New 3-way coming up.

[Jed]

I'm not sure if there is a way to make it label correctly in software, but why don't you plot in relative dB, and find what your reference is with a Radio Shack SPL meter (referenced in its accurate range). Then since it is in dB, it's much easier to just add in reference, of course maybe I'm just used to log scale rather than percent in reading these things.

I can plot it that way, just for everyone wondering how things look as percentages, I decided to plot it this way. Good suggestion with the SPL meter. I'll get one tomorrow.

Jed

[mazurek]

I still haven't played with ARTA that much, probably one of the easiest measurement programs out there though. I still need to figure out how to do gated measurements with it. The measurements I posted were with a made up dB scale, I don't know how it got assigned, but only the relative difference was meaningful. I figure the system could be calibrated once with the external dB meter, if the mic preamp gain was known, but my gain knob has no detents so its hard to know where it is.

Just be careful, the RadioShack SPL meter is not flat sensitivity. There are posted correction curves though.

I remeasured my bass bins with better stuffing, and now their distortion profile is very similar to yours, I may take them higher, but I will have to start thinking about putting in a notch for the breakup. I'm still thinking about a 5" midrange, I don't know if I'd be willing to go all the way down to a three inch like yourself. BTW, the 15M4531K says it has SD-1 motor, do you have information that still says its not as good as its bigger brothers?

[Jed]

BTW, the 15M4531K says it has SD-1 motor, do you have information that still says its not as good as its bigger brothers?

I'll dig up the files capslock sent me for the 15M. As I recall, the 12M was slightly better- but I'll go through my emails and PM you the info.

What woofer are you using?

I just picked up the SPL meter from radioshack and will play around with it to get the levels closer to where I think they should be and post updated plots.

Edit: I forgot I deleted all the emails with attachments in my inbox folder

As you mention it does make sense the 15M would have lower distortion than 12M, but if my memory serves me right they didn't. You can't go wrong with the 15W series, as that is tested by Zaph.

[mazurek]

The sub-woofer that is measuring very similar is Dayton RS265HF, my current woofer is 6.5" Scanspeak 18W4531G00. I figure the Dayton could be taken up to 200Hz maybe to stay well away from the breakup, I feel I could get away with a 5" midwoofer instead of a 6". I could save up and get on the accuton boat and do a C90-T6-89 5" woofer, or do a 15W8530K like Zaph. I'm shooting for 108dB/m and have a total of ~110 Watts at 8ohms to get me there or 180 watts at 4 ohms. If I did this all over I might start with the Aura like yourself which maybe could be taken marginally higher, but the Dayton RS10" do quite respectably.

I'll stay away from the 15M, must be a reason nobody has any projects, too bad on the attachment, thanks for looking.

[Jed]

I'll stay away from the 15M, must be a reason nobody has any projects, too bad on the attachment, thanks for looking.

No problem, you could PM Jon or Capslock for the files as well, just to verify things- they might still have it.

Sounds like you have a nice system in the works. Choosing drivers is difficult work. The new excel W16NX looks interesting as well. I'd buy some to test, just I've used up a lot of my speaker building money.

[Jed]

Aurasound NS10 513 at ~85DB with mic about 5" away

Assessments after the distortion measurements:

Starting with the Aurasound NS10. This driver appears to have very low distortion below 300HZ, with a slight peak in 3D products at 400HZ. However, above that we have decreasing distortion so this one could theoretically get crossed pretty high for a "subwoofer". As Jon mentioned somewhere, these are really woofers. I'm pretty impressed with the sound quality of this driver, since using it up to about 400hz LR2. Actually, the slope I used was a bit steeper than LR2 and it sounded pretty darn nice.

My final design calls for 2 of these Aurasound NS10 drivers wired in series. Sensitivity seems to be around 85-86DB per 1 Watt (single driver) or so based on rough estimates with my new RadioShack SPL meter. Very low even order distortion and higher order products.

Image not available

Accuton C79 at ~88-90DB with mic about 5" away

The Accuton seems to perform best from 200HZ to 2.5Kish. In my past design using this driver, I had it running up to about 3.5K crossed over with a modified LR4 slope. Now that I see some more distortion numbers, the trend suggests an optimum crossover point at around 2K and a max crossover point for low distortion at 2.5K. I remember posting something about how the treble just didn't have the same tonal qualities to match the speed of the Accuton's midrange. I think what I was hearing was slightly elevated D3 products in this area, in combination with a tweeter that gets stressed playing below 3.5K.

Image not available

Visaton MHT12 at 92~95DB

The Visaton MHT12 is a really sweet sounding tweeter, but as the levels increase, so does the even order products and there seems to be a peak in 3rd order distortion (not too large) but noticeable at 3K. Based on some new information, it appears the MHT12 would like a crossover in the 3.5-4K range with a steep slope. Ideally, I don't think this is the best match for the Accuton C79 based on my listening sessions crossed at 3.5K, and its distortion profile. I think a nice mid-dome would work well with the Visaton MHT12 at around 4K LR4.

Hopefully, when my Visaton Ke25s that I ordered come in next week they will have pleasing results for a crossover in the 2.25-3K range. I wish I had the C79 performance information a bit earlier, as I probably would have picked a tweeter that is a little cleaner down to 1K- as you can see from augerpro's results of the Ke25, there is a bit of noise in the 1.5K range. Willl an LR4 slope at 2.5K be steep enough to clean things up a bit in the lower treble for the Ke25? I'll soon fine out.

Image not available

[Johnloudb]

I'm looking for a good OpAmp for an active crossover. Only using it for low freq only, at 350Hz crossover point, L-R 4th order. I'm thinking AD811. I welcome any suggestions. ???

[Jim Holtz]

I'm looking for a good OpAmp for an active crossover. Only using it for low freq only, at 350Hz crossover point, L-R 4th order. I'm thinking AD811. I welcome any suggestions. ???

I'm not sure if the LM4562 is appropriate for your situation but is was designed for audio applications and is the nicest, cleanest sounding opamp I've used. Exceptionally clean, tight and accurate sounding.

Jim

[Johnloudb]

Nice OpAmp

I'm not sure if the LM4562 is appropriate for your situation but is was designed for audio applications and is the nicest, cleanest sounding opamp I've used. Exceptionally clean, tight and accurate sounding.

Jim

Yes, that is perfect for my application. I'll use the LME49710 single OpAmp package though. Thanks!

[Jed]

I'm looking into the implementation of some asymmetrical slopes for my project to see if I can get a flatter power response. The ever popular Linkwitz topologies have a lot of advantages, but I'd like to see if I could perhaps get something smoother off axis using asymmetrical bessel and butterworth configurations- without resorting to omni designs.

A long time ago, Jon recommended I look into a midrange bandpass with Bessel 2nd order LP and 3rd Order Butterworth on the HP for tweeter if using drivers with an extended midrange and lower treble response. I'm going to do some research to see how the above will sum and how phase is influenced, as well as power response, and how individual off axis measurements will look.

One observation I had with the Aurasound NS10, is that perfect LR2 didnt sound as good when I had the low end of the C79 summed in. I increased the capacitance of the shunt and the result was a steeper slope, I didn't overlay a 3rd Order Butterworth target slope over it, but I wouldn't be suprised if it was something close to that.

Traditionally, I've used LR2, and LR4 slopes for their flat on axis response characteristic, but as can be seen in a study by Roman B. at www.rjbaudio.com, sometimes LR4 results in a peak in the power response just past Fc. If crossing over in the critical 2-5K presence region, that peak in the power response can flare up and get pretty annoying. Some drivers, like stiff pistonic drivers, may be more noticeable to create a sense of brightness, if there is elevated distortion in this area as well. Even the C79, with its damped "ears" exhibits a cone breakup induced D3 harmonic peak at 3.5kish. If I do have end up using a Bessel 2nd order filter at 2.5K, this peak will have to be notched to smooth out the final response and reach the target response.

More to come as I investigate these options. I'm sure LR4 will work fine, but I think I should look into some alternatives, especially after taking the distortion measurements and making some new observations.

Edit: Some interesting reading about Bessel filters- http://www.rane.com/note147.html

Here is a link to a Linkwitz filter article http://www.rane.com/note160.html

[Paul Ebert]

Hopefully, when my Visaton Ke25s that I ordered come in next week they will have pleasing results for a crossover in the 2.25-3K range. I wish I had the C79 performance information a bit earlier, as I probably would have picked a tweeter that is a little cleaner down to 1K- as you can see from augerpro's results of the Ke25, there is a bit of noise in the 1.5K range. Willl an LR4 slope at 2.5K be steep enough to clean things up a bit in the lower treble for the Ke25? I'll soon fine out.

Seems like a waveguide might be appropriate?

[cjd]

I'm not sure if the LM4562 is appropriate for your situation but is was designed for audio applications and is the nicest, cleanest sounding opamp I've used. Exceptionally clean, tight and accurate sounding.

Jim

I've got two in my pre as an instrumentation amp (paired with the THS4131). Slick piece. (the pre outputs a fully balanced signal)

[Jed]

Seems like a waveguide might be appropriate?

Hmmm. I'll look into it.

Jed

[JonMarsh]

Well, my ONE KE25 came in today, so after I was done doing another series of driver measurements for the Modula CC center channel, I did a quick pass for HD and SPL using very near field measurement, at one watt drive, just to get a feel for the behavior.

Though the test conditions and equipment aren't that similar to other posts here, the general trends look pretty similar, with a fair amount of excursion related HD2, and fairly low HD3.






Would prefer to see somewhat lower HD2, but in my proposed M8ta three way upgrade, these should work fine, and they fit well in the existing opening.

[Jed]

Looks good Jon. I also got my package of Ke25s today. I'm pleasantly suprised how nice they are. Very beefy and nicely constructed.

I took some distortion measurments as well, but not at the levels you test at. This is at about 1 watt.

Image not available

For those of you who like distortion %s above and DB below.

Image not available

In recent developments, I'm testing a small 3-way with MHT12, RS52, and SEAS ER18RNX in an old box. This will be a speaker for my TV- later maybe I'll start getting into HT, I've always been a Hifi 2 channel kinda guy but I have to find uses for all these extra speakers!

Image not available

And here is the predicted response after taking measurements of the individual drivers and using LSPCAD 6.0 Standard for crossover design.

Image not available

I'll build the crossovers tomorrow and give it a listen and measure the system together. Then, I'll more than likely plug the Ke25s into my larger floorstander design with Aurasound NS10, and Accuton C79. Looks like the Ke25s have potential for a nice match in this low distortion system crossed around 2.5K.

[TacoD]

Why don't you attenuate the rising HF response? Success with building the x-overs .

[Jed]

Why don't you attenuate the rising HF response? Success with building the x-overs .

I'll see how it looks at 2meters in a different room.

[JamesK]

Hey Jed,

Maybe you can use these instead of your C13's.

1/2 price! That or Jon should add them to his driver collection. :T

[Jed]

Hey Jed,

Maybe you can use these instead of your C13's.

1/2 price! That or Jon should add them to his driver collection. :T

Good find on that one! :T

In other news I've spent a good portion of the last 2 days testing the little speaker above. For $34 the Dayton Rs52 is a steal. They are not without limitations however, as most of you are aware.

The bass of the ER18RNX is right up there with the CA18RNX, if not better. Very nice for $62.

I'm really getting the hang of LSPCad. The optimizer is fun to watch.

[JonMarsh]

Hey Jed,

Maybe you can use these instead of your C13's.

1/2 price! That or Jon should add them to his driver collection. :T

Nah, I want to hold out for finding a pair of D30, I wonder why Madisound doesn't even have them up on their online ordering page? :W 8O

[JonMarsh]

Good find on that one! :T

In other news I've spent a good portion of the last 2 days testing the little speaker above. For $34 the Dayton Rs52 is a steal. They are not without limitations however, as most of you are aware.

The bass of the ER18RNX is right up there with the CA18RNX, if not better. Very nice for $62.

I'm really getting the hang of LSPCad. The optimizer is fun to watch.

The RS52 is a very nice midrange for that money- SPL capabiity and low end extension being the main limitations, of course, though in my brief experience you can push a small system with them to some pretty nice in room SPL. (I'm not looking to go deaf any sooner than I have to)

[Jed]

The RS52 is a very nice midrange for that money- SPL capabiity and low end extension being the main limitations, of course, though in my brief experience you can push a small system with them to some pretty nice in room SPL. (I'm not looking to go deaf any sooner than I have to)

Yup, they'll do 98-100DB @ 2meters no problem (crossed over modified LR4 800-3.5K). Not that I've tried it. Right now I'm shaping the response of my little three way and I'm tempted to add another ER18RNX. Using one ER18RNX I really have to cut the mid and tweeter down substantially. Jon, I can see why you use 2 woofers in your new speaker with RS52- especailly if people are going to use it for HT.

As for Madisound carrying the Diamond tweeters? That's a lot of SEAS inventory for one pair of D30s!!!

[Jed]

What took me 2 days to figure out probably would take Jon 5 minutes. The motivation behind this curve is to get that nice summing I saw in Jon's design. This response represents the Er18RNX, Dayton Rs52, Visaton MHT12 response. I wasn't sure how Jon got his response so perfect, but then I started over from scratch with my design and added the Zobel networks first. I think having the Zobels remain constant and then working off a flat impedance really helped the phase alignment. I haven't built this crossover yet, but this is encouraging and good practise before I start the Accuton 3-way!

Image not available

[JonMarsh]

Looking very nice, Jed! :T

You're "connecting the dots", too... each new project I usually learn or figure out a few more things- it never ends. :W

The other thing I find I have to be real careful about is the measurements and delay compensation so I have good data to work with going into LspCAD.

~Jon

[Jed]

Looking very nice, Jed! :T



The other thing I find I have to be real careful about is the measurements and delay compensation so I have good data to work with going into LspCAD.

~Jon

I made sure I took all the measurements for the project with the same settings, then spliced the nearfield response, adjusted baffle step, and delay for the woofers. I didn't bother to splice in the port response though so these are more extended in the bass than the graph indicates, just in case someone was wondering.

I've also cross referenced the measurements with ARTA verus JustMLS to make sure everything is accurate.

[Jed]

Visaton MHT12 tweeter, Dayton RS52, SEAS ER18RNX 3 way frequency response measured in room at about 1 meter. Ignore the data below 200hz. I'm pretty happy with the results. When I have a few hours in on listening I'll post my impressions of the drivers.

Image not available

Here's the impedance plot.

Image not available

[Evil Twin]

Most impressive, young Jedi.

Room modes and floor bounce are rather annoying, though, aren't they?

[Jed]

Most impressive, young Jedi.

Room modes and floor bounce are rather annoying, though, aren't they?

Thankyou ET. Crossing at 800HZ to the mid somewhats limits the placement of the woofer closer to the midrange, so avoiding floor bounce cancellations is impossible given this format. Perhaps I'd have better luck with say an TMWWW 3.5way or 4 way?

I'm actually giving more thought to 4-ways lately, especially after recently visiting Linkwitz's website and he mentions using multiple drivers over a narrower range is ideal, of course the downside being crossover complexity. Wish I could afford some Accuton 7" mid woofers right now. Imagine a dual aurasound NS10 push pull bandpass subwoofer, Accuton C95, C79, and Visaton Ke25SC 4-way in a Vandersteen Model 5ish box. Whatayou'll think? Can't really afford C95's right now so I could use ER18 until the time is right. Since everyone seems to be doing the ISIS thing, I might change my tune and do something different.


Are there any downsides to the push/pull bandpass sub enclosure?

[TacoD]

Good work Jed, I do not like the sound of bandpass subs. But that's just my opinion .

[Jed]

Good work Jed, I do not like the sound of bandpass subs. But that's just my opinion .

Thanks Taco!


I do recall hearing one bandpass sub, and it seemed peaky in its response. But that company went out of business. Can you describe the response you heard?

[Jed]

Getting very close to finalizing the MHT12, RS52, and SEAS ER18 3-way. Here is some more info about the design. I made some recent changes to the woofer filter because I felt I wasn't crossing them steep enough and there was a peak in the response around 600HZ that got annoying with some source material.

Slopes are LR4 all around.

Images not available

And the not so finalized but very close filter for the 3-way: Note, I haven't listened to this exact woofer LP network yet and my extra parts will be here on Tuesday. I'll have some final pics of the system and layout with final listening impressions next week. This has been quite a learning experience.

Image not available

[Jed]

Planning for the Dual NS10, Accuton C79, Visaton KE25SC begins.


Question- In an effort to reduce box volume using a sealed box and improve transient response, could I have one Aurasound NS10 firing downward in a low Q box, while the upper Aurasound fires forward in a higher Q box, all while crossed over at the same Fc, which will be around 300-400HZ?

I think Avalon does a downfiring woofer, but its a 4-way. Now that I think of it, perhaps it doesn't make sense to cross over a woofer up that high if it is downfiring. I don't even know if the Aurasounds are designed to mount that way.

[JonMarsh]

You could mount the Aurasound that way, and it might be an interesting side experiment, but I doubt it would produce the results you want. The upper driver in the higher Q configuration with smaller volume would have the less damped transient response, the downward firing driver would be pretty messed up in frequency response much above 100 Hz, as it would effectively be slot loaded. Been there, done that, can be used to raise the Qtc for an overdamped drive and lower Fb at the same time, but the NS10 doesn't have the right parameters, IMO, for that approach- you'd be looking for a driver with Qts in the 0.2 to 0.25 range, in my experience. Just my thoughts- but don't let my inputs prevent or stop your experimentation- sometimes that's one of the best ways to figure things out. Just think through how it's all going to work together before you start making the first MDF dust. :W

Personally, I'd encourage you to attempt something more like the Indra, but with dual 10's. And a downward firing port tuned along the lines suggested for the Three Way Design Study. I don't think ET would mind.... :W

[Jed]

Personally, I'd encourage you to attempt something more like the Indra, but with dual 10's. And a downward firing port tuned along the lines suggested for the Three Way Design Study. I don't think ET would mind.... :W

I was hoping you'd chime in with your experiences on this.

OK, next question.

Are the added benefits of increasing the baffle from 2" to 4" multidensity layers really worth the added weight? EDIT: Don't be rediculous JED, they should be 6" thick not a whimpy 4"!!!

[Jed]

The Duet 10 concept.

2X Aurasound NS10, Accuton C79, and Visaton KE25SC. More to come.

Image not available

Personally, I'd encourage you to attempt something more like the Indra, but with dual 10's. And a downward firing port tuned along the lines suggested for the Three Way Design Study. I don't think ET would mind.... :W

I feel a bit more comfortable working on something like this. Baffles will be 4" thick, width is around 12".

[JonMarsh]

The Duet 10 concept.

2X Aurasound NS10, Accuton C79, and Visaton KE25SC. More to come.

Image not available

I feel a bit more comfortable working on something like this. Baffles will be 4" thick, width is around 12".

I like, Jed! :T You know, 2" is good, but I'd done "bookshelf" class speakers (like the Unorthodox 8" two way, published in AuxioXpress; which is sort of the uncle to the M8ta) with 2" thick front panels. The front baffle is the launch panel- 6" is justifiable with really deep bevels, but unless you get a good 12" cabinet saw, it will be tough to go that far. 4" is pretty dang solid....

At this rate, we'll have to have an Avalon Klone sub forum soon... :B :rofl:

[Jed]

At this rate, we'll have to have an Avalon Klone sub forum soon... :B :rofl:

Thanks Jon,

I updated the plans a bit. This just feels right for my next project. 4" thick baffles it is! I figure- why not? If I'm going this far I might as well go for the gusto.

Next week I'll be picking up the MDF and Baltic Birch at my lumber store.

Image not available

I like the idea of the removeable top head unit for midrange and tweeter, I just think I need to take things one step at a time. ColoradoTom, Tyler, and others will have to make that one.

Image not available

There's not much sense in me doing what everyone else is doing either. The more options the better!

[chrismercurio]

Looks good,

I sold Avalon for a number of years and most people miss some critical details to the cabinet construction when they build clones. The drivers are all braced inside. I never took pictures so it will be difficult to explain. Imagine a frame of MDF that is built into the cabinet with a foam rubber (or similar damping material between the back of the magnet and the frame. The drivers are coupled very firmly to the cabinet in this way with little to no play at all in the installation of the drivers. It does make the design and execution of the cabinets much more difficult.

In my view Avalon creates one of the finest cabinet loudspeakers available, second only to Magico.

C

[Jed]

In my view Avalon creates one of the finest cabinet loudspeakers available, second only to Magico.

C

I'm willing to compromise some details and save ... oh... $20,000-30000. :W

I know what you mean about the bracing, I've seen pictures of it at the Opus clone website. Not sure if it is worth the extra 200hrs or so. I mean, come on, how many speakers have 4" thick baffles.

Images not available

I think there is room for improvement even with their cabinets. For example, not much breathing room for the Accuton C79 above.

[Johnloudb]

Question for Jon Marsh (and anyone who knows)

Jon,

I'm working on this B-G RD50 (dipole) two way design with an active crossover. You mentioned that you're partial to dipoles. I'm a little concerned about the front baffle thickness of 2.25" of maple. If I made the baffle 3" to 4" the rear part of dipole would be in a 1.5" to 2.5" sink hole. I was worried about diffraction and that it would make the rear wave sound horn loaded. I'd greatly appreciate any suggestions.

Thanks, John Loudenback.



Image not available

[Jed]

I'm a little concerned about the front baffle thickness of 2.25" of maple.

For your application, the 2.25" baffle will work nicely. If you want, for the lower section of your design where the woofers are housed, you can beef it up to 4".

I'm going with 4" because the design calls for large facets/cuts to bevel the sides, and the woofers are fairly large and cause quite a bit of vibration.

I'd also invest in a measurements system so you can see the effects the baffle has on the RD50. A simple out of the box active crossover will not be optimum for the RD50 in my opinion because it doesn't deal with diffraction and baffle step loss. There is a lot of info out there about passive crossovers, and schematics for the RD50 active crossover as well. That's where I would start.

[JonMarsh]

"Sound" comments and recommendations from Jed all around.

PARTICULARLY for a dipole configuration of any kind, I do a fair amount of simulation with EDGE before deciding on baffle shape, configuration, and driver locations. Saves problem solving and hair pulling later.

Also, a reminder that it's tricky going from a line source in the midrange to a point source monopole bass source; you can really only balance it correctly at one distance from the speaker. But this might be a good way to get some experience with the RD50, listen to some tools, then do a more optimum design later. I've got a pair of RD50's in storage for a while, and have gone through many "thought experiments" with them, but no hardware yet. Have a whole mess of Fountek ribbon tweeters to use with them, too. I'm still looking for the optimum midwoofer to use with them in dipole, I had hopes for the 7" Aurasound, but it was never released. May end up going with 10" or 12" Aurasound, but that is a bit pricey, and may push the baffle size I'm looking for. Another possibility is the RS270, I still haven't done detailed measurements, and they would be a lot heavier than I would care for (Neo rules!)

Considering a Duelund style 3 way crossover implementation, using the RD from about 400-500 to 4 kHz.

[Jed]

I had hopes for the 7" Aurasound, but it was never released.

Aurasound has the potential to really push the envelope. Maybe we can sign a petition to get some ceramic mids 4-7".

[Johnloudb]

I'd also invest in a measurements system so you can see the effects the baffle has on the RD50. A simple out of the box active crossover will not be optimum for the RD50 in my opinion because it doesn't deal with diffraction and baffle step loss. There is a lot of info out there about passive crossovers, and schematics for the RD50 active crossover as well.

Thanks Jed,

I'm designing my own active crossover. I'll definitely get some measurement software. I'll have more questions when I get that far.

[ColoradoTom]

I like, Jed! :T You know, 2" is good, but I'd done "bookshelf" class speakers (like the Unorthodox 8" two way, published in AuxioXpress; which is sort of the uncle to the M8ta) with 2" thick front panels. The front baffle is the launch panel- 6" is justifiable with really deep bevels, but unless you get a good 12" cabinet saw, it will be tough to go that far. 4" is pretty dang solid....

At this rate, we'll have to have an Avalon Klone sub forum soon... :B :rofl:

As some of you might know... I had a problem with water damage to my M8ta's and I decided to make some fixes to get them to an acceptable look. As part of the design process for the "Isis" like project I will be required to make some rather large bevel cuts and decided to try a new technique.

I purchased a Festool TS 55 Eq plunge/circular saw a couple of weeks ago along with a 96" guide rail. The guide rail makes bevel cutting highly accurate and I seem to be getting within 1/32" with my cuts. I set the saw up for a 45 degree bevel and recut the bevels on my M8ta's to about a 4 1/4" deep. This makes the M8ta's look more like an Avalon Opus from the front - I believe that Kris also cut deeper bevels on his pair. I've included pictures showing the pair from a distance and a closeup of the bevels.

To make the bevels I cut from one side and then the other which leaves about a 1" center strip that I cut using a handsaw to complete the bevel cut. Because you have the previous cuts from the plunge/circular saw as a guide, the cut with the handsaw is very easy. With about 10 -15 minutes of sanding the bevel is complete. This is very simple, accurate, and very safe. The thought of running a 6" thick piece of MDF through a jig on my table saw was really freaking me out - this is a better method and I'm pretty sure I could get close to 6" bevels if I needed to with the same method. The Festool guide system makes this trivial - and someone could probably make a rail guide for their skill saw with a little effort that could do the same thing. Total time to cut the facets and sand them was about 1 hour.

Hopefully, the modification to the design won't require any major changes to the crossover or provoke the rath of ET! 8O


Tom