Wavecor Ardent Journal - the first Builds

[benthe8track]

So in terms of laminating different materials I came up with this:



Blue is hardboard, green is bamboo, red is birch, and MDF color is 3/4" MDF, pink is 1/2". Some of the 3/4" MDF can be swapped with birch I guess. Any thoughts?
Let me know If I should start a new thread, don't want to hijack

[JonMarsh]

Welcome to the new thread! this topic deserves it's own space; it's benthe8track's baby, with my assistance and participation, given that I proposed the idea a while back and have a lot of interest in it myself!


Now, one of the advantages to the stiffer materials is that we don't need to layer up so much stuff to get good strength. I used this to good effect in the Isiris build (Three Way Design study for now, independent thread pending time and completing the last details of construction). In that system, I was able to reduce the front panel thickness from 4" to 3"- i'd propose doing that here also, as it increases the working enclosure volume for a given outside dimension, and reduces the tunnel effects for the mid woofers and midrange.

In the original Ardent front panels I actually fabricated up the panels in two sections, as this facilitated the necessary bevel cuts on the front.

Here are the original Ardent front panels after the two sections have been glued up:




The main working section of the front panel was 2-3/4" in depth; this was the part cut with the bevels:





Then it was glued up with the "back/inner" 1-1/2" panel section:




One must be creative and work around the limitations of available tools...

[wkhanna]

As shown we have three different materials being used, all with unique rigidity & damping characteristics (I do not include the hardboard since its function & effect appears to be mostly ornamental).

Since each material comes in a thickness where two pieces are used to achieve the desired finished thickness, what, if any effect would interlacing have, I wonder....?


The individual elements and layup as shown:

B = Bamboo
B = Bamboo
P = Baltic Birch (ply)
P = Baltic Birch (ply)
M = MDF
M = MDF


One possible Interlaced iteration:

B = Bamboo
P = Baltic Birch (ply)
M = MDF
B = Bamboo
P = Baltic Birch (ply)
M = MDF

[dar47]

Good question,

Bamboo is hard to source where we are located at 1.5" so the initial thought was to use 3/4" at $250. a sheet. Although we may now get all materials at wholesale , so maybe the bamboo at 1.5" is doable for the outer layer and is best for mounting the drivers. Maybe 3 layers of 3/4". The middle and inner baffle may be reduced in thickness as the total thickness will be base on how aggressive of a facet we end up with as, (closer you are to drivers the larger the facet). Alternating MDF with Birch could be done. My thought was to use birch on the outside and as much as possible to lighten the cab and use 1/2' MDF on the inside for damping. I want 1/8" hard board on the outside to limit telegraphing of the paper backed veneer covering the cabs, only 1/8' joints show. If we stick to 3/4" and 1/2" material it lends itself to CNCing whole sheets for the parts, then build sub assemblies and CNC dados, half laps, facets, holes for driver bolts and etc.

We will try to optimize by sheet to machine as this may be 4 pairs of cabs.

[benthe8track]

As shown we have three different materials being used, all with unique rigidity & damping characteristics (I do not include the hardboard since its function & effect appears to be mostly ornamental).

Since each material comes in a thickness where two pieces are used to achieve the desired finished thickness, what, if any effect would interlacing have, I wonder....?


The individual elements and layup as shown:

B = Bamboo
B = Bamboo
P = Baltic Birch (ply)
P = Baltic Birch (ply)
M = MDF
M = MDF


One possible Interlaced iteration:

B = Bamboo
P = Baltic Birch (ply)
M = MDF
B = Bamboo
P = Baltic Birch (ply)
M = MDF

This is an interesting idea. If the goal of the baffle is based on stiffness then the effects on modulus of elasticity (and subsequent stiffness) then the order in which it's laminated should be immaterial. Although wood isn't homogeneous or isotropic so maybe there is something more to it.

I'm game for ditching the last 1.5" of MDF, my dad and I were discussing the merits of just having a 1/2" layer on the inside. Not sure if there would be any value to this however.

[wkhanna]

I was wondering where the material with the best damping would best be placed, & where the one with the most stiffness should go.
I would think, without consulting my engineering text books, the greatest stiffness would best be placed where the drivers mount.
But beyond that, I am not sure what other advantages could be realized with interlacing from there.

Off the top of my head, it would think having the 'damping' material in or near the middle of the lay-up would prove most beneficial.

As for calling the hardboard 'ornamental', I was figuring you were using it as base to apply what ever finish you were planning to use.

[benthe8track]

I'm curious too, I tried consulting my mechanics of materials book and I think the answer lies above my undergrad pay-grade haha.

I tracked down some BB for 25 bucks a sheet and the bamboo for ~200. I don't think I can do better than that.

[wkhanna]

I'm curious too, I tried consulting my mechanics of materials book and I think the answer lies above my undergrad pay-grade haha.

Ha!
Me too!

I may be over thinking the whole issue and leading you guys off down a rabbit whole tangent!

Sometimes it is better to just make sawdust. :W

[technodanvan]

Given that you plan on covering the baffle with hardboard and that lots of lamination is planned already, I would think completing a psuedo-translam baffle (a la the Ansonica) would maximize the benefits of including the stiffer materials by having them extend through the full thickness. If done this way I would evenly space your materials throughout with bamboo at least dead center and on edge.

You'd have to be pretty careful cutting/routing holes in it though as bits/blades could get beat up pretty badly going from birch to bamboo to plywood and back again.

Also - no lead layer? :P

[JonMarsh]

Hmmmm, if a camel is a horse designed by a committee, what is this turning into? :W

From my personal experience I'm inclined to recommend the hardboard, two layers of bamboo, and two layers of BB ply, or if you which a layer of BB and a Layer of MDF. But really, I'd rather leave the MDF out of the front panel...

Maybe we should structure this like soap box derby- general parameters become the outside enclosure dimensions and the drivers and crossover execution, but each soap box team is free to innovate in materials and construction to their hearts content... I mean, this is an "open source" project, right? :W

[wkhanna]

Thank goodness the voice of reason finally appears!

But I wonder if we could get crowd funding for this project?


I promise to stay in my rabbit whole now.

[JonMarsh]

Thank goodness the voice of reason finally appears!

But I wonder if we could get crowd funding for this project?


I promise to stay in my rabbit whole now.

Automatic spell corrector strikes again.... :B

You're welcome to come out of your hole anytime you want...

[JonMarsh]

Also - no lead layer? :P

Nah, I think we should use reactor grade Thorium instead...

With an FP this thick, lead layer isn't needed. For side walls, it might be IF you were going well up into the midrange, but I don't think it would buy you much here, based on my experience with the first set.

[benthe8track]

Looks like we have access to a 5-axis and a 3-axis to cut the panels/facets. Just need to make sure the envelope is big enough on the 5-axis and figure out how to program it.

Edit: http://www.shopbottools.com/mproducts/5axis.htm
So 34"x34". Would have to do 2 setups per baffle. Might just be easier to make a tool and do it on a table saw JonMarsh style.

[benthe8track]

Had some time to work on the facets for the front baffle. This is taking Jon's recommendation of a 3" baffle made from 2 layers of 0.75" bamboo and 2 of 18mm BB.
On the Ardents the baffles were at 45 degrees, was this to accommodate cutting on a table saw? After playing around with it a bit it seemed to work better using 35 degrees (could get closer to the tweeter), which would have to be done on the 5-axis. Any input is welcome.

[JonMarsh]

There's nothing religious about the 45 degrees... let's just say tradition dating back to the late 70's. But I'd say this looks pretty good. Certainly some of the dimensions were guided by the limitations of my BT3100 saw, which actually has a much deeper cutting depth than most 10" saws (I haven't found one yet that matches it) and deeper even than some 12" saws.

What tool are you using for your CAD work?

As I've got one cabinet converted for the 7" aluminum Illuminators, I'm thinking about mounting drivers and taking new data just for the purpose of evaluating a more Isiris like crossover design.

[cjd]

I went 3 layers of birch ply for the Nebbiolo, which features the same baffle design - I also went 3d bracing, which it turns out does change the good 'ole knuckle-rap test. Whether it's audible in the finished build is an entirely different question - resonant frequencies are well above the frequencies being delivered to the woofers, leading to potential harmonic excitation alone - the energy is probably low enough at this point that even surround or spider induced bobbles are more audible. Or reflections off the extra thick baffle you just can't get away from the driver in the end...

I built the Nebbiolo baffle up a little differently to try to help the last noted problem on my mid - without building two I can't tell if it's helped, but I suspect it will.

C

[benthe8track]

I use solidworks, there are better systems out there but that's what I learned on. This project has been good for cleaning off my CAD cobwebs.
I suppose you wouldn't be able to stuff a wavecor in there?

I did some tweaking and settled on 40 degrees, any more and there isn't enough 'meat' along the edge at the back of the top. This leaves about 0.25" as shown here:



I'm pretty happy with this layout, I tried some more extreme angles as well and the wife did not approve so this is what it will look like:



Next up is the fun stuff like figuring out how to machine a grill that will hold the felt nicely and attach with magnets:

[benthe8track]

I went 3 layers of birch ply for the Nebbiolo, which features the same baffle design - I also went 3d bracing, which it turns out does change the good 'ole knuckle-rap test. Whether it's audible in the finished build is an entirely different question - resonant frequencies are well above the frequencies being delivered to the woofers, leading to potential harmonic excitation alone - the energy is probably low enough at this point that even surround or spider induced bobbles are more audible. Or reflections off the extra thick baffle you just can't get away from the driver in the end...

I built the Nebbiolo baffle up a little differently to try to help the last noted problem on my mid - without building two I can't tell if it's helped, but I suspect it will.

C

I'll take a look, I'll have to rethink the bracing as illustrated in the first picture in this thread anyways. With the 3" baffle I think there will be interference with Wavecors and the brace as drawn.

[cjd]

I didn't plan bracing as carefully as I should have and ended up doing a little post-assembly surgery to fix it so the braces cleared the magnets. I glued up each baffle layer to the box separately though, so I did this work before the face layer was added. I did this because I don't have CNC and the bracing wasn't perfect.

The boxes I have will not pass any real inspection for top notch assembly - the sides are not flat, but slightly wavy, and I've given up trying to fix it more than it is. Lessons learned for next time I guess.

C

[benchtester]

I use solidworks, there are better systems out there but that's what I learned on.

And a lot worse, Pro-E for example. Personally I don't know of a better CAD system than SolidWorks for someone not using it 8 hours a day. (Fair disclosure: employment issues made me go from Solidworks to Pro-E in 2009, and I still don't like it. I will be using Pro-E for all my projects, because it requires all the practice you can get.)

[JonMarsh]

For anyone curious, Solen shows having a dozen of the SW223BD01 in stock; I'm going to order another pair some time in November.

https://www.solen.ca/pub/index.php?catalog=description&nocol&nodiapo&s1=1&s2=&s3=&id=2781&page=2781&nobut=1

[JonMarsh]

Some modeling verification in Unibox:

Base model and response; single unit parameters verified from input data:



This is the Unibox predicted anechoic response (no boundaries) for dual drivers in 40L, fully stuffed, with 100W input, for one cabinet. With boundary effects this will be lifted 6 dB typical below 40 Hz.




Just seeing if using the supplied vendor parameters results in a model that agrees for all the red items in Unibox is a pretty good sanity check on the vendor supplied T/S info; you might be surprised how often that doesn't work out...

Last, the calculated step response, which sure looks nice compared with ported alignments...

[dar47]

For anyone curious, Solen shows having a dozen of the SW223BD01 in stock; I'm going to order another pair some time in November.

https://www.solen.ca/pub/index.php?c...e=2781&nobut=1

Jon,

For the 8 cabs we are doing that's 14 drivers. I'll contact them and see if we can get a good discount and I'm going to see how they are on the mid and tweet. I'm buying soon and the other 2 may go for the woof's if it's discounted.

For the cabs we got so far with no adjustment to the model 0.35 ft^3 for the mid chamber and 1.5 ft^3 subs. If you can let Ben know if you want some adjustment.

I did some prelim sheet counting and the material should be vary cost affective. I'm trying to hang you with not much more them shipping .

[dar47]

Ah, should have looked before first response, at 0.73 ft^3 we are right there!

[JonMarsh]

Ah, should have looked before first response, at 0.73 ft^3 we are right there!

Yup! I'd thought this out pretty carefully in research, and the SW223BD01 is pretty optimum in the original Ardent LF volume.

OK, I'll hold off on the drivers until we can sort out the best Qty deal for everyone put together- I'm glad to see that Solen has this many in stock right now, when I bought previously they just had a couple.

[benthe8track]

I double checked the volume and we are a bit over 40l due to the larger baffle required. I'll do some tweaking to shoot for 40L for the woofers. Was the target for the mid 0.35 ft^3?

[JonMarsh]

A little over is no problem at all- could help, what with the volume of the crossovers

[benthe8track]

Finally had some time to get back to the model.
Reworked the bracing to clear the drivers. Gave a 0.25" clearance around the drivers:



Just realized I need to add some clearance for the drivers (between the baffle and outer edge). When doing it by hand with a jasper jig I typically just rounded up to the next 1/16th. What's a reasonable value to use, maybe 1/32?






The model is almost done at least, just need to experiment with the magnets for the grill to see if we can put them under the hardboard. Then need to finish up the insert locations for the base and we can move on to programming the cnc.

[benthe8track]

And a lot worse, Pro-E for example. Personally I don't know of a better CAD system than SolidWorks for someone not using it 8 hours a day. (Fair disclosure: employment issues made me go from Solidworks to Pro-E in 2009, and I still don't like it. I will be using Pro-E for all my projects, because it requires all the practice you can get.)

I've only tried Pro-E once or twice. I thought it was supposed to be the cats ass once you get over the steep learning curve. I'm with you, SolidWorks is great for how intuitive it is. I used it in my first engineering summer job years ago and basically BS'd my way in there saying I knew how to use it. Took the books home and watched a lot of youtube to get up to snuff.

[dar47]

I like it in wood. Jon, let use know if your happy with 0.35 ft^3 for the mid?

Image not available

[meb46]

Excellent work - me likes!

[knowledgebass]

I'm curious to know how much room you've left between the grill frame and baffle edges? From your image in Post#18 it's not clear where you are planning to put the wool felt as the frames appear to be flat with woofer cutouts? May have missed it elsewhere in the thread. I'm currently working through this exercise on my build (CJD's Ansonica) and came up with 1/2" minimum frame thickness for strength/stiffness which I based solely on my experience with MDF - having never built a speaker grill. I feel like that might be too conservative? MDF is cheap but my time isn't so hoping to get it right the first time even if I'm a little overly conservative in my construction. I've also attempted to leave a little extra between the frame and drivers or tweeter for wool felt resulting in facets that maybe weren't as steep as I had originally planned in order to surround the drivers with some amount of will. I'm planning similar facets for my translaminated baffle and my MDF grill frames will be fabric wrapped.

I'm also using rare-earth magnet fasteners in the form of a ring and rod magnet. The ring will be embedded into the baffle and the rod will pass through a 1/4" hole in the baffle. I've not seen other baffles mounted this way so I'll try to get my post describing what I'm planning up this week with some photos. My techniques are somewhat more modest, not having access to a CNC

[dar47]

I'll throw my thoughts so far, I still have to review with Ben.

1. Grill thickness is 3/8" HDF, I have used it and it's less susceptible to the grill warping. You can bevel or round over the outer edge to make the grill look even more thinner.

2. The grill will be cut 1/8" shy of the facet edge, seems close to Avalon.

3. Felt will be inset in the grill surrounding just the tweet and mid. It comes to just outside of the outer driver recess hole. If we use 3/8" felt it will have to be attached to the Grill cloth with 2 side tape. If we use 1/4" felt it can be machined in the grill. I haven't seen a Avalon grill and maybe Jon has other ideas.

From my experimenting, grill magnet just under the cab veneer facilitated just metal contact attached to the grill, this was great for 6 contact points, not 4. For these I wanted to put a magnet under the veneer and 1/8" hard board (outer baffle layer) and a magnet imbedded in the grill. Ben the Mech. Eng. guy tells me he can calculate the forces.

Never heard of rod and ring mags, Yes please post sounds vary interesting.

[JonMarsh]

I like it in wood. Jon, let use know if your happy with 0.35 ft^3 for the mid?

I'm happy- and it's fairly non-critical given the crossover point.

[JonMarsh]

Finally had some time to get back to the model.
Reworked the bracing to clear the drivers. Gave a 0.25" clearance around the drivers:



Just realized I need to add some clearance for the drivers (between the baffle and outer edge). When doing it by hand with a jasper jig I typically just rounded up to the next 1/16th. What's a reasonable value to use, maybe 1/32?






The model is almost done at least, just need to experiment with the magnets for the grill to see if we can put them under the hardboard. Then need to finish up the insert locations for the base and we can move on to programming the cnc.

How accurate do you expect the CNC process to be? Then, how accurate/consistent are the driver dimensions? (we don't really know...) There's no advantage for the woofer rebates (especially covered with a grille) to have them ultra close fit, and it can sometimes make driver installation and removal a little more tricky. My thought is that 1/16" sounds fine...

Just my 0.02

BTW, CAD work looking very nice! :T

[dar47]

Thanks, Now I'll start laying out cut sheets for machining and get our materials ordered .

[benthe8track]

I'm curious to know how much room you've left between the grill frame and baffle edges? From your image in Post#18 it's not clear where you are planning to put the wool felt as the frames appear to be flat with woofer cutouts? May have missed it elsewhere in the thread. I'm currently working through this exercise on my build (CJD's Ansonica) and came up with 1/2" minimum frame thickness for strength/stiffness which I based solely on my experience with MDF - having never built a speaker grill. I feel like that might be too conservative? MDF is cheap but my time isn't so hoping to get it right the first time even if I'm a little overly conservative in my construction. I've also attempted to leave a little extra between the frame and drivers or tweeter for wool felt resulting in facets that maybe weren't as steep as I had originally planned in order to surround the drivers with some amount of will. I'm planning similar facets for my translaminated baffle and my MDF grill frames will be fabric wrapped.

I'm also using rare-earth magnet fasteners in the form of a ring and rod magnet. The ring will be embedded into the baffle and the rod will pass through a 1/4" hole in the baffle. I've not seen other baffles mounted this way so I'll try to get my post describing what I'm planning up this week with some photos. My techniques are somewhat more modest, not having access to a CNC

I think my dad touched on most of it. In that model the grill is in contact with the baffle. I am looking at some thin felt from mcmaster.com to surround the edge of the hardboard frame where it will come in contact with the baffle. I need to open the grill up a bit so the thicker felt can rest inside. I have an idea how I'll support it but haven't drawn it yet. I've seen lots of great grills made by hand, no need for a CNC
Please post up more info about those magnets I'm curious.

How accurate do you expect the CNC process to be? Then, how accurate/consistent are the driver dimensions? (we don't really know...) There's no advantage for the woofer rebates (especially covered with a grille) to have them ultra close fit, and it can sometimes make driver installation and removal a little more tricky. My thought is that 1/16" sounds fine...

Just my 0.02

BTW, CAD work looking very nice! :T

Thanks! I keep fine tuning it the models, it's not so bad when you're cutting MDF but I don't want to scrap a $200 sheet of bamboo.
The machine claims a positional accuracy of +/- .002 and a step resolution of .0004” but I'll believe that when I see it. I would be impressed if a HASS mill would hit that. I changed the gap to 1/16th around the drivers but may bring it in just a bit depending how the machine preforms.

[benchtester]

The machine claims a positional accuracy of +/- .002 and a step resolution of .0004” but I'll believe that when I see it. I would be impressed if a HASS mill would hit that. I changed the gap to 1/16th around the drivers but may bring it in just a bit depending how the machine preforms.

1/16" is huge (radially?) You shouldn't have any problem with .010" radial (.020 diametral). I have cut (only) a few baffles with CNC. My actual holes were about .005" small, which I suspect is mostly due to tool bit issues. I ended up with a slight press fit, which looks great; although I know it's not quite right. Next time I will learn how adjust the program on the fly to compensate for the cutter.

[wkhanna]

The machine claims a positional accuracy of +/- .002 and a step resolution of .0004” but I'll believe that when I see it. I would be impressed if a HASS mill would hit that.

My actual holes were about .005" small, which I suspect is mostly due to tool bit issues.

I have been in machine shops for years, currently as quality & production engr where we make medical implants, among other things.

Maintaining tolerance over an entire production run on CNC mills & lathes or wire EDM of better than .002” is common for any precision machine shop. We have some jobs were the total tolerance is one-half of .001 (+/-.00025).

When dealing with such tight tolerances, as benchtester discovered, tool wear is actually more critical than machine capability.

[Finleyville]

I have been in machine shops for years, currently as quality & production engr where we make medical implants, among other things.

Maintaining tolerance over an entire production run on CNC mills & lathes or wire EDM of better than .002” is common for any precision machine shop. We have some jobs were the total tolerance is one-half of .001 (+/-.00025).

When dealing with such tight tolerances, as benchtester discovered, tool wear is actually more critical than machine capability.

Experience always trumps theory. BTW, I though you were staying in your rabbit hole?

[JonMarsh]

When dealing with such tight tolerances, as benchtester discovered, tool wear is actually more critical than machine capability.

This I would expect. Then, one also has to consider material dimensional stability with regards to temperature and humidity- wood products are in a different class there than aluminum, for example!

[meb46]

Don't forget the thickness of your final finish... i.e. Paint/Lacquer can build to a significant thickness and the last thing you want is to mess it up when you install the drivers due to lack of clearance.

[dar47]

Don't forget the thickness of your final finish... i.e. Paint/Lacquer can build to a significant thickness and the last thing you want is to mess it up when you install the drivers due to lack of clearance.

Good point, I cut 4 routing jigs to rout my C90 in my CC with the driver in hand. Each 1 was a closer fit and the last 1 had to be careful screwing in the C90 not to damage the finish. I think we will try a test run on some BB at 1/16" over for the subs (your not going to see a gap a 1' away, at least not me with my bad eyes) and a 1/32" for the mid and tweet. I have those 2 drivers to see the fit.

[dar47]

1 other thing I'm going to test is the half laps, 1 with the d1mention right on and 1 adding a 1/32" to the lap just to see how good the edges on the backs to sides line up. The dados are easy add a 1/16" for fit.