measurements if they hit you in the face.

I don't know what the heck you're asking for that I haven't already posted, here and in other threads. I posted the impedance plots, and you accused me of doing a sim. You said that there is no pattern control, so I show you measurements that show there is. You say there's no 10dB "gain" over a flat baffle and I post the plot that shows more than 10dB.

How about this?

https://encrypted-tbn2.google.com/images?q=tbn:ANd9GcTp4OJyfbqVevHfyu5Zcs3cUyBylDq28 Hesa8vls1hGUrFZA-Rd

[Pete Schumacher ®]

If someone could cut me a break on a pair, I can take measurements. Pretty hard to justify that expense without any measurements to back it up.

Alex at RAAL says they've tried doing a waveguide, but he was unhappy with the resulting time domain response that it created. Of course, that's just the owner's opinion, and perhaps with the right guide . . . .

[Deward Hastings]

You say there's no 10dB "gain" over a flat baffle and I post the plot that shows more than 10dB.No, you didn't. No polars, either. And no explanation of how "diffraction" (without beaming) supposedly causes gain.

But the picture of your b*tt plug was at least funny . . .

[Pete Schumacher ®]

No, you didn't. No polars, either. And no explanation of how "diffraction" (without beaming) supposedly causes gain.

But the picture of your b*tt plug was at least funny . . .

No Dewie, that's your little whiner plug.

Sorry if the data posted is too difficult for you to wrap your head around. If you dispute my data, that's your unsubstantiated prerogative I suppose. You think I made up the impedance plots? You think that raw RS28 in guide plot a few posts back, I simply drew up in photoshop? Or those off axis plots of the completed system were also just the result of simulations?

Seriously, what is your problem tonight? Didn't get your nap today?

This guide was first worked on over a year ago, with DanP sending a total of 8 different shapes and depths for me to use with the RS28.

If you're too lazy to look it up, well, that's not my problem. I certainly don't have to justify the measured results to you. I know what I've measured and others have seen the data as well. You'll just have to deal with it and move on.

But you calling me a liar, is really weak.

By the way . . . when are you going to post something original?

Or are you just going to sit in the bushes lobbing grenades?

[ryanbouma]

Even though there is probably some commercial interest in Pete's work, my findings are in agreement.

First here is a "waveguide" that is 6.5" diameter. Impedance before and after.

http://i874.photobucket.com/albums/ab306/tuxedocivic/TN%20Allen%20waveguides/SBbareimpedance.png

http://i874.photobucket.com/albums/ab306/tuxedocivic/TN%20Allen%20waveguides/SBWGimpedance.png

Here is a frequency response.

http://i874.photobucket.com/albums/ab306/tuxedocivic/TN%20Allen%20waveguides/SBA0to30.jpg

*Note my top octave isn't that accurate, I had some mic problems a few months ago. I've changed my setup quite a bit since and have worked out the kinks.*

And here it is against a flush mounted baffle (no edge treatment).

http://i874.photobucket.com/albums/ab306/tuxedocivic/TN%20Allen%20waveguides/SBflushAand3.png

I have similar results for an even smaller guide and for the large Dayton 8". All I don't have for those is the impedance. The Dayton 8" gave the SB29 around 10 or 12db worth of gain. All my filtering indicates the impedance didn't change much, compared to the free air impedance measurement used in PCD. Maybe some.

My thought process also aligns with Pete's that the radiating sound is being directed forward towards the microphone. At high frequencies there is no radiated sound, just directional sound. At low frequencies it's very omni, so it gets pushed forward towards the listener / microphone.

I don't see what's so hard to understand about these measurements. What would it take to convince people these things work. Sure, they're hard to make. The only draw back I can see.

[Pete Schumacher ®]

Don't be surprised to see Deward claim that your last plot there was fabricated, and that 10 dB isn't 10dB. :rolleyes:

[critofur]

On axis level appears to be increased by 10db, the limited dispersion seems superior [in this application] in that it more closely matches the woofers he's using. Distortion measurements look very good, at, what to most, seems a fairly high level. Looks good to me, if the listener does not hear any unwanted "waveguide sound" (which may be often due to sighted bias?) which one of my friends dislikes very much, he basically won't listen seriously to any speakers with a waveguide...

The problem is not with (what little) "the mic tells you", it is with your mistaken interpretation and understanding of it.
Instead of telling someone they don't understand something, why don't you explain (in a nice way) what the "correct" understanding is, if there is a practical difference...

*EDIT* OMG - both of you kids need a "time out"! The post I'm expecting to read next is: "meet me at the playground after school, I'm gonna beat you up!" If this keeps up for years..., I'm gonna see in the news: "fight at the nursing home, two old guys argued about speakers then one pushed the other out of his wheelchair"


If someone could cut me a break on a pair, I can take measurements. Pretty hard to justify that expense without any measurements to back it up.

Expensive + afraid to let people post measurements publicly is a suspicious combination to me, one that I would avoid... Reminds me of the diamond tweeters that "Zaph" measured years ago...

[Pete Schumacher ®]

On axis level appears to be increased by 10db, the limited dispersion seems superior [in this application] in that it more closely matches the woofers he's using. Distortion measurements look very good, at, what to most, seems a fairly high level. Looks good to me, if the listener does not hear any unwanted "waveguide sound" (which may be often due to sighted bias?) which one of my friends dislikes very much, he basically won't listen seriously to any speakers with a waveguide...

Instead of telling someone they don't understand something, why don't you explain (in a nice way) what the "correct" understanding is, if there is a practical difference...

*EDIT* OMG - both of you kids need a "time out"! The post I'm expecting to read next is: "meet me at the playground after school, I'm gonna beat you up!" If this keeps up for years..., I'm gonna see in the news: "fight at the nursing home, two old guys argued about speakers then one pushed the other out of his wheelchair"



Expensive + afraid to let people post measurements publicly is a suspicious combination to me, one that I would avoid... Reminds me of the diamond tweeters that "Zaph" measured years ago...

Yeah, I got a little bent out of shape being called a "lying sack" by Deward, accusing me of hyping and fabricating data. Pretty uncalled for, and my response to his accusation was, well, pretty poor.

At least you can see the data for what it is. Deward on the other hand, who knows what his issues with it are.

So, what are your thoughts on the actions of the guide? Is it due to acoustic loading, even though there's no throat or upper frequency "gain" as well as no modification to driver impedance? Or is it due to "steering" the wavefronts forward, a type of diffraction control, where the pattern is pushed forward as the tweeter becomes more omni?