dumptruck is finally building ewaves for real

Just whatever, I'm not really a fancy amp kinda guy. A Rotel and a Sony ES at the moment, both a little over 100W.

Sounds good, probably comparable to what I've had on my setup so far, Yamaha M45 and Kenwood KM-X1000, both ~125W. I've replaced the Yamaha by a small tube amp a while ago, sounds good as well.

I'm trying to work out an all-analog active Xo based on op-amps that replicates what I have on the DSP right now, I'll see where this goes. I started kinda blindly, but most of what I want is already outlined on S. Linkwitz's webpage, which has been tremendous help.

az
 
Last edited:
I started kinda blindly, but most of what I want is already outlined on S. Linkwitz's webpage, which has been tremendous help.

az

Siegfried's the man. He's all about the full mathematical modeling of his circuit's. He headed up a team that made one of the first hp spectrum analyzer's @ hp labs (Palo Alto). His designs are very well thought out. He, also, is very methodical in his approach to validating designs. I wish we had an army of speaker designers like these. I think that the state of the art would move more quickly.

I've heard a colleague of his, Brian [J.] Elliot, own dipole designs. Subs [dipole] too. I think the best speaker's that I've ever heard and I worked at a high end store selling Wilson, Aerial, B&W, Thiel, etc.

Anyway, kind of a sidetrack. Back to the E-Waves thread.
 
Siegfried's the man. He's all about the full mathematical modeling of his circuit's. He headed up a team that made one of the first hp spectrum analyzer's @ hp labs (Palo Alto). His designs are very well thought out. He, also, is very methodical in his approach to validating designs. I wish we had an army of speaker designers like these. I think that the state of the art would move more quickly.

The equations and transfer functions for the various active crossover building blocks are pretty well outlined on there. I've worked out most everything to replicate the functions I currently implement via miniDSP, mostly using SPICE: LR24 high-pass and low-pass slopes, a shelving filter for BSC and a notch/EQ filter for CD-compensation. I'll also need to delay the woofer by ~0.12ms, if nothing changes. This will likely require two or three all-pass stages, one of which I should make adjustable to nail things down and account for unforseen variations.

az
 
I've heard a colleague of his, Brian [J.] Elliot, own dipole designs. Subs [dipole] too. I think the best speaker's that I've ever heard and I worked at a high end store selling Wilson, Aerial, B&W, Thiel, etc.

Sorry to take it back off track, but can you point me in the direction of some more info on these, Jon?

- Michael
 
Sorry to take it back off track, but can you point me in the direction of some more info on these, Jon?

- Michael

Hi Michael, The last time that I worked with Brian was about 10 yrs ago. He was starting to partner up with Chips Davis who, if you didn't know, is kind of a listening space Guru. Brian (Elliott) never was interested in marketing his technologies/designs and, quite frankly, has no skills or interest in running a company. That guy's interest is really on the research side and just when you think a design is frozen, he'll say, "Hey I've got an idea, let's try this.". That's the last thing you want to hear when there's a deadline.

I have not heard Siegfried's Open Baffle design's (Orion, Pluto, etc), but Brian's were easily the best speaker's that I have heard. Extremely low distortion and very dynamic at the same time. The subs were phenomenal because you'd feel it in your body, but without the "congested" sound that higher distortion designs bring.

Brian doesn't really discuss his designs and seems to just do one-off builds for wealthy clients. He lives in Los Altos Hills, but I think you're best bet is to read about Brian and Siegfried's design concepts on Siegfried's website. Siegfried looks to be very meticulous to a point that he could have made an awesome accountant. Judging what I've seen on Siegfried's site, there are more similarities, than differences in these two guy's designs and it looks like that they either shared info or just came to similar conclusion about open baffle designs.

I know that Siegfried mentions Brian on his site, but I don't remember how to navigate to get there. The topic to search for is "H-Frame subwoofer".

Brian also borrowed speakers from Thiel and NHT to evaluate and measure. He really like some 12" inch speakers from NHT at that time, but said, "NHT was going to stop making them". I don't know if Ken (Kantor) was there at that time or if Brian was dealing with Ken.

You could also look for an AES paper that Brian and Siegfried wrote about open baffle design. I think it's a pretty old paper. Maybe from '92??

Open baffle are great, but the design is critical and you will need an eq to compensate for the bass cancellation that occurs with an H-frame design, but the results are spectacular, if done well.
 
I found some links that might prove interesting:

http://www.zoominfo.com/p/Don-Barringer/172523901
http://www.6moons.com/industryfeatures/linkwitz2/linkwitz.html
http://www.stereophile.com/content/siegfried-linkwitz-page-5
http://www.linkwitzlab.com/AES'89/AES'89.htm
http://www.aes.org/e-lib/browse.cfm?elib=16879

It's kind of funny: In a previous post, I had mentioned that Brian and Siegfried had presented a white paper about loudspeaker design. As I tried to find that, I saw a third name on the paper..... Floyd Toole. So it has that going for it too.
 
Hi Michael, The last time that I worked with Brian was about 10 yrs ago. He was starting to partner up with Chips Davis who, if you didn't know, is kind of a listening space Guru. Brian (Elliott) never was interested in marketing his technologies/designs and, quite frankly, has no skills or interest in running a company. That guy's interest is really on the research side and just when you think a design is frozen, he'll say, "Hey I've got an idea, let's try this.". That's the last thing you want to hear when there's a deadline.

I have not heard Siegfried's Open Baffle design's (Orion, Pluto, etc), but Brian's were easily the best speaker's that I have heard. Extremely low distortion and very dynamic at the same time. The subs were phenomenal because you'd feel it in your body, but without the "congested" sound that higher distortion designs bring.

Brian doesn't really discuss his designs and seems to just do one-off builds for wealthy clients. He lives in Los Altos Hills, but I think you're best bet is to read about Brian and Siegfried's design concepts on Siegfried's website. Siegfried looks to be very meticulous to a point that he could have made an awesome accountant. Judging what I've seen on Siegfried's site, there are more similarities, than differences in these two guy's designs and it looks like that they either shared info or just came to similar conclusion about open baffle designs.

I know that Siegfried mentions Brian on his site, but I don't remember how to navigate to get there. The topic to search for is "H-Frame subwoofer".

Brian also borrowed speakers from Thiel and NHT to evaluate and measure. He really like some 12" inch speakers from NHT at that time, but said, "NHT was going to stop making them". I don't know if Ken (Kantor) was there at that time or if Brian was dealing with Ken.

You could also look for an AES paper that Brian and Siegfried wrote about open baffle design. I think it's a pretty old paper. Maybe from '92??

Open baffle are great, but the design is critical and you will need an eq to compensate for the bass cancellation that occurs with an H-frame design, but the results are spectacular, if done well.


Thanks Jon. My first DIY attempt was a 3-way OB design that utilized H-frame subs for the subwoofer section. I love the H-frame subs...extremely musical and accurate/quick and really appreciate the sonic qualities of OB designs in general. I'm scouting for my next DIY build and find myself leaning toward something along the lines of the NaO Note II RS or a spinoff of the Orion. Always looking for more designs/ideas to consider so thank you for the links and info.

Sorry for the threadjack, dump. I think you once suggested the NaO Note to me, come to think of it. :scratch2:
 
Not too worried about threadjacking this far down the road. Wouldn't it be Note II RS vs LX521?

Admittedly, I haven't studied the 521 as much as I have the Orion and NaO Note II RS. I picked up the IRS Betas around the time I became familiar with the 521 so I haven't read up on it in great detail as I kind of lost interest in another DIY once the Beta project fell in my lap. But absolutely...that design would be a consideration as well. Have you ever heard any of the three, dump?
 
Not a one, but my understanding is that the 521 pretty much makes the Orion obsolete. It is also extremely similar to the Note II RS in design (but the voicing might differ a lot, don't know).
 
Man, I can't find anybody who has heard the Note II RS. There's just that one testimonial floating around from some audio meet up in the northeast where the listener claims he liked it better than the Orion. At just over a grand to build that's a pretty big leap of faith....for me, at least. It's been a while since I visited the Note II RS thread over at DIYaudio....need to go revisit and see what's up with that design these days.
 
True, but it's a lot less than the LX521, which is approaching $3k, and the $800-$1000 xover can't be used for anything else whereas the Note II uses miniDSP that you can adjust or repurposes as you see fit (or you can use a PC-based xover, too).

edit: I was totally wrong (or outdated, to be fair) about this as far as DSP options, detailed correction several posts ahead.
 
Last edited:
True, but there is so much positive press and name rec with Linkwitz's designs I'd actually stand a good chance recapturing a good chunk of my investment if I went to sell them. My gut tells me that would not be the case with Note II RS. Don't ya think? Hell, I may not be able to sell any of it here in Texas! Probably just have to part them out. :sigh:
 
True, but it's a lot less than the LX521, which is approaching $3k, and the $800-$1000 xover can't be used for anything else whereas the Note II uses miniDSP that you can adjust or repurposes as you see fit (or you can use a PC-based xover, too).

Eww. That part kinda sucks. At the end of the day I always seem to come back to the Note II RS. And that design is approaching, what, a few years old now? Surely updates/new version are near.
 
Heh, well actually...
John K. said:
And at this point the design has been finalized for almost a year. I'm not planning on any driver changes. I'm not planning any new designs either.
 
Chilling out at home on a very cold night with some potent seasonal beers, doing some audio experiments that I'm sure I'll have to redo, I came to look at a graph in this thread and thought I'd respond to some stale comments here:
True, but it's a lot less than the LX521, which is approaching $3k, and the $800-$1000 xover can't be used for anything else whereas the Note II uses miniDSP that you can adjust or repurposes as you see fit (or you can use a PC-based xover, too).
As I mentioned to Tom in PM recently, I found out I was wrong about this; thanks to some motivated DIYers, the LX521 has extensive options for more affordable, non-specific DSP crossovers that were meticulously designed for miniDSP and Linux software, approved (and even tested) by Siegfried himself. I take it back :eek:. Some really nice work done on these software solutions (google LX521 DSP, I'm too lazy for links).

As for the Geddes approach, seems like his round WG's with generous mouth radii would have the same issues as the TD15M/QSC ctc spacing problem. Might this be why he seems not to produce his 15" Summa anymore? I can't say I've followed all of this closely. I guess the trade-off is not so bad for an earlier onset of constant-directivity.
IIRC (lord help me if I don't), there was some sort of price increase needed where he felt they already cost more than people interested were willing to pay, and he didn't like building them very much. The speaker worked just fine, though.

The squashed type of CD guides limit crossover frequency with their vertical size to some extent, and maybe more importantly they make crossover design difficult as you approach their limits. Geddes' big ones using lower xovers, he hasn't put out vertical data, only descriptions, but I wish he would.

I have come to think I understand that not only would they be only a little worse than competitors as far as lobe spacing and magnitude, but they would be better inside and outside the nulls, which I think may actually be more important. Also, they simply have to be much better on all other axes (there are infinitely more than two, remember). That's what I meant by "beginning to side with Geddes' take on this type of 2-way". Don't get me wrong, I'd sign up for another QSC waveguide speaker tomorrow - just talking ideals here.

Well-produced jazz recordings have a way of sounding better than most other music I think.z
Yes, for various reasons they lend themselves to normal stereo hifi very well. I am kind of embarrassed about making that reference. I made it because I had just been listening to some Three Blind Mice jazz records on these speakers and meant it as a counterpoint to my "sterile" comment, because they sounded quite alive and real. I do NOT use such records for critical evaluation, I use them for maximum appreciation once I'm done. For critical listening, I start with various pop/rock records I'm familiar with (no particular quality needed) to make sure I know where I'm at with bass/midbass, and then I use various fancy speech recordings, symphony recordings (I need a better library of excellent ones there, keep meaning to research that), and the EBU SQAM test CD that I'm in love with.

Side note on that SQAM CD/set: It's horribly boring, but SO well made and helpful if you are familiar with hearing common acoustic instruments. Actually, even if you aren't, the simple test of various instruments playing a scale or other pattern (it has a lot of those) can reveal gross errors in speakers and rooms that are commonly described in vague audiophile nonsense terms.

I really think there are so many speakers that people stretch their vocabulary to describe when it could be as simple as "when the clarinet hits C and C#, everything goes to hell", and that kind of thing can be easily located in measurements, and addressed. It's also free (google). If anyone has similar test sources, I'd love to hear about them.
 
The squashed type of CD guides limit crossover frequency with their vertical size to some extent, and maybe more importantly they make crossover design difficult as you approach their limits. Geddes' big ones using lower xovers, he hasn't put out vertical data, only descriptions, but I wish he would.

I have come to think I understand that not only would they be only a little worse than competitors as far as lobe spacing and magnitude, but they would be better inside and outside the nulls, which I think may actually be more important. Also, they simply have to be much better on all other axes (there are infinitely more than two, remember). That's what I meant by "beginning to side with Geddes' take on this type of 2-way". Don't get me wrong, I'd sign up for another QSC waveguide speaker tomorrow - just talking ideals here.

My last iteration on the QSC is symmetrical LR24 at 1250Hz. It seems to be working well enough on horizontal measurements, but I never did take verticals much. I never even considered attempting the "other axes". I have a real hard time getting gating to be below ~1500Hz and lifting/rotating these 110lb boxes to take polars is a pain in its own. I did have consistent ~500-600Hz gating in another room for my smaller Peerless/JBL-PT build though.

So I understand your own latest crossover was a mix of miniDSP and passive - which parts are what exactly? What about delay, how much do you have on which driver? I think our builds are close enough in the that we both have rear-mounted TD15M and front-mounted QSC, yours flush-mounted though. 18mm ply separates then on my box. Aside from that, the CD used might be the main difference. My BMS might have a deeper apparent acoustic center because of its annular diaphragm and re-entrant (?) phaseplug. With the 1250Hz LR24 slopes and a 2500Hz/-10dB/Q=0.5 notch used for CD compensation, I need 0.16ms delay on the bass driver. That ~180° phase shift I get through the Xover region appears to be something normal for LR24 slopes.

I completed the design for my active analog Xover, which just duplicates what I implement via miniDSP, and have my BOM and everything, but chickened out after having filled my Digikey shopping cart. :) Cost was higher than I cared to spend on that at the moment, especially since it'd just replace the miniDSP, which I don't have an issue with in itself, just the bit loss due to upstream or local attenuation. The design is not readily adjustable and since I'm not 100% certain about my data (high gating, no verticals) I'll wait it out, assuming I come back to the idea... I'll just be using a 4-gang ALPS Blue Velvet pot (buffered) at the input of my incoming ICEpower boards.

az
 
No it's all DSP currently, the hybrid one was passive parts combined with an analog LR4 active. Well, except I do have passive attenuation on the HF, but that's because of my continual troubles getting noise low enough with any amps.

The woofer is ~0.3ms ahead in raw response, but there is no delay currently because after the phase shift from driver EQ and the LR4 filters, everything seems to work out perfectly. I have not totally verified that, though, because I want to re-do the woofer EQ and there's no point in messing with it before that as long as they are summing pretty well on the straight forward axis, and they are.
 
I'm soon to travel this road with the same combination of drivers. One "plan B" may be to use the L'cleach 600 with a BA750...that to lower the crossover point.

Thanks for your contributions...I follow in your footsteps.
 
Back
Top Bottom