Nebbiolo: Another 3-way experiment

[JonMarsh]

http://www.parts-express.com/pe/show...&FTR=w4-1320sj

Nevermind. PE just being daft, and that stupid tiny "F" at the end. Impedance is completely different. Oh well. I suppose the thing to do is snag some of this new ferrite model and use that instead.

Sigh.

I was just joking at the Northern CA DIY during my presentation that as soon as you finish a design and lock every thing down solid, at least one or more of the drivers will go obsolete- this brought a round of laughter from everyone including Siegfried- but it is a special group for which the driver goes obsolete during the development... Welcome to the club!

In good related side news, Charlie has a couple of NS12-513's it looks like he's willing to sell me, so I have some back up on my woofers.

Good luck getting your TB's sorted out!

[cjd]

Interestingly, since I've been taking the slow lane on this, it seems PE is again carrying the Neo version so that's back in the game. Now that I have the ferrite versions in-hand...

Is that extra extra special? :P Or special revoked?

[TEK]

I was just joking at the Northern CA DIY during my presentation that as soon as you finish a design and lock every thing down solid, at least one or more of the drivers will go obsolete- this brought a round of laughter from everyone including Siegfried- but it is a special group for which the driver goes obsolete during the development... Welcome to the club!

Wonder if that might bee seen as a suttle reminder telling someone that the timeframe for the projects beeing a bit to long :W
But as we all know - slow work takes time 8)

[cjd]

Getting closer.



I have to figure out how I'm going to do the wiring before I seal them up all the way I think. I may also go ahead and take some measurements before I shape the baffles, just to see just how much difference the chamfers make.

[---k---]

Progress. Very brave to do the chamfers afterwards. :T

[JonMarsh]

Progress. Very brave to do the chamfers afterwards. :T

Adds to Olympic degree of difficulty! :W

[JonMarsh]

Wonder if that might bee seen as a suttle reminder telling someone that the timeframe for the projects beeing a bit to long :W
But as we all know - slow work takes time 8)

Very much so- and I was holding that mirror up in front of myself!

[cjd]

port plug insert glued in and planed flush. Wire for the back terminals to a terminal block (I had just enough left of the 12ga versions from ApexJr) - I'll run from that to the crossovers since it'll be sealed up soon. I had spent a little time sharpening up the plane and tuning it - works noticeably better now. That's how I'll be finishing the bevels - rough-cut with a circular saw and final cleanup and plane to the mark by hand.

Progress.

C

[---k---]

Truly hand crafted.

[cjd]

Backs buttoned up! Well, glued up really... Some measurements pre-chamfer up next I think.

I think this is the most nerve-wracking part because the back is truly closed up for good; there's just no easy way to change anything behind the vertical brace I have (goes side to side, in addition to the one front-to-back and the shelf braces.) The chamfers don't seem to be scaring me like maybe they should.

C

[cjd]

Pretty nice response is do-able with first order electric everywhere but the mid low-pass, which does OK with just one component but prefers two. This is pre-chamfer, so not final, targeting 500/2000Hz crossover points.

Unfortunately, that means impedance compensation like crazy; zobel + RLC on mid, RLC on tweeter, zobel on woofers. As you might imagine, the RLC on the mid is a problem: 12mH inductor, 200uF cap!

It appears to be largely acoustic 2nd order Butterworth.

Further experimentation tightens up some things, but not everything. That RLC on the mid does seem rather important (not surprising given the Fs and the slopes.)

More experimenting forthcoming...

[JonMarsh]

Pretty nice response is do-able with first order electric everywhere but the mid low-pass, which does OK with just one component but prefers two. This is pre-chamfer, so not final, targeting 500/2000Hz crossover points.

Unfortunately, that means impedance compensation like crazy; zobel + RLC on mid, RLC on tweeter, zobel on woofers. As you might imagine, the RLC on the mid is a problem: 12mH inductor, 200uF cap!

It appears to be largely acoustic 2nd order Butterworth.

Further experimentation tightens up some things, but not everything. That RLC on the mid does seem rather important (not surprising given the Fs and the slopes.)

More experimenting forthcoming...

Yup, all those zobels sound like what I was going through with the Duelund crossover approach, too... pardon the pun...

[cjd]

Well - going to LR2 slope targets at 800Hz and 3000Hz (and getting pretty close through -40dB) gets flatter and cleaner, but also does away with the RLC on the mid. Tweeter gets away with a cap and RLC, optional resistor (.5ohm to 1ohm looks like it may be nice.) Mid is two component high-pass and low-pass, RC, and a bit of padding (~1ohm out front so tunable to taste.) Woofers are zobel + two components (LC in parallel - C is .25uF though...) None of the values are super silly, though the parallel inductor for the mids isn't exactly small at 5.5mH. Not big either.

What is interesting about this one is that, were one to want to build a flat-faced box that stands vertically (no slant backwards) it only takes some tweaking on the tweeter levels to be reasonably flat - switching to being on-axis instead of ~6deg off-axis on the tweeter is the probable cause. Phase integration isn't quite as snug, but it's workable - and optimization for that axis should be do-able. Off-axis is similarly solid (I took one ~20deg off-axis data set today.) That means that overall power response should be good and around-the-room sound should also be good.

I ran into the problems I expected trying a minimal component crossover - it really does want relatively ruler-flat impedance or everything just falls apart.

I won't be running out to buy components just yet, but this is promising. LR2 slopes is still quite low for me. I think some of the trouble-spots running the tweeter at a 1st order slope will change once the baffle chamfers are cut, so this will change.

I'm sure I'll keep fiddling. This is fairly traditional 3-way so far, after all.

[cjd]

What the heck. Early sketching...

1st order electric with lots of impedance management mentioned:


The LR2 variant:

[cjd]

The (a?) day of reckoning has arrived...

pics with chamfers cut soon I hope! (and more fervently hope for no mistakes!)

[cjd]

A little detail of the wiring here.



Chamfers.

[TEK]

Any pictures of how you did the chamfers?

[cjd]

Any pictures of how you did the chamfers?

Nothing fancy. Clamped 1/4" hardboard as a guide for a circular saw, lots of care setting it up, and finish with a hand plane. No pictures though. Didn't even occur to me. The saw was only deep enough to cut about 90% of the depth, but the plywood separated off cleanly witnout tearing, so the plane took care of it easily.

C

[cjd]

Veneer showed up!

[---k---]

You're actually going to veneer them this time? :P

[JonMarsh]

Veneer showed up!

What kind?

[meb46]

Sweet build... Looking across the current forum builds it would seem we have our own complete Avalon Line-up

[cjd]

Yes, I'll really be veneering these.

It's lightly figured cherry (except for the backs which get a nice but cheap plain cherry) - I'll be attempting cold press, so I have platens in progress (some of which have been around and were used for the glue-up of baffles and back and have worked quite nicely.)

I'll probably spend this weekend preparing the surfaces. Not sure if I should be attempting any kind of sealing prior to veneering or not either (and if so, how,) so a bit more research is in order.

I'll do the bottoms first, and if that fails have a quick abort and regroup handy. If it seems to go well, I'll proceed to a back, after which there'll be another analysis...

Meanwhile, my projector lost half its control circuit (left side DMD controller? something like that - I guess it's a common problem in the BenQ W5000,) but ~3000 hours is still ~1/4 the price of just theater tickets for all these movies we've watched... So evaluating my options. The good news is, my wife's first response was "Shit! How fast can we get a replacement?!"

[PMazz]

You're going to have to do some Bondo type work on the facets and any other edge grain. Plywood laminations will telegraph thru the veneer otherwise. Another option on the facets would be to add some vertical grade (.03") laminate or maybe 1/8" hardboard first and veneer over that.

[cjd]

You're going to have to do some Bondo type work on the facets and any other edge grain. Plywood laminations will telegraph thru the veneer otherwise. Another option on the facets would be to add some vertical grade (.03") laminate or maybe 1/8" hardboard first and veneer over that.

I assumed the end grain would be an issue, but wasn't aware it took that kind of prep or I might have approached this slightly differently - I guess I'll have to find some time when I can do that work outside (it's pouring at the moment.) I hate bondo, but it does the job well so I guess I can't complain. That might make it a "not this weekend" project though.

[TEK]

my wife's first response was "Shit! How fast can we get a replacement?!"

That's the right kind of reply :-)

[Steve Manning]

I hate bondo, but it does the job well so I guess I can't complain. That might make it a "not this weekend" project though.

cjd,

With the bondo work I did on my cabinets I had some U-Pol Fibral Lite recommended to me by a paint shop, when I ran out of the regular bondo. The stuff is much easier to work with than regular bondo, it stays pliable longer and goes on smoother and can be built up thicker. It has some short hair glass fiber in it so it ends up being stronger as well. It also sands nicely and leaves a very smooth finish. I turned it on to my paint guy and he loves the stuff.

Steve

[cjd]

For painting, I know those all work well - I'm just not sure how they bond for veneer adhesion with a cold press glue.

[Steve Manning]

For painting, I know those all work well - I'm just not sure how they bond for veneer adhesion with a cold press glue.

I believe it should work fine, though I would not recommend going for a super smooth finish on the bondo. Use either an 80 or 100 grit sander paper and you will leave a better bonding surface for the glue on the veneer. I ran into that usefully tip on Joe Woodworkers site after I got the bubbles in the veneer from poor bonding. He actually recommends it on the back of the veneer as well if you are using paper backed veneer.

[---k---]

You're wifes response to the projector is awesome. She's a keeper.

Sent from my DROID2 GLOBAL using Tapatalk 2

[cjd]

I finally risked further back pain, hauled stuff up and got some surface prep done. I'll be sore, but if I keep up the exercises it should remain simple muscle soreness.

May get a first-whirl at veneering the bottom very soon indeed.

C

[cjd]

Finally got all the pieces together and a bit of time (and not a little courage) and tried my hand at applying veneer. I've never done this before, so really wasn't sure what to expect. This is the cheap veneer I'll be using for the bottoms and the backs, with enough to start over should I mess something up. I think I may be alright.



Cold press, unbacked cherry.

[benthe8track]

Looks good! I'm curious to see how you prep the facets for veneer as I'll need to do something for those Wavecor Ardents. Probably 1/8" hardboard but even considered using the car audio trick of the bondo and epoxy "milkshake".

[cjd]

Looks good! I'm curious to see how you prep the facets for veneer as I'll need to do something for those Wavecor Ardents. Probably 1/8" hardboard but even considered using the car audio trick of the bondo and epoxy "milkshake".

Those are prepped. The base had a similar 3 layers of exposed plywood edge, and at the moment it's clean. I'll give this a little more time before I veneer the baffles (it's going to take me time to get there anyway, no way around it) to be sure the prep is adequate and the plys aren't telegraphing, but based on surface feel they should be good. I followed Pete's advice earlier in the thread and went with "bondo type work" - here's hoping it really was up to the task. But, that's yet one more reason I did the bottom first.

[dar47]

Sorry wrong place

[WilZirkle]

That's some good looking cheap veneer Chris. Nice job.

[JonMarsh]

That's some good looking cheap veneer Chris. Nice job.

+1 on that. The only delicate thing I hear about working with Cherry is it's photo sensitivity; you might want to drape the speakers when you're not working on them, keep the light off of them. Otherwise, looking good- good enough I'm going to have to go back through your thread and see if you mentioned where you got your "cheap" veneer!

[JonMarsh]

Looks good! I'm curious to see how you prep the facets for veneer as I'll need to do something for those Wavecor Ardents. Probably 1/8" hardboard but even considered using the car audio trick of the bondo and epoxy "milkshake".

Thinking along the same lines. For my original Ardent's I did cold press veneer; when it worked, it worked very well...

[WilZirkle]

Jon, I was thinking about that too. Depending on where he ends up letting these live, the cherry on the bottom may never match the rest of the project again.

[cjd]

Too well do I know the joys of Cherry and its light sensitivity. I'll probably do a shellac wipe through a bunch of layers and finish with an oil/poly blend to finish - and possibly get them LOTS of light exposure before I apply that finish. I have to see where I can source danish oil - I'd like to do a 1:1:1 danish/tung/poly but we'll see.

Veneer is from veneersupplies.com It's not straight its full length (but straight enough it'll work cleanly the 42" of the back) and it has some spots that go through - worm holes or something? Not exactly sure. The baffle/top will be the real test of my patience since I'll be book matching that, and I also have to account for the facets - it's a rather nice figured cherry though.

C

[cjd]

These are going to be interesting...

I took new measurements today, including off-axis. Modeling a preliminary crossover, and picks up a ~3dB tilt from 200Hz to 20kHz at 30 degrees off-axis, but otherwise remains scary flat on-axis through off-axis. Well, and the very top end of the tweeter takes a hit off-axis (15kHz+, this is an OW1 after all)

2nd order LR.

There's a pretty big dip on the woofer response that's there every way I measure - different distance, on or off axis vertically or horizontally. I guess it's consistent enough I can rely on it but it's a little unexpected.

I need to order some parts, I really want to hear this.

Pics just need to be transferred when I'm not otherwise occupied - right now I'm listening to music. They're not finished yet, but working up layers of shellac.

[cjd]

I think this is where I'll start on a real crossover build and tweak from here. Roughly LR2, appears to be 950Hz and 3kHz where things seemed to settle in. I may try to push the mid-woofer lower before jumping in, we'll see. I'll definitely be ridiculed in some circles for the parts count Lots of impedance compensation in the mix.

On-axis measurements and response, with mid reverse null


15 degrees off-axis horizontally (modeled response on real off-axis measurements)


30 degrees off-axis horizontally (modeled response on real off-axis measurements)


Nulls from woofer or tweeter reversed alone


Impedance

[cjd]

Too lazy to get the real camera out so not quite as good a photo as this deserves but...

Assembled so I can do measurements


Still working on finishing - a way to go, but I think it'll turn out OK

[JonMarsh]

Nice work, Chris! :T Does that house curve you've got overall work for you? I bet they sound nice with a lot of program.

Is this how you're finishing up the finishing, as you'd originally planned? Or has this been updated?

Too well do I know the joys of Cherry and its light sensitivity. I'll probably do a shellac wipe through a bunch of layers and finish with an oil/poly blend to finish - and possibly get them LOTS of light exposure before I apply that finish. I have to see where I can source danish oil - I'd like to do a 11 danish/tung/poly but we'll see.

How do you plan to set them up in your room?

Happy New Year!

[cjd]

There's not much house curve is there? Overall, I'm not sure. I tend to go for pretty flat overall - these are a touch hot on the top end on-axis but that balances out well off-axis. We'll see. The crossover points aren't what I'd expected to target, but it's where things seemed to come together. I may try other variations as well, but gotta start somewhere. I had to order more parts so haven't been able to throw anything together and start listening yet.

I'm not even sure what room these'll be in, so even that is up in the air. Pending some of the other projects I have in the queue these may end up as gifts in the end, too. They'll probably be set up in front of my HT screen for the initial listening (that wall is lightly treated, and the room is carpeted) and voicing, with some trials elsewhere if things start to sound good. The HT setup would put them ~2.5M away probably 2M apart with some room to fiddle. That keeps them a bit away from any walls too.

Finishing, I'll be doing shellac with the oil/poly finish, though I can only do the shellac with what little ventilation I have. The rest will have to wait till spring. ~10 coats of that I suspect (the pics are with 2 coats) and they should look pretty dang good when that's done, but the final oil/poly seems to reduce the need for manual polishing, and gives them a bit better waterproofing - whether in my house or as a gift there's a high possibility these will also be plant stands sometimes... I'd gotten my oils mixed up before (danish is a mix already) but it'll probably be a plain poly/tung oil.