Official Luthiers Forum!

Before glueing the top plates one must true the sides of the top plates. I have tried a technique of sanding with a long level. Also tried one with sandpaper stuck on a glass board. Also tried to plane them. It seems everytime i get so dang close and there is a hair of light shining through to my saddened eyes, as if to spite my long efforts in getting a perfect join. I have tried it on 3 sets already but i dont seem to be progressing. I have watched videos of guys doing it in about 30 seconds and bam its done. What the heck am i doing wrong here?

_________________
久保
美智え

Michiyuki, I struggle with the same problem every time. It's possible that you are closer to perfection than you realize. It can be very difficult to hold the joining surfaces snugly together while checking for a light leak. Try butting them together on a flat surface and taping across the joint every four or five inches snugly before holding them up to the light. Then, if you still see a tell-tale leak, you will have a better notion of where and how much correction is needed.

I confess that I still haven't managed to plane a perfect joint. I get close and then I finish with sandpaper on the edge of a 36" level. I do this with the joint edge slightly overhanging a piece of thin plywood laid on top of my table saw. And I sand the joint edges one at a time. If you sand your joint, be sure to sweep all the dust off when you're finished and then wipe it with some alcohol or naphtha on a clean cloth or paper towel before glueing.

Patrick

Lots of folks struggle with this but you will master it with practice and time.

Sanded joints are inferior in my view for a number of reasons. First a sanded joint does more damage to the wood cells, leaves debris, and if not properly cleaned up can leave a darker colored line in the center seam. A sanded joint also has less joint energy for gluing purposes than a freshly planed joint. Remember we are joining two very thin edges so this joint, for lack of much gluing surface, must be very well prepared.

This is why I favor a planed joint. Something that really helped me learn to shoot this joint with my Record #5 was Link Van Cleave who talked me though the idea of starting the pass with the hand pressure on the front of the plane, shifting to the middle in the middle of the stroke, and having pressure on the back of the plane by the end of the stroke. This was the piece that I was missing in my own lack or technique.

It also all starts too with a decent plane, well tuned, and an uber sharp blade.

I used to candle the joint but these days I simply inspect it closely. You can see any gap that exists, it will appear slightly darker. I don't have great vision either but can still see when I have a gap even a very small one.

You might want to simply practice once you have a decent plane and it's set-up well. Dedicate a junky top top learning to shoot the joint. My bet is that you will make real progress with some practice.

Patrick I will tape them up and see what it looks like when I get back home. What grit did you use for the paper? I used a pretty fine grit since I didn't want to take to much too fast. This had me going for a long while though.


Hesh - I agree the planed surface is better. Surfaces will glue better for sure compared to sanded. Without a question. I switched to sand paper strictly because I tried my plane with the same results. I sharpened the plane blade before I went in. The technique you described is kind of what I tried. Except I was switching weight more to the back. I will try to switch to the middle this time.

Appreciate the advice fellas. Let's hope I don't end up with 5" wide too plates by the time I get it down hehe

_________________
久保
美智え

I don't candle mine. I joint them with a plane on a shooting board, then place them down on the clamping table flat and see how the joint looks closed on the table. If it disappears I know I have good joint.
Any gaps are clearly visible and it's less frustrating than trying to see if any light comes through.
the plates have to be sanded to an initial thickness and cleaned up though.
I join my tops at about .130-.140 thickness, install the rosette, then thickness the rest of the way on the sander and clean up the rosette.
I thought about making a candling platform that I could hang on the wall and set it down flat when I needed it to lay the plates down on top of to reveal light gaps, but I decided that would be overkill.

_________________
Old growth, shmold growth!

When doing the weight shifts look and think about what is happening with the plane. When starting out the back of the plane behind the blade is out in space. So weight toward the front. But if you watch you find you need to keep the part of the plane in front of the blade flat on the wood. It is really easy to rock it to far forward so that the blade is riding a little high. Once the back of the plane is on the wood, you have both the leading and trailing edge of the plane on the wood, Keep them both there, thus weight in the middle. As you are going to exit the plate you need to remove the weight from the front of the plane as it is going off into space. But there again it is really easy during that weight shift to go to far back and lift the blade off of the plate a bit. Watch behind the blade to be sure it stays flat on the plates. On the final passes there is no need to go fast, just take care that the plane is riding properly on the plates.

_________________
http://www.Harvestmoonguitars.com

A few comments:

Candling -- even joints that look absolutely perfect may yet reveal slivers of light passing through when view under a light. This is an important joint and you really want it perfect. Additionally, tape the set together at the top, middle, and bottom when candling. Sometimes a shift in the orientation of the plates will cause light to show even when the joint is actually perfect.

Technique -- you want a well setup plane that has some length to it. You also want to be taking very thin shavings with the throat as narrow as possible to avoid tearout. Something like a block plane would really require excellent technique. Start the movement with pressure on the front of the plane, then pressure in the middle, and then pressure on the back of the plane as you finish the cut.

Troubleshooting -- if you are really having trouble, then here are a few things you can do. Get the joint mostly perfect and then begin to take a few strokes starting and ending a few inches from the ends of the plates. This will carve out the center. Then come back with full passes with the plane. This sometimes makes it easier to get the joint closed completely. If you still have trouble, apply this same strategy but just sand the plates instead of planning. I have had some hardwood with weird grain that wanted to tearout badly at the ends. I put some 3M gold stikit 320 grit on my jointer and used the fence to sand the set level. As soon as you see a sanding scratch across the entire length then you know the joint is good to go. In practice, a 320 grit surface will be just fine.

Lastly, it does get easier!!! -- I remember going through this same process many years ago when I started building -- but this will eventually click and then you can do it with ease. You might check out YouTube videos and see if that might help.

IME, it has little to do with practice, and everything to do with your plane. It needs to be reasonably long (mine is 13"), absolutely flat, sharpened to perfection, chipbreaker set reeeeally close to the edge (if you need to go against the grain at any point... otherwise don't need a chipbreaker), blade set for a very fine shaving, and preferably with a very small mouth opening in front of it. It took me 2 years before I could get a plane flat enough and blade sharp enough to cut perfect joints. Before that it was basically impossible to get a perfect joint, and since then it's been pretty easy. I've never had luck with sanding them.

I've also switched to jointing the plate halves one at a time, rather than trying to plane them together like most tutorials say. Mostly it's just a lot faster than trying to perfectly re-align them on the shooting board after every time you stop to check the joint.

Softwoods are a bit more forgiving because they have a tiny bit of give to the surface, if you put pressure on the joint when gluing. Rosewoods and other super hard types have basically zero give, and the joint must be flawless or it will fail (starting from the weak point and spreading into the good part). On my first ones, I just routed a channel half the thickness of the plate and glued a solid wood backstrip to compensate for my imperfect jointing (along with the usual cross grain reinforcement inside).

Decent Plane, absolutely sharp, finely set. That's it. It's nearly always due to a blade that is not as sharp as it should be or set to take too great a shaving. There is a bit of technique required, namely transferring weight to the front or the rear of the Plane where necessary. Practice but practice blade sharpening more than anything.

I am no expert by any means and have only joined a few tops but have done a fair amount of joining wood for other projects. I only use my #7 Baily plane which is made just for that purpose. I bought it a number of years ago for a song in a garage sale but it needed some TLC to bring it to perfection (It is almost 100 years old). But a new SW blade, flattening out the base and truing up the frog made this one of my best planes I have.

As far as shooting an edge, I don't really worry about putting pressure on the front or back while shooting. I hold the wood down on the shooting board with one hand and push the plane with the other. I try to hold the plane in the middle (not by the handle) and give it a nice even push all the way through the pass. I do adjust the blade so it is just barely kissing the wood and taking the thinnest of shavings. I have never had a problem with my #7 as long as it is tuned and sharp.

I have tried sandpaper but it just doesn't give you a clean edge. I have a tendency to look at everything through my 10X photographers loup and if you try it and actually look at the surface of a sanded vs plained surface there is just no comparison.

Bob

While the family went to sleep I snuck in the storage room and worked on it a little. I got one top set about 99% so I am going to call that one done. Maple back came in second and another spruce top will need some more work tomorrow. It's amazing how much it takes to make something completely flush like that. I started adjusting the plane more and it helped a little. So I will Resharpen it inspect and set it right again to see if that helps more when I finish the other two. Thanks again. The evils of hair width gaps can frustrate the most even tempered.

_________________
久保
美智え

I'll just chime back in one more time to say that I believe Hesh is correct when he says a planed joint is superior. I just haven't yet been able to manage one. This is why I said you must be ultra fastidious about sweeping and then washing away your dust debris. But if you can get the hang of that planing technique, it is definitely the best way to go. You asked about the grit I use. I have used 220 after getting as close as I can with the plane.

+ for DennisK. I had problems until I got the right plane. Now it if fairly easy.

_________________
_____________

EddieLee

Scary sharp well set up plane, set very fine, one of my favorite jobs in the shop. I use a #4-1/2 bailey with a hock blade (got $15.00 in the plane $28 in the blade) I'm betting most of your problem is the blade is not sharp enough.

I'm going with not sharp enough blade. You don't NEED a perfect plane that takes two years to learn to perfect, although I'm sure that helps.

The joint is in the middle of the dark stripe down the middle of this one.

Incidentally I test all my joints after I cut the soundhole out, break the little circle to see where it fails.

_________________
Old growth, shmold growth!

Agreed - I really struggled getting the perfect joint until I scored a Veritas #4-1/2 and learned how to get it insanely sharp.
What used to take hours now only takes a few minutes...

Hang on in there. And technique matters as well as the tools, so practice.
It's taken me 6 guitars and just this most recent guitar, got a perfect joint for the top, candled over a perspex sheet supported on a workmate with a strip lamp 2" behind it. (before, I was happy to have a bit of a gap under the fingerboard extension area)
Scary sharp blade, tuned plane, set fine at the mouth, of a decent length and set it for whisper thin shavings. Don't be shy to touch up the blade if you've had 30/40 swipes at it with no luck, even with softwood.
And stagger your joint slightly or you won't be able to find the seam after gluing. dang nuisance.

_________________
The name catgut is confusing. There are two explanations for the mix up.

Catgut is an abbreviation of the word cattle gut. Gut strings are made from sheep or goat intestines, in the past even from horse, mule or donkey intestines.

Otherwise it could be from the word kitgut or kitstring. Kit meant fiddle, not kitten.

Two thoughts come to mind:

1) It is almost impossible to shoot a closed joint with a plane with a concave sole. The sad fact is, most planes suffer from that deficiency. Even ones I've flattened (20, 30 and 90 year old planes) have reverted over the years and have to be re-trued every few years. The new generation planes (LN, LV) don't seem to suffer from this malady. Or at least the ones I have don't. With a properly flat plane, I can shoot a perfect joint in about a dozen strokes or less than 2 minutes. With a blade of the same sharpness with a concave sole to the plane, best not to even start, because it's almost impossible.

2) You need to see a hair of light anyway. If you're gluing up with a water based glue, like most of us do, the wood at the joint absorbs the water and expands, causing that dead straight joint you just shot to go convex, i.e. gaps at the ends. Jacking the ends tight can be a real pain, especially on hard woods, so if you leave a hair of light in the middle it's much easier to get a totally tight glued joint over the full length.

OK, one more thought:

Changes in grain direction on the shot edge cause changes in the depth of cut that a plane will take. Small, but enough to spoil your day. So in these cases it can be quicker to do a series of spot faces with a short, high angle block plane to achieve that graduated hair of light when candled.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

As an aside, I've noticed, with my 3 anyway, as supplied, LN planes have been pretty flat on the surface plate round the edges, but concave side to side by about a thou., i.e slightly dished in the middle.
Anyone else noticed that?
I just wonder if that's deliberate?

_________________
The name catgut is confusing. There are two explanations for the mix up.

Catgut is an abbreviation of the word cattle gut. Gut strings are made from sheep or goat intestines, in the past even from horse, mule or donkey intestines.

Otherwise it could be from the word kitgut or kitstring. Kit meant fiddle, not kitten.

Hi Michiyuki,
A perfectly planed edge from top to bottom is an admirable thing to strive for, but is not absolutely necessary. On the sound board the only part that _has_ to be perfect is the area from the soundhole to the bottom of the guitar (the lower bout). The soundhole will be cut away, so obviously it doesn't have to be perfect. The upper bout seam is generally hidden and reinforced by the fingerboard, so "close" is good enough. Get the lower half perfect and the rest close and avoid some frustration (and an ever shrinking soundboard).
If you inlay a back strip and use a back graft then the back seam doesn't have to be perfect either - just close.
It's good to make perfect joints, and is a skill you will eventually acquire if you work at it, but don't let less than perfection hold up your guitar building, and always keep in mind what really matters.

So what happens after the glue dries and the wood reverts back? Seems to me that it's better jack the ends tight and have a straight joint after the glue dries that's not trying to pull apart.

I never seem to have that problem though.

_________________
Old growth, shmold growth!

What Trevor describes is what I was taught to call a spring joint. It guarantee's that your ends are locked tight where they are most likely to fail. Though as Clay mentioned at the top end you will have cross struts and a fretboard to hold the seam together and at the bottom you have the tail block. I've always accepted a bit of light in the middle on a perfect arch. And I mean the tiniest of all protrusions of light, barely there, you can push it together with your hands to close it off. I've never had a top fail though I have had a back fail but that was under poor conditions.

I got the spruce tops pretty dang good now. I still see a sliver but i got it closer, if i push them together on those areas they close up since the wood is so soft. The maple back are much harder, therefore there is no give at all on the joint. I am having a much tougher time with them. I worked on them for about 2 hours and got close but it will need further trial and error as it does not meet my expectations. Thanks to everyone for all the info that is really invaluable here. I am interested in what will be the outcome!

_________________
久保
美智え

I did my first few with a block plane, but it's not the right tool for the job. It is good practice for setting up, sharpening and using a plane though... The blade must be scary sharp, the plane bottom ground flat. Even then, it takes time and skill to get a perfect joint. Like I said, good practice...

Nowadays I use my grandfather's pre war Bailey #8, which is much faster and easier.

That's what back-strip inlays are for
I think after 2 hours you should look at your technique or equipment. Is the gap is always in the same area?? Maybe your surface isn't flat or you're pressing harder at some point along your sanding stroke.