How to make a grown man cry .....

[Steve Manning]

The answer .... after spending a rather long time building a rather good sounding and nice looking pair of speakers ....



You discover the lamination's on your cabinets are coming apart, right at the glue joints. I used Titebond III, which I had heard good things about, and after ~14 months the stuff is coming apart. At this point I don't know if it was the wrong type of glue to use for a translam project, I didn't use enough, though I thought I did, or what. At this point I'm rather bummed because I don't think there will be a way of fixing things and have the fix not be rather apparent.
So anyone got any ideas what I did wrong so I don't repeat it in the future? Thanks for any suggestions.

Steve

[ThomasW]

Can't see a problem. Is there room on the inside for you to layup a thickness of epoxy and fiberglass matt?

[vinceb]

That's got to be hard, what a beautiful cabinet. If you drink beer, go get one.

I had a bit of issue laminating with TB1 where I think I was too slow assembling and maybe not enough glue, I think it's okay but on the edges you can see some small gaps. It made me appreciate Gorilla Glue for this stuff due to the foaming action.

Maybe you can strategically re-glue the bits and clamp it and see what happens. If you wipe it well while drying it might not be that bad on the finish. Worth a shot anyway...

[---k---]

I can't see the problem in the photo either, but I'm sure it sucks. Those look amazing.

I've never attempted a translam project like that, so I don't have any direct experience. But in this months Woodsmith magazine that I got yesterday, it has a short article on glue. Not sure if it applies, but in the side bar there is this paragraph:

When it comes to glueing up a bent lamination, yellow glue will work, but you can expect some sprinback due to the elasticity of cured PVA glue. When you need to avoid the "creeping," there's another glue I like to use - urea formaldehyde.

[Steve Manning]

Clarification .....

Sorry guys, that picture is what the speaker used to look like, today I could stick a 6 inch metal ruler through one of the cracks all the way into the speaker.
Vince, I've already had a beer, I'm eyeing the rum as a follow up ....
Thomas, there is no room on the inside for what you have suggested, though I'm thinking of trying a clear epoxy to fill the cracks on the outside and then redo the finish. Not sure how that will do though.
K, I know David Marks uses plastic resin glue for his bent wood laminations. It's supposed to have longer air time and sets up very strong. I'll have to find a copy of that magazine and check that article out.
Steve

[Dennis H]

Wow that sucks! Are the cracks near the baffle? Maybe it cracked because the plywood shrank in thickness and something had to give with the solid baffle. Maybe floating the baffle on the front rather than gluing it solid could help prevent that in a future build.

I think your idea of stripping the finish, working epoxy into the cracks and refinishing is a good one. Even if the epoxy isn't clear, it will just look like another horizontal line among many and it won't really show.

[ThomasW]

If you remove everything from inside the box it looks to me like there's enough room to at least partially fiberglass the inside....If you do that with the unit clamped together it's IMO your best bet. Then redo the outside with a clear coat of epoxy.

Yes it will be a stinky mess but it may just save the project.

[gmed]

I know how you must feel. It took me more than a year to finish my project. This is why I used between 20-30 nails (using a gun) on each layer. And I used regular wood glue. I didn't rely on the glue alone.

It must feel like crap. I honestly dont think you can do much to salvage it to something like it used to be.

[DancesWithBeers]

Steve,

I'd look more into Dennis H's suggestion of wood shrinkage/expansion issues.

I'd be inclined to think the hardwood on front and back ripped the enclosure apart, since this type of lamination isn't under much stress, relatively speaking.

Is the baffle still attached to the plywood securely? If so, if the total height of the plys behind the baffle with gaps is still the same height as the baffle, then that would be an indication that the enclosure was ripped apart by expansion of the wood. No glue on earth would have prevented this if that's what happened.

If expansion was what caused this, it should give some hope; if you can detach the baffle. Without baffle, you could maybe squeegee some glue into the cracks and re-clamp the whole thing. I'm not sure how complex your baffle attachment is, though.

If attached with just a simple butt joint, you could get a bunch of safety razors (the kind with one sharp and one edge with bent metal around it) and slowly hammer these in around the gap between baffle and speaker to break the joint. If the joint is more complex, and there is not a way to remove, the baffle, then you might want to consider a wood filler or some other suggestion.

If the wood delaminated for some other reason, then I am, frankly, frightened for my own project and surprised. The glue joint with Titebond III should be stronger than the wood itself when cured, so if it did come apart due to driver stress, then something is very wrong. Was the temperature above 47 degrees F. during glue-up? Were the speakers in direct sunlight or exposed to high heat?

Hope this helps and keep us updated. At the very least, besides the very painful loss, this is an interesting case study.

-DWB

[Rolex]

Almost all expansion in wood will be perpendicular to the grain. This is dependent on how the wood is cut from the log, so I am speculating here. Based on that, I doubt it is it the front baffle causing problems. I used titebond II in both of my stacked lamination projects. I would say I used a fair amount.

Regarding repairs, you can use a filler in the epoxy that has similar tones to the plywood you used. That might help with hiding the repair.

How did you seal the cabinet?

[Paul Ebert]

Rolex is right about the expansion being along the width of the speaker, but it may still be the culprit. The expansion will vary linearly with the amount of wood along the width, so where there are cut-outs, there will be less expansion. Perhaps this variation is resulting in some forces in the vertical direction. Just a hypothesis. Do the cracks show any pattern with respect to their location along the height of the speaker?

Also, sometimes too little or too much clamping pressure can cause joints to fail (too much can 'starve' the joint of glue). Only you can tell if this might have happened.

Did you glue up all of the laminations at once or did you do them in stages? If you did it all at once, you would have needed quite a bit of clamping pressure to give each joint enough. There might have been an unequal distribution of pressure. Again, a pattern to the failed joints might give a clue.

Regardless, what a drag after so much work and to such a gorgeous speaker.

[fbov]

Steve,
I had a similar issue with gaps showing after glue-up. I found I could wedge the gaps enough to get glue deep into the gap (waterproof sandpaper was thin enough) and since the exterior wood was finished, the partially-dried glue came off like rubber cement.

In your case, are the gaps even your fault? I would not be surprised if the plys were de-laminating.

Have fun (eventually)
Frank

[fjhuerta]

That's a gorgeous speaker... save it!!!

[CupCak3]

Can you take some close-ups of the problem areas?

Thanks.

[Hdale85]

Sounds like the wood shrank or expanded..... maybe the quality of birch ply you used was not the greatest and they started to split and come apart. I think throwing some clamps around it to get the openings back together and then fiber glassing the inside is a good idea.

[ThomasW]

I think throwing some clamps around it to get the openings back together and then fiber glassing the inside is a good idea.

Yeap with one exception. Here's what I was typing as you posted....

With the insides of the cabinet gutted and no clamps on the cab.

Make up a batch of epoxy (a lot of it) pour that into the cabinet.

Next rotate the cab to coat the entire inside surface, make sure some goes into any open gaps.

Now clamp the cab to close the gaps.

Then layup the inside using small strips of precut fiberglass mat (fiberglass cloth is actually a better choice than since it's inherently stronger than mat).

[jkrutke]

I watch this thread with interest. I worry the same thing may happen some day with my attempt at a lamination speaker. I used good old yellow Elmers wood glue, with high and even clamping force. Something I may consider adding is a couple turnbuckles on the inside, pulling the top to the bottom. It should massively enhance the rigidity of the top and bottom too.

One thing that may help in my case is that I allow the front and rear baffles to float. They are removable and held in place with bolts through clearance holes. One good sign is that my system has survived massive temperature changes as these go in and out of my cold Wisconsin garage. (1.5 years working on these and they still aren't done)

[ThomasW]

(1.5 years working on these and they still aren't done)

Slow work takes time..... :B

[Steve Manning]

Hi Guys,
Thanks for all the great support and ideas. Here is what the major crack looks like.



To answer some of your questions and ideas .... Thomas I like your idea about gutting the speaker and filling things with epoxy, but have a couple of road blocks. The first is the walls have sonic barrier attached. It has its own adhesive backing and then I stapled in place just to make sure it did not come off. I also did this with the top and the bottom of the speaker off.



The idea of trying to get that out and then reinstalled might be a real pain. I also found after trying to fill the crack with glue and clamp it closed, is that I could not get the gap to move at all, even with 3 clamps!!
So what I'm trying at this point is filling the crack with some epoxy from the outside and see what happens. I figured the sonic barrier will act as a dam on the inside so I can fill the crack level.



With regard to the general opinion on the baffle expanding ..... Here is how it is put together. To help with reducing expansion issue the padauk is mounted to mdf which should be very stable.



As you can see the backside is doweled and the entire baffle is glued into the cabinet.



I recently took a class with David Marks of DIY Wood Works fame and we talked about wood movement. My understanding from our conversation is that you will get a fair bit of motion across the board but very little over the length. This is also the reason I mounted the hardwood to the mdf, to help stop movement. I did not have any problems with the padauk cupping at all once that was done. Of course it might not take much in this case.

DWB, your question about the temperature is very valid, my wife and I have already been discussing that. I am fairly certain I did all my glue up's above 47 deg., but ..... I have a nagging feeling in the back of my head that the bottle of glue that I used on that cabinet, being stored in the garage, might have gone below freezing at some point during the winter. Reading the bottle yesterday it states, in bold, not to let it freeze. So this is a possible culprit, I will have to wait and see if the other cabinet follows the self destructive path as well.

GMED, good idea on the nails. If I go this route again I will throw in a few for good measure.

So at this point I'm waiting on the epoxy to set and thinking through some of your ideas. I'll let you know how to goes.

Steve

[DancesWithBeers]

Almost all expansion in wood will be perpendicular to the grain. This is dependent on how the wood is cut from the log, so I am speculating here. Based on that, I doubt it is it the front baffle causing problems. I used titebond II in both of my stacked lamination projects. I would say I used a fair amount.

Rolex, you are 100% correct, but, from what I remember, even the small amount of expansion in the direction of the grain that occurs will be more than the expansion of plywood. I will look up the percentages tonight, but, if I remember right, plywood expands a fraction of what solid wood expands in any direction. Though, I'm not sure if this also applies to the thickness of the plywood.

-DWB

[---k---]

I recently took a class with David Marks of DIY Wood Works fame

That is very cool. I'm jealous.

[orbifold]

I feel your pain, Steve, I'm at work and can't post pics, but I have a 6+cu ft Tempest XJ that was done (but in MDF) remarkably, like the pics John K posted. You and he chose ply to work in. I have the Dayton RS MTWW R & L for mains, all 40+ inches tall and no problem. They are MDF translams with John's bolted front baffles, I guess seen in my avatar.
But with the extra push of the Tempest 15" in a sealed enclosure, I now see cracks. Mine are in the center of the MDF where it is weak. I did pump up the Volume just a while earlier. ops: Well dosen't everybody?
I've thought and thought and the fix I will try will be to drill right through with a 3/8" ship auger bit. This will necessitate cutting my dear in two in the middle horizontally-- the bit is maybe 15" and I have to drill 24" plus. Four holes on a side. The sides are 2" thick. I will start on top and use a Forstner for a clean hole that i can cut a plug for to hide the hole, since I'm doing piano black, heh,heh. :B:B Planning on running 3/8" threaded rod in each hole and using epoxy on the nut or tee-type nut up top, and then the bottom one can be torqued as needed.
Er, wish me luck..:E:E:E:E:E:E

I'm not suggesting this for you, just letting you know you are not the only suffering bastard out there!

BTW the way, those guys are Swweeet! Surely worth saving!

[Brian Walter]

Steve, after reading over this discussion, I think what you are experiencing is shrinkage cracks. My guess is that the laminations were not as dry as the current average moisture condition and that when the laminations dried out the baffle prevented the laminations from shrinking at the baffle developing tension in the laminations (from top to bottom) resulting in splitting. Shrinkage cracking is common in concrete and is similar to the what you see in a dried up mud puddle.

I tried filling shrinkage cracks in a wooden bar top one time and I was less than satisfied with the results. I would recommend trying to put some epoxy in the crack and squeezing it together in hopes that the additional strength of the epoxy might keep the crack from returning. If you could remove the front baffle first that would be ideal as it would relieve the tension minimizing the likelihood of the cracks returning. Applying fiberglass to the inside as suggested by others would also be a good idea.

Brian Walter

[jkrutke]

With regard to the general opinion on the baffle expanding ..... Here is how it is put together. To help with reducing expansion issue the padauk is mounted to mdf which should be very stable.

Rather than expansion of the baffle, has anyone considered if the issue might be contraction of the side layers, such as what happens to wood as the moisture goes out of it? Perhaps the issue was that the wood had a little more moisture content than normal. So the baffle is in compression while the sides are in tension, for opposing forces that rip it open.

Maybe the best way to avoid this issue is to make sure your wood is dry, and maybe allowing the glued sides to set for a good long time before connecting the baffle to it.

[jkrutke]

I think what you are experiencing is shrinkage cracks. My guess is that the laminations were not as dry as the current average moisture condition

Your post beat mine by 2 minutes. Damn that hunt-and-peck typing method. :B

[DancesWithBeers]

After looking at the pictures, I wonder, to preserve the look of the translam, why not cut a thin peice of plywood the width of the crack, on a tablesaw or bandsaw, and then put it in the crack? You could rough trim the curve and then finish bringing it flush with sandpaper...

If the crack is not an equal width (wider in the middle would support the shrinkage theory), then the thin peices might require some shaping beforehand.

[Dennis H]

My understanding from our conversation is that you will get a fair bit of motion across the board but very little over the length.

Right, but in the case of plywood, 'across' the grain is the thickness direction. The length and width directions have the plies arranged at angles so they are pretty stable. However the thickness direction of every ply is cross grain.

[fbov]

Steve,
I like you're approach for filling the cracks with epoxy, of the clear variety, of course. I got some epoxy fingerprints on the finish of my NatP towers when I glued up the front baffle and they sanded off nicely with 400-600 grit wet/dry. Yours will be even easier; at worse, one more run through your last finishing steps and all will look good - until Spring when humidity returns. Perhaps a little dessicant in the box would guard against an expansion problem?

HAve fun,
Frank

PS In my experience, long pot time epoxies do not thicken quickly. You may need to tape over the filled crack to keep it from flowing out of the crack at the ends until it firms up, them pull the tape before it fully cures. I used ferrite-loaded epoxy to line the walls of my enclosures, and even with a 5:1 iron ratio (very thick) I still had flow issues on one panel on a flat work bench. That stuff would not sand, and removal took the wood underneath with it!

[Steve Manning]

I think guys might be on the right track with the contraction idea. On Sunday I had to tighten the threaded rods that run down through my speaker stands since they had come loose. It has been rather dry here in Virginia of late and I have gas heat in the house, so that dry things out as well. We have to run a humidifier in the bedroom at night during the winter. Looks like I should run one in my living room as well to keep my speakers happy.

Steve

[bob barkto]

It's a distinct possibility.
Wood doesn't normally self destruct this far along in the process.

That stack will move in thickness (top-bottom) quite a bit. ~3/16" from one extreme to the other would seem about right knowing how much birch likes to move.
The rigidly affixed solid wood/mdf baffle won't move but a few thousandths. It will create stress on the stack and the result is as seen.

Epoxy is the correct fix, but it will split again in all likleyhood.

The joint failed at a glue line. That means the joint wasn't glued properly.
Normally you'll see failure just to either side of the glue line where the stresses get concentrated.

If that joint was sound would the cabinet have split between the glue lines, somewhere in one of the plies?
I think so. Eventually.

Rather than expansion of the baffle, has anyone considered if the issue might be contraction of the side layers, such as what happens to wood as the moisture goes out of it? Perhaps the issue was that the wood had a little more moisture content than normal. So the baffle is in compression while the sides are in tension, for opposing forces that rip it open.

Maybe the best way to avoid this issue is to make sure your wood is dry, and maybe allowing the glued sides to set for a good long time before connecting the baffle to it.

[bob barkto]

Plywood will expand about the same in thickness as a solid wood panel of the same species.

Plywood expands about equally in width and length.
The amount is ~1/4 that of the cross grain expansion of a solid wood panel of the same species. Thickness, wood species, glue and number of plies affect the amount. Extremes can be found.

Rolex, you are 100% correct, but, from what I remember, even the small amount of expansion in the direction of the grain that occurs will be more than the expansion of plywood. I will look up the percentages tonight, but, if I remember right, plywood expands a fraction of what solid wood expands in any direction. Though, I'm not sure if this also applies to the thickness of the plywood.

-DWB

[Dennis H]

That stack will move in thickness (top-bottom) quite a bit. ~3/16" from one extreme to the other would seem about right knowing how much birch likes to move.
The rigidly affixed solid wood/mdf baffle won't move but a few thousandths. It will create stress on the stack and the result is as seen.

Yup. That's why the baffle needs to float on the plywood stack, so it can move.

Back in the day, I built some quilted maple cabinet doors about as wide as those speakers are tall. They were a solid wood, no frame style. They were in Idaho where it's always dry and the wood had been sitting in Idaho for several months. Still, out of an abundance of caution, I mounted the full-width top board that served as a pull with a dado joint, no glue, and a single screw in the center. Good thing too. After a couple of years, the vertical panels had shrunk 1/8"+. They would have cracked for sure and we're talking big bucks labor + materials to build them -- drive to Washington to buy the wood, bandsaw it into bookmatched pieces, plane it, cut it, glue it, finish it, yikes!

[craigk]

save the cab's. if you can find some place local that does fiberglass work they can help you. in fiberglass layup you often have to make custom parts. the correct way this is done in under a vacuum. you could take the cab's to the shop and apply liberal amt of glus to cracks then have the shop put the cab under a vacuum. this will draw all the cracks closed and seal them at the same time. hope this helps.

[AvFan]

Those are beautiful speakers and others have suggested some good possible solutions. I hope you are able to salvage them.

However, I think the cabinet's movement compared to the baffle and a possible dry joint/glue failure were accentuated by the cabinet structure. The center part of the cabinet is stronger than the top and bottom due to the solid portion behind the baffle and tweeter. It looks (third picture in post #19) like the crack is very close to the inner edge of a woofer opening and stresses concentrate at the corners and edges of openings. I bet the crack started near the baffle and worked its way towards the back because that is where the stress could be relieved. I agree the crack was at least partially caused by fixing the baffle to the ply cabinet but the structure contributed too. Wood movement can't be stopped; it must be accommodated. What caused the movement? The air in most homes is much drier during the winter than during the rest of the year resulting in considerable wood movement. A couple of more questions:

Was it warm and humid when you put the finish on the speakers?

Did you seal the inside of the cabinet with finish or a sanding sealer? (could have caused the crack to start on the inside of the cabinet and work outwards).

All the best at trying to repair two very pretty speakers.

Bill

[DancesWithBeers]

Plywood will expand about the same in thickness as a solid wood panel of the same species.

Plywood expands about equally in width and length.
The amount is ~1/4 that of the cross grain expansion of a solid wood panel of the same species. Thickness, wood species, glue and number of plies affect the amount. Extremes can be found.

Yeah, you're right about the thickness changes.

I guess the solution when doing these types of speakers with plywood would be 1) figure out the extremes of dryness/moisture in your area and account for movement within those points, 2) seal from all sides to minimize the expansion and contraction, and 3) be very careful with the type and attachment method of the baffle.

Scary.

[JonP]

Wow... that does suck... Hope the crack fill goes well.

Wonder if you have more of the glue you used. Might be worth making some test joints, and comparing with a fresh bottle, to try and see if the glue did suffer from freezing and was weak.

Otherwise, the thought that there could be more expansion/contraction than joints could stand is kind of worrying. Then, allowing a long drying period after glue up (lot of water applied in the water based wood glue) before final finishing and assembly might be an idea. Maybe also trying to seal the wood (inside as well as out) after construction would help with long term moisture loss/gain.

After you do your crack filling, might be worth pulling all the internal foam, and giving the interior several layers of Zinnser Seal Coat (shellac) or similar, to add a moisture barrier.

[---k---]

I was going to mention something similar to JonP. In my searching for people using Minwax Wood Hardener, I noted that people use it to "stabilize wood". They soak the wood in it to fully harden the wood, after which it is no longer affected by moisture changes. I would assume that if you could get it to a dry state, then just sealing the surface with shellac or epoxy would do the trick.

[Hank]

Difficult to guess at the cause, but I'd go with:
#1 starved glue joint
#2 there may have been MORE moisture in the laminations than the ambient environment your finished speakers reside in
The only problem I've ever had gluing wood with all kinds of glue was not enough glue or too high clamping pressure.

[gmed]

Hi steve. One question. Is the enclosure opening up on the baffle surface also? If not then it cant be due to the expansion of the baffle. Also the hardwood is glued to MDF which I highly doubt that would expand.

If the splitting is on the wall of the enclosure, I really think it could be due to shrinking of the plies, rather than the expansion of the baffle. Especially since you cant clamp it back together. This could be due to moisture in the wood when you started and as the wood dried, it shrunk.

Again, I still think it may also be due to insufficient glueing/quality issue, along with lack of nails.

One thing you can do is to fill the gap with wood putty and refinish. I used it on some of my small gaps and after sanding an finishing, you cant tell. Of course, I would fill with epoxy as much as you can, but the surface I would fix with the wood putty. you can find them with the same shade as the Birch.

[Steve Manning]

Update ....

Just a quick update while I eat lunch here at work. The epoxy work is going well, I’ll post a picture later. It took a couple of rounds to get the surface nice and level with the cabinet. I’m going to give it a few days to set really well before I sand it, though I’m thinking of waiting a little while before I do a re-finish just to see if I’m done filling cracks in, I would expect not.
At this point I’m hoping that the cracking will not be too severe, and I can stay ahead of it and keep the cabinets looking good and save the day. Either way it’s given me a few extra things to add to the “what I would do different next time” list …. Making sure the glue is “good” both in the type of glue, as well as not being possibly compromised. Adding nails and or glue dowels through the BB layers for additional layer/joint support. Seal the inside as well as the outside, nope I did not do that, and here in Virginia, with the Dismal swamp just down the road, it gets rather humid during the summer time. One could easily see 50% humidity swings from winter to summer. Since I had heard of how stable Baltic birch was, I did not think of sealing the inside of the box. The last thing would be to attach the baffle and back plate with bolts/screws rather than gluing them in place and constraining the birch layers. Lessons learned I guess.

Steve

[bobhowell]

Your stack will move again, with the season, just like hard wood floors squeak in the winter. No amount of glue will stop it. You might have had wood that had 15-20% moister content. Here in North Georgia I have monitored wood I have drying for 15 years or more. It gets down to 12-14% outside, and stabilizes. When I move it inside it drops to 9-10%.

Yellow birch shrinks 16-17 % volumetric, when going from about 30% to oven dry. The shrinkage is linear, so, 5% drying is 2.5-3% of the original length lost. Over a 24" tall speaker, that is a half inch, to, three quarters. So just a 3% drop would open up a bunch of cracks up to about 1/4 to 3/8 inch.

I left a half sheet of MDF outside this fall for 3 months and it expanded 1/16 inch in thickness, from3/4 to 13/16. My dado's were too narrow and had to be recut.

I think you are planning to let the baffles float, now. so that will keep the cracks from reopening. Floating originally might have saved you.

Zaph discusses a method of attaching baffles he uses on his site. His method would probably allow the necessary movement , also. Drill oversize holes in the baffle and use washers to catch the fasteners. As the stack moves, up or down, the washer will slide along the baffle. You must allow for 3/8" movement, so use a big washer. You must also find a way to seal around the hole.

[Steve Manning]

Up and running ... for now anyway

Well I'm back in business at least for the time being. Here's a few pictures of what I did and how it turned out. The first one is after the first coat of epoxy, I had to lay down two. The second is after a little hand scraping, and the last is getting ready to go back on the stand. My wife could not tell where I had done the repair when she got home from work, so I guess that's good. Now it's wait and see where things go from here.

Steve

[David_D]

Great Job Steve!
Looks great! Smiles all around.