Open baffle plan/layout

[Saurav]

I'm trying to figure out how to design my OB to minimize the amount of cutting I need to do (since my woodworking skills are close to zero). Drivers:

* 12" woofer
* 6.5" midrange
* 10" WG (probably)

Based on advice in ABCDipole, I'd like the baffle around the midrange to be ~12" wide, and I'll keep the driver centered. Not sure about the tweeter yet, I don't know how much baffle edge diffraction is a factor with a WG/compression tweeter.

I'd initially planned on a trapezoid, narrowing from 16" on the ground to 12" on top, but there's no way to get that cut at Home Depot, so it would be up to me to get clean cuts for that. Instead, I've been thinking about something like this:

Image not available

The lower part is roughly 16" square, and I'll probably put some wings on the sides for stiffness (and increased baffle width). The upper part is 12" wide, which should be an easy size to get. I was thinking of extending the narrow baffle down to the ground and gluing it to the back of the lower baffle. That gives me extra thickness around the 12" driver, and should hopefully be attached well enough to the lower baffle. I might make the upper baffle 14" wide instead of 12", that might make it easier to cut the woofer hole in the doubled thickness.

Does this seem doable? Or will I run into any construction issues I haven't thought of? I've modeled this in EDGE and the response isn't too far off my trapezoid baffle.

[dlneubec]

You didn't say what drivers or crossover points you were considering, so I'm just guessing here.

One thing that occurs to me is that you might want to consider the effect of floor bounce cancellation on the mid and woofer responses. Attached is a quick look I took in BoxyCAD2 using your drawing and assuming a listening distance of 11' and height of 38". You can see that it predicts a null in the midrange FR due to floor bounce at around 540 hz. Unless you cross enough above this so that the null is not in the passband, it might be noticable. Where you crossover could also have some effect on BSC, your open baffle simulations in ABCDipole and also effect just how much of the midrange is truly dipole in it's response. Just some food for thought.

[Saurav]

Thanks for that reminder. I had run Boxycad2 floor bounce sims on some of my earlier iterations, but not for this one. The midrange driver in my current speakers is higher, but that's because the tweeter isn't 10" in diameter.

I'm not quite sure how to solve that problem without making the speakers taller than I was planning. I could try the XT1086 waveguide instead of the DDS, which means I should probably switch to a bolt-on driver instead of the threaded one I'm thinking of using now.

Drivers...

Tweeter: I bought a pair of BMS 4540NDs, and I'm waiting to hear back from Assistance Audio about the DDS waveguides.

Midrange: Most likely my current Audax PR170M0

Woofer: I'm calling it the AE OB12 - it's a variation on the IB15 that John and Nick thought up that should work better as an OB woofer. Details here.

Crossover points: 300-500Hz, and 1500-2500Hz, roughly.

So you're right, the midrange height in my drawing will not work for the crossover to the woofer.

The simplest solution seems to be to make the speaker taller. If the midrange is closer to 29" off the ground, its dip moves to 350Hz, and now a 500Hz LR4 XO is more feasible. This means the speakers are now ~45" tall. I was trying to keep them < 40" because they overlap the bottom of my screen, but I can think of other solutions for that problem.

[dlneubec]

I'm not sure floor bounce nulls are that critical anyway, but I thought it worth a reminder to consider along with other factors (as if there aren't enough already :E ).

It seems like the Audax, with only .5mm xmax, would do better with a higher crossover anyway, from a power handling standpoint. If you are using ABCDipole, you have probably already tried different crossover points for the mid driver, and looked at what your max spl would be at that Xmax. I can tell you that Jeff B., in his open baffle Salk design, was much less concerned about the dipole radiation pattern than just wanting to have an open back mid with a high enough spl capability without reaching Xmax at the low end.

[Saurav]

It seems like the Audax, with only .5mm xmax, would do better with a higher crossover anyway, from a power handling standpoint.

Well, for me, that's tempered by the fact that I'll be driving it with 4W, but yes, absolutely. And I don't want to design a speaker that's broken if somewhere down the line I decide to change amps.

Yes, I've been using ABCDipole, as well as MJK's worksheets and Linkwitz's spl_max spreadsheet. 400Hz and higher, I'm OK on the max SPL capability of the driver. Right now my SPL bottleneck will be the tube amp driving the mid/tweeter, which is fine, it gets loud enough for me.

I can tell you that Jeff B., in his open baffle Salk design, was much less concerned about the dipole radiation pattern

Interesting. If I understood ABCDipole correctly, John K. recommends a baffle width that's ~2x the driver diameter, and the reasoning behind that is to have the baffle provide the dipole loading up to the point where the driver directionality takes over. If the baffle is too wide, you get a transition from dipole behavior at low frequencies, to monopole behavior in a range where the baffle isn't helping but the driver is still a point source, and then narrowing back up again at higher frequencies. The wider baffle also creates a deeper dip above the dipole peak, which makes EQ harder.

My current speakers have the same Audax mid on a 16" open baffle, and I can see the dip in the measurements. I've mostly ignored it (this was meant to be a test/learning project, that ended up in my living room for a few years now), but this time around I want to do a better job.

So this time I'm going with a 12" baffle which should provide dipole loading to a higher frequency. Above that the driver starts to become directional, and hopefully I can cross to the waveguide at a frequency where the dispersion of the two drivers is close, but it's still within the usable range of the mid and the compression driver. That's not full-range dipole, but the front lobe should stay somewhat consistent in dispersion, and the back lobe will narrow out and disappear (since I don't have a rear tweeter).

We'll see how that goes once I get the drivers and measure them. But at least, that's the plan. Again, lesson learned from my current speakers, which use a ribbon tweeter. The XO frequency is a stretch for both drivers, and the treble dispersion is much wider than the upper midrange, which makes it sound splashy when it measures flat on-axis. I currently have the on-axis response tipped down a few dB to get the in-room balance to where I like it.

About the Audax: This is the manufacturer's data that shows up on most sites, including Madisound:



But I found this at an Audax archive site that Amphiprion posted some time back, and this matches my own measurements more closely:



Notice how this recommends a 600Hz XO @ 18dB... I'm thinking 500 @ 24dB shouldn't be too far off from that.

What do you think about my plan of gluing the two sizes of baffles together? Do you foresee any problems with that? Would I need more stiffening of the upper baffle, like the Isiris does with a brace running down the middle/back?

[dlneubec]

I think gluing them together will make the mid and woofer baffle more rigid, but mostly in the vicinity of where they overlap. I'm guessing you might need some more support on the mid/tweeter baffle.

I have two of those Audax, so I'm interested in how it works out for you for some unknown future project.

Maybe the Audax is a little larger in diameter, but the max size that ABCDipole suggested for the 6.5" B&C 6md44 in my Bagby/Salk inspired design is just over 10". I have my tapered baffle at just under 10" at the center of the mid. I passed it by John K. for his feedback and he thought it looked just fine. There is hardly any dipole dip at all in the modelling. Interestingly, the original design that Jeff Bagby did that ended up later as the Salk design was also only 10" wide, though the final Salk design is more like 15".

[Saurav]

Have you run any distortion measurements on it? That's the one piece of information I don't have, that I'd like to see. Overall I like the sound of the driver. augerpro had offered to measure mine once he comes back from vacation later this month, and I might send mine to him. If those numbers don't turn up something horrible, I'll probably use this driver.

Maybe the Audax is a little larger in diameter, but the max size that ABCDipole suggested for the 6.5" B&C 6md44 in my Bagby/Salk inspired design is just over 10".

Huh. Maybe I did the cm -> in conversion wrong I'll have to go back and look at that. Or maybe there are other factors than just driver diameter that go into that recommendation.

A 10" baffle will make mounting a 10.5" waveguide a little complicated, unless I taper it back up, and I'm not sure I can pull that off with anything that looks remotely aesthetic. One option might be to go for a top-mounted-horn kind of a look... which usually works better on top of a box, I'm not sure how that'll look on an OB.

Hmm... OK, I clearly need to fire up ABCDipole tonight and look at those numbers more closely.

I'm guessing you might need some more support on the mid/tweeter baffle.

I was wondering about that, which is why I thought I'd run it by you guys here. I might try strips glued to the baffle edges, I've seen some OB designs here that did that to stiffen the baffle.

[Saurav]

OK, you were right. The recommended baffle width is 10.5", and 12" models pretty close (and actually has a smaller dip). The recommendation is calculated from the driver diameter, which is calculated from Sd, and the 6MDN44 has a slightly lower Sd than the Audax.

[Saurav]

Well, in a few months I expect I'll look back at this and laugh at how differently the final thing turned out, but here's a model of what I'm thinking of for now:

Image not available

The midrange is at 30", the WG at 38", and my listening height is around 35". And with these dimensions, it's pretty clear that the upper panel will need bracing. I might move those drivers down a little and see how much the floor bounce actually affects things. I'll probably also start with shallower wings, anecdotal evidence seems to suggest that deep wings don't sound as 'open'. That'll depend on measurements.

And I doubt I'll have those nice roundovers. If only my woodworking skills were anywhere near my 3D modeling skills...

[dlneubec]

I have no measurements for the Audax and probably won't have until later in July, at best, since I've got my hands full finishing up two other projects for the Dayton DIY on July 12.

Wouldn't ribs along the outside edges act just like mini wings? I would think they would to be modeled. It seems to me that one central rib from the top of the bass driver to the bottom of the midrange and then a short one between the mid and waveguide, as long as will fit would do a decent job of stiffening the panel.

Another option would be a central rib that is deeper with cutouts over the drivers. Here is how I did my NaO mini MTM section. I just used my Jasper jig to cut the radii in mdf and then a 3/8" roundover.

[fjhuerta]

Are you guys planning to use passive or active x-overs?

[Saurav]

Wouldn't ribs along the outside edges act just like mini wings?

Maybe. I was thinking of a 1" strip of wood glued to the outside edges (let's say the panels are 3/4"). That shouldn't add too much to the baffle width, and should help spread the diffraction effects out a bit? Something like the edges on top panel of an Orion.

It seems to me that one central rib from the top of the bass driver to the bottom of the midrange and then a short one between the mid and waveguide, as long as will fit would do a decent job of stiffening the panel.

Hmm, I hadn't thought about that, I was thinking any rib would have to connect the panel to the base (like you show in your photo) for it to be effective, or the panel will still just 'flap around'. But that's not true, since the panel doesn't have a single flex point. OK, I'll probably add a brace like that.

Are you guys planning to use passive or active x-overs?

I think Dan's design is passive? Mine will be a hybrid, passive between the mid and tweeter and active to the woofer.

[Saurav]

So something like this:

Image not available

I'm thinking the brace needs to be thick (in the front-to-back direction), or it'll just flex with the panel, right. I could go thicker, like a 2"x4" strip.

[dlneubec]

I would suggest narrow in width (3/4" stock should be fine, preferably plywood), but deeper for a much more effective brace (think I-beam or floor truss). It would be most effective if the brace bridged the mid driver, if possible. Otherwise, the short piece between the mid and waveguide might not be much help, IMO.

[Saurav]

Otherwise, the short piece between the mid and waveguide might not be much help, IMO.

I was thinking the same thing after I put that in there. That's just a block of wood glued to the panel. I'll leave that out, and see if I can get the brace to bridge the mid. I'll be trying for as little gap between the mid and tweeter as I can, so I'll probably just end up with the lower brace.

I would suggest narrow in width (3/4" stock should be fine, preferably plywood), but deeper

OK. I drew that at roughly 2" square. How deep should I go? 4/6/8"? I know I've seen 1x4 and 1x6 at HD, not sure about 3/4" thickness. Or I could cut it out of plywood, like you said.

[AlanH]

Another option would be a cental rib that is deeper with cutouts over the drivers. Here is how I did my NaO mini MTM section. I just used my Jasper jig to cut the radii in mdf and then a 3/8" roundover.

Hi Dan,

Sorry this is a little off topic, but in the picture of your NaO mini above, it looks like you used the Peerless/Vifa XG18 glass fiber woofers rather than the Peerless exclusive drivers spec'd on the Music and Design website.

I'd like to experiment with dipoles and since I already have the XG18 and TDFC drivers in a Zaph design, it would be easy to build a dipole baffle like yours and experiment. Did you use an active or a passive crossover?

Thanks!

[CraigJ]

Saurav,
Did you get my p.m.? If not, congrats on the BMS 4540NDs, I think you will find them a nice upgrade. Are you making the baffle out of 3/4" mdf? If so, Mark K demonstrates an excellent way to stiffen his baffle with maple strips. If you go here and scroll down to "A new Baffle is born": https://www.audioheuristics.org/curr.../whats_new.htm Another idea, have you considered separate panels? From your drawing, you could easily put a "foot" on the top baffle, and slide it up to your woofer.

Craig

[Saurav]

Did you get my p.m.?

Nope. Was that today? On this forum? It says I have 0 private messages.

Are you making the baffle out of 3/4" mdf?

Or plywood, maybe the nice 13-ply stuff if I can find it, or the 7/9-ply stuff (which is all I've found at HD/Lowes). Ideally I'll build a test mule out of MDF and the final speakers out of something else, but knowing my track record, I'll probably just build this once.

Mark K demonstrates an excellent way to stiffen his baffle with maple strips.

I've looked at that. The exact thing he did might be a little beyond my construction abilities, so I was thinking of a simpler variation. It wouldn't be flush with the front surface, but I'm not sure if that would be a bad thing. If I look at the Orion:



The side strips extend out in front of, and behind, the baffle. I remember reading somewhere that the point of this is to spread out the diffraction, so that instead of seeing one edge, the wave sees multiple edges at multiple distances, so you don't get a sharp diffraction null at one frequency. At least, that's what I think the idea is, I could be totally wrong. Anyway, so if I take a strip that slightly thicker than the panel, glue it to the edges and round over the corners, that might get close.

Another idea, have you considered separate panels? From your drawing, you could easily put a "foot" on the top baffle, and slide it up to your woofer.

Haven't really thought about separate panels. Would make it easier to move this thing What exactly do you mean by a foot? Extend legs down to the ground, maybe with an angle brace at the bottom, and just place that against the back of the lower baffle?

I guess I should mention that one other design goal is to make this somewhat hard to tip over. My 9-month old daughter has just discovered the vertical direction, and spends every waking moment trying to pull herself up onto anything she finds. She also likes to kneel/stand in front of my speakers, hand on the woofer (I guess she enjoys feeling it vibrate). I'm not sure it's the best idea for her ears to be 6" from a woofer that's playing at 85dB 10' away... so I generally try to keep her away from the speakers...

So anyway, one concern with a separate baffle would be that it would be top heavy, and easier to tip over. That's partly why I made the base with a lip out in front, that should make it harder to tip it forward.

Edit: Oh, I think I see what you mean - keep the baffle extending to the floor, with the hole cut out for the woofer, but put a separate base on it that sits on the main base?

[dlneubec]

Hi Dan,

Sorry this is a little off topic, but in the picture of your NaO mini above, it looks like you used the Peerless/Vifa XG18 glass fiber woofers rather than the Peerless exclusive drivers spec'd on the Music and Design website.

I'd like to experiment with dipoles and since I already have the XG18 and TDFC drivers in a Zaph design, it would be easy to build a dipole baffle like yours and experiment. Did you use an active or a passive crossover?

Thanks!

Yes, that is the XG18. John K. did a custom design for me to work with that driver, both the passive and active section was redesigned. The Nao Mini's are a hybrid passive/active design. I probably can't share those deails with you, however, since it was John's crossover design, not mine. Perhaps if you contact John, you could work something out. There is a photo, or used to be, of my NaO Mini's (with subs) on his gallery pages.

John is great to work with and did a lot of mentoring with me through that project an a couple more after that as I started to do my own measuring and crossover design.

[CraigJ]

Nope. Was that today? On this forum? It says I have 0 private messages.

I've looked at that. The exact thing he did might be a little beyond my construction abilities, so I was thinking of a simpler variation. It wouldn't be flush with the front surface, but I'm not sure if that would be a bad thing. If I look at the Orion:

Edit: Oh, I think I see what you mean - keep the baffle extending to the floor, with the hole cut out for the woofer, but put a separate base on it that sits on the main base?

My outgoing pms seem to get locked up lately, in short; Paul W has had excellent results with an 11" waveguide for your BMS driver, if you choose not to use the DDS model. Jack, of Assistance Audio, may have a pair of waveguides that he'll let you demo. I believe Mark K's maple strips are just two different widths glued together. Add a 45 degree angle, or two round overs and your done. Multiple ways to accomplish the same thing. Finally, I think you understand the dual baffle comment and I too have young children and have to think about things tipping over or being poked. Good luck and have fun.
Craig
p.s. the waveguide is wide, so you may have to move the mids closer to the tweeters for better center-to-center spacing.

[Saurav]

I believe Mark K's maple strips are just two different widths glued together. Add a 45 degree angle, or two round overs and your done.

They are. You're over-estimating my woodworking skills OK, I could probably build that, but for the most part, I'm trying to keep this to something I can get cut at Home Depot, followed by a jigsaw at home.

Paul W has had excellent results with an 11" waveguide for your BMS driver

I'll have to look for that. I thought I'd read pretty much everything I could find on that driver in the forum archives. Was this something he fabricated himself?

[CraigJ]

If you do an advanced search with 12" msc and Paul W, you will find a lot of excellent info on waveguides studies by Paul. Regarding woodworking skills, I've found once you get away from "the box", you need fewer tools and skills.
Craig

[Saurav]

Thanks. I searched through the long waveguide threads for posts by Paul W, and found a reference to an MSC waveguide. However, a Google search for "Music Supply Center" doesn't turn up anything (one link returns a 404 error).

I'll have to do more research on the MCM and PE waveguides as well. And there's the XT1086, though that will need an adaptor.

[augerpro]

Looking good Saurav! I'll be watching Dan's and your's progress since it seems you guys will finish your dipoles long before me.

I just got back a few days ago from Europe so I'm working on test baffles today. I'm shooting for measuring with an IB15 this week, so that may be of interest. I'll be comparing flat versus U baffle of two different lengths.

I agree with Dan on using a "spine" instead of side wings. The deeper the spine the stiffer the baffle will be, like an I beam. I wonder if oak or birch plywood would be better than mdf? Or maybe something else?

I'll be curious to see the condition of the DDS waveguides you get. I spoke to someone who uses those in a commercial loudspeaker and he said he has to refinish every one he gets. That sort of made be bite the bullet and put a deposit down on a pair of 10" waveguides from Earl Geddes. Earl is a perfectionist when it comes to tolerance, especially at the throat entry. Does anyone know of a place that carries plastic stock that one could turn on a lathe?

What were you going to use as the active crossover?

BTW I probably won't be able to measure your drivers anytime soon. Between getting my own projects moving again and overtime at work I'd hate to make a promise and not be able to keep it. Next time I do driver testing will probably be when I receive my Geddes waveguides and B&C 15NW76's from usspeaker.

[AlanH]

Yes, that is the XG18. John K. did a custom design for me to work with that driver, both the passive and active section was redesigned. The Nao Mini's are a hybrid passive/active design. I probably can't share those deails with you, however, since it was John's crossover design, not mine. Perhaps if you contact John, you could work something out. There is a photo, or used to be, of my NaO Mini's (with subs) on his gallery pages.

John is great to work with and did a lot of mentoring with me through that project an a couple more after that as I started to do my own measuring and crossover design.

Thanks for the info!

[Saurav]

Ah, he's back!

BTW I probably won't be able to measure your drivers anytime soon.

That's fine, I've more or less made up my mind to use them anyway.

BTW, I know you already have the IB15s, but you might be interested in this:

https://www.htguide.com/forum/showthread.php?t=30022

Look at the bottom of the 1st page. John is going to build me a couple of those.

I wonder if oak or birch plywood would be better than mdf?

Probably. If nothing else, I can 'finish' it without needing to worry about veneering the kinds of panel shapes I'm thinking up. I've never seen the 13-ply Baltic Birch plywood in my local Home Depot or Lowes. The best they have is 7 or 9 ply, and it's usually oak or maple (on the outside surface veneer, not sure what's inside). So... I might start with MDF, or the HD plywood, or look around for local lumber places, or try and buy plywood online. Don't know yet.

I'll be curious to see the condition of the DDS waveguides you get.

That's certainly a concern. It seems that the threaded WGs from Assistance Audio are in better shape than the bolt-on drivers from US Speakers, but that's based on a very small sample size of recent forum posts. If I don't hear back from them in a few days, I'll probably look for other options.

That sort of made be bite the bullet and put a deposit down on a pair of 10" waveguides from Earl Geddes.

I didn't realize you could buy just the waveguides. Got a link?

What were you going to use as the active crossover?

Probably one of these:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=74420

Or use a generic PCB. I have a bunch of opamps and resistors/caps left over from an earlier active XO project, and a decent power supply.

[CraigJ]

I'll be curious to see the condition of the DDS waveguides you get. I spoke to someone who uses those in a commercial loudspeaker and he said he has to refinish every one he gets.

Hello Brandon,
The condition of the current DDS waveguides doesn't appear very good. Ten days ago, I emailed pictures from the "this worked out well..bms/dds" thread, directly to DDS's sales department. So far. zero response regarding their quality control. Personally, I'd ask Jack to hand pick a pair. In the meantime, I'll pm Paul to get his opinion on the PE verses MSC waveguide.

Craig

[augerpro]

I'm waiting for those boards from Bob too. Have you contacted him? I know he will have some extras, but my impression was not that many, unless you got in early and already have a deposit. They are mostly for some subwoofers, but I may use a pair just for the dipole correction. For the woofer to mid/tweet crossover I'm just going to use either the Behringer CS2310 or dbx 223. I have one of each.

Geddes will sell the baffle with waveguide already mounted from his new 10" kit. I think it's 1.25 thick mdf, and you have to cut the waveguide out. Basically just cut the bottom off the baffle and trim the sides so you end up with a 10"x10" square to work with. Currently he is not selling just the waveguide, since it is molded to the baffle at the factory or something. But it's workable with the baffle included. And it's got the geddes contour, foam plug, and matches perfectly my B&C DE250's, so I'm satisfied. And should be good build quality/tolerance. A bit spendy at $400 a pair though.

[Saurav]

I was first in line for the extras (just missed the GB cutoff date), but I haven't paid him anything yet. I've been in email contact with him, but didn't want to bug him too much, since he was pretty unwell for some time. Not a big deal if I don't get one of those boards. It keeps things cleaner, but I can build one of these on a generic PCB (with lots more wire). I have a 2310 and a 3400, and was thinking about ways to combine that with a custom board that just handles the EQ.

Geddes will sell the baffle with waveguide already mounted from his new 10" kit.

OK, I'd seen that on his page. For now I'm going to try the threaded BMS 4540ND, and see where I get. If I end up with a bolt-on horn like the XT1086, I may switch to a bolt-on driver, and the DE250 is the first choice. If you have the DE250, the Geddes waveguide makes a lot of sense, since he said it's designed for that driver, right.

The PE horn throat adaptor seems to be pretty bad based on some comments here. I searched around and found these based on a comment on PE:

http://www.loudspeakersplus.com/product.asp?specific=jokmlpi4

These may work better, who knows. If I buy the XT1086 from these guys, I'll throw in a pair of these. Though it seems like this cannot work as well as using matched horn/driver pairs.

Regarding the DDS waveguides - it seems like most of the good comments are from before the company changed hands.

[Saurav]

In the meantime, I'll pm Paul to get his opinion on the PE verses MSC waveguide.

Where can one buy this MSC waveguide? All the links I found are dead.

[ttan98]

Saurav,

Use Econowaveguide($9.99 each), which I am using now... performance seems quite good together with Selenium D220Ti see Attachment, with NO equalization and filtering, sample at 1/24th oct.


Just joking, all the best in your design.

[Saurav]

Heh, I followed that whole exchange on those two forums. Thanks for reminding me of this, I might just try it out.

[Saurav]

So... something like this?

Image not available

That's about 1'x4".

[Mark K]

Really,

You can use my method with sandpaper, a router, and a circular saw. Really. It works suprisingly well! The basic stock is HD, so you don't have to rip anything. Really. Glue two stock from HD, roundover and 45 deg miter from your router. Uh, that's it.

Hey,
I've got 4 kids between 3 and 7. The GR130//SS MLTL's have had pushed in cones for years now!!!! So I know about tall things falling over.

I too would love a largish waveguide that would go to 1k. When someone makes a decent one, I'll buy it.

[dlneubec]

If it were me, I would make the vertical brace deep enough to have a cutout in it behind the woofer and tie it into the base below the woofer. I would also bridge the mid and tie it into the baffle between the mid and tweeter waveguide and maybe even over the tweeter to the top of the panel, like I did in my NaO mini design or as Jon Marsh did in his open baffle floor standers. You'll have to decide what you are able to accomplish with your skills and tools.

[Saurav]

You'll have to decide what you are able to accomplish with your skills and tools.

That being the key Thanks.

[CraigJ]

Ah, you mean a brace like the one below which was cut using a jig saw?

[Saurav]

Hmm... I do have a jigsaw...

Is that just glued to the front baffle and the base, or are there screws involved? How do you clamp something like that?

[dlneubec]

Simple, clamp it through the driver holes after they are cut and at the top and bottom. I recessed mine into the beack of the baffles using a router, so that it had more glue surface and something to keep it well anchored witout screws, but that is probably not necessary.

[Saurav]

Thanks, I was going to ask you for a clearer shot of the back.

I guess I'll see how doable this looks once I start cutting things. I just think that this will need more accuracy than I've generally been able to manage in the past And I'd like to keep the baffle as short as possible, and the mid/tweeter as close as possible, which increases the cutting accuracy required.

[CraigJ]

Hmm... I do have a jigsaw...
Is that just glued to the front baffle and the base, or are there screws involved? How do you clamp something like that?

Saurav,
Don't worry, you aren't building a complicated thing like a box. Routing cutouts for the drivers with Krutke's accuracy was my biggest challenge. To assemble the Isiris Juniors back, I laid the baffle face down, and propped the back and bottom up into place. Long screws could work for attachment, though I used biscuits and glue because I didn't want screws showing through the Corian.
Craig

[Saurav]

Still thinking about ways to make this easier If I make the main baffle 13" wide, so it can fit the 12" woofer, then I could simplify this to something like:

Image not available

Would that be 'structurally sound'? I've seen driver cutouts get that close to the edge in boxes, I'm not sure if this is a bad idea with an OB. The supports on the sides should help strengthen that area, right?

And I drew that as a 12" cutout, in reality a 12" driver with have a < 12" cutout, right. So the hole is really the rim of the driver.

[Saurav]

So here's what I'm going to try to build:

Image not available

I decided to make the lower baffle taller, that'll give me more glue area between the two pieces and hopefully won't affect FR for either driver.

I got one upper baffle done today:

Image not available

I got the pieces cut to size at Home Depot, and then cut the holes out with a jigsaw, mostly freehand.

Image not available

Image not available

I started trying to flush-mount the midrange, but that's not practical with a dremel and router attachment. So I flipped it over and used the other side. My aesthetic requirements are lower than most of you guys And it'll be the back side of the baffle anyway, you'll only see that if you go back there and look.

I might borrow a router from a friend and make larger roundovers... I might try the countersink again on a fresh piece if is seems critical. I'm not sure how much it'll help with this driver:

Image not available

There's a lip around the surround that's almost the depth of the rim. It looks like I'll get diffraction off that inner edge anyway, so I'm not sure it buys me anything (besides looks) to flush mount this driver. Also, I'm crossing over to the tweeter around 2kHz, which is a wavelength of over 6" - another reason why I don't think a 1/4" countersink will make a huge difference.

Maybe I'm just trying to talk myself out of needing to do this again with a router I know I'm sacrificing some performance, but I have to balance that against what I think I can realistically build.

The waveguide:



I don't want to flush mount that just because it's going to be a difficult shape to route. Also, the outer edge of the WG is slightly rounded, so if I countersunk that all the way in, I'd actually add a sharp edge where there isn't one now. I think a larger roundover on the baffle edge might help, it'll get it closer to the edge of the WG. And the directivity of the WG should help too. Again, I know I'm giving up some performance, but if it's not a case of "there's no point doing this if you don't flush mount that WG", then I think I'll call it good.

Actually, I'll probably call it good anyway, so I guess I'm just here to get my decisions validated

[Paul Ebert]

In general, countersinking is pretty important for tweeters and mids. Perhaps less so on the tweeter with the WG. An easy way for you to accomplish a similar thing would be to get some wood or fiberboard or something that is similar thickness to your mid frame, cut a hole in the right place and the size of the mid frame and laminate it to your baffle.

[cjd]

You're going to pick up some diffraction off the waveguide edge either way - it'll be better than what you'll pick up not having the mid flush-mounted.

C