Having some trouble with my dovetail saw build

[Chris Griggs]

So I got around to starting my dovetail saw build last weekend and up until today everything was going fantastically. The lessons I learned from my first saw build were proving to be invaluable, and I was able to get a nice tight mortise for the back and a nice straight kerf for the saw plate right of the bat. With the assembled back/plate inserted into the handle everything looked tight and straight, so I went about roughing out the rest of the handle, doing the intitial cleanup and then drilling the holes in the handle and plate.

Again, everything seemed perfect until I inserted the nuts. Everything appeared aligned and went in easily but when I tightened down the split nuts my plate would bend at the heel.

I checked to make sure the wholes in the plate were aligning with the holes in the handle - seemed good.

I checked to make sure there were no burrs around the holes I drilled in the plate, filed around around the holes again to make extra sure no burrs were throwing things out of alignment - no burrs, seems good - still got warp when I tightened down the nuts.

So I assumed maybe the mortise for the back needed just a bit more tweaking - took a little away from here and there - still no luck.

Took more way, mortise is no longer nice and tight.... ARRRRRRG.... slight improvement in warping, but plate does still bend when I tighten down the nuts.

The mortise is loose enough now that I really don't think it is causing the misalignment/bend when I tighten the nuts. It's also loose enough that I'm going to need go back and start from scratch. Actually, already went back and started from scratch, but messed up, so technically my next attempt is no. 4 - much harder now that the plate is already drilled.... ARRRRRGGGG again! I kinda wish I had just left well enough alone and accepted some bend at the heal - the handle was great in all other respects and its just in that last inch of the heal - probably wouldn't have even impacted performance. Of course, me being the obsessive perfectionist that I am just could leave well enough alone....

So anyway, I still can't figure out what went wrong with the first handle - it seemed like everything was going fine, and I cannot for the life of me figure out what is causing the plate to bend when I tighten down the nuts. It is dead straight when the nuts are in place and not tightened down.

I'd like to figure out what happened before going for handle no. 4. Any idea's what going wrong?

[Steve Thomas]

Could one of the screws be engaging with the plate as it passes through?
Or when they are inserted do they just drop into place?
Is there junk in the kerf, of the handle?
I bet its painfully obvious when we figure it out.

Steve

[David Weaver]

There is a wander somewhere in the saw cut between the cheeks in the handle. I don't think that's uncommon, and not a big deal as long as the rest of the plate is straight and the bend is only right under the cheeks. That part of the plate won't touch wood, anyway.

I have a slight amount of that in my first saw, but it still cuts great.

My nurse dovetail saw with the original handle is like that also.

[Chris Griggs]

Yeah it really wouldn't have effected anything. I wish I'd left the mortise alone - it was a really nice fit until I mucked around with it.

Anyway, so you're saying that whatever wander is their is too slight to cause any warp with the plate just seated, but that when I tighten down down the nuts and the cheeks close one the plate it causes the warp?

So how do I prevent that? I sawed the kerf dead on the gauge line with my other LN dovetail saw. A far as I can see there is no wander on the visible part of the kerf. How do I prevent it on the inside if the cut? I mean I sawed really really carefully. Is there anything else to keep in mind other then the standard accurate sawing?

[David Weaver]

I can't think of any way to eliminate it completely. I used a ryoba to do mine, the one that wandered, and then I did it again for the other two, and they were OK. It probably took 5-10 minutes for me to make that one cut the second and third saws because I was concentrating on pulling the saw with no pressure and as straight as possible. I think a skilled sawyer could make a cut like that with speed and precision, but I don't have that much speed or precision and I like to make that cut after the mortise has been cut.

[Chris Griggs]

Yeah, I go really slow for this too - it definitely needs to be more accurate than most wood-wood joinery. I wonder if to some extent its just the nature of the thin (.018) plate. Once I got the mortise fitted and properly aligned on my sash (.025) I had no issues with warp once the nuts were tightened. I'm finding I like to cut the mortise before I cut the kerf too. I did the opposite on my first saw (kerf than mortise) and for whatever reason, like you, I'm having an easier time getting everything aligned when I cut the mortise and then cut the kerf.

I'll give handle no. 4 a crack. If that doesn't work maybe I'l go back to handle number 1. The mistakes I making on these latter attempts are a result of trying to make a handle for a plate that's already been drilled.

The mortise on no. 1 is a bit gappy now but not terrible. I can always glue in a really thin shim, do some refitting and should be able to get a pretty invisible repair for the gaps - I did that on my first saw and it came out quite well. Though for this one I really wanted everything to come out pretty seamless - hopefully the 4th times a charm!

[Chris Griggs]

Could one of the screws be engaging with the plate as it passes through?
Or when they are inserted do they just drop into place?
Is there junk in the kerf, of the handle?
I bet its painfully obvious when we figure it out.

Steve

Hi Steve, I just realized I didn't respond to your response. Nope, its not the alignment of the hole, everything goes in nicely. One of the holes in my plate was initially a little off, but I elongated it with a chainsaw file until everything fit together, so I know thats not the issue.

I checked for gunk in the kerf as well, ran the saw that I had cut it with back through to make sure it was clean.

I think Dave's right that its probably just a very very slight wander in the cut that is undetectable until the bolts are tightened. I'm guessing that with a thicker plate it wouldn't even have shown up at all - but I guess the thin plates quite a bit more sensitive. The kerf sure seems accurately sawn, but the saw plate seems to disagree with my assessment of the kerf. Good thing I made a bunch of copies of the handle pattern I'm using....

[Ron Bontz]

Hi Chris. +1 on the possible deflection inside the cheeks of the handle. Also take a good close look at the alignment of the spine slot and plate slot starting at the front and working your way back. If the spine and plate slots don't match, centered, up you will get that bend at the heel of the plate when you tighten the bolts. It will go away or become less when you loosen the bolts trying to find the problem. (Don't ask me how I know that. :() Lastly, if the slot in the spine itself is not straight vertically as well as horizontally or has a slight twist in it, it will bind when the cheeks are tightened. ( don't ask how I know that either.) Live and learn. :) Best of luck.

[Jessica Pierce-LaRose]

Does releasing one nut or the other on it's own relieve the bend? I wonder if the countersink wasn't perfectly perpendicular or something if it would pull things out of whack?

[david herbert]

Perhaps you could try shimming the bend out using masking tape,trying it in various places with differing thicknesses - layers of tape. If it works,replace the tape with steel
feeler gauge stock.If that does not work you have lost nothing but time an possibly learned something. just a guess,though.

[Isaac Smith]

Can you remove the spine and just fasten the plate in place? That could help you narrow the problem down.

[Jim Koepke]

I sawed the kerf dead on the gauge line with my other LN dovetail saw.

Maybe luck was with me for my first back saw build. My understanding was the best way to cut the kerf for a saw was to use the saw plate being mounted with unset teeth. This is to make a tight fit on the saw plate.

My post doesn't make it clear how the holes were laid out since the old saw plate already had holes. There were some lines drawn on the handle to match the saw positioning and then the plate was used like a template. There was a mess up with the forstner bit, but it still worked out. The holes were drilled just a touch undersize and clearance was made with a round file.

http://www.sawmillcreek.org/showthread.php?180712-Back-Saw-Build

jtk

[Chris Griggs]

Thanks for all the input folks.

Josh, no, the bend happens when either nut is tightened - I was think too, that it could be that the drill isn't perfect, but the nuts are fitting in smoothly and it doesn't matter which one tightened. Pretty sure it has something to do with the kerf tightening down in the plate.

Dave H, the shimming idea is an interesting one - if nothing else it could help me with the final fitting of the mortise on the next handle attempt. That's something I'll mess with.

Ron, I initially though it had to do with the alignment of the spine mortise with the kerf, but for the life of me no matter where I take material away I can't get rid of the bend (which is why I'm now thinking it might just be the kerf). Trying to adjust the alignment of the kerf with the spine is actually how got to the point where I now have a gappy mortise - it's so hard to tell where to remove material to get in the mortise to get is inline with the kerf. Any suggestions, for how to go about evaluating the alignment and determining where to remove material in the mortise - this fitting/aligning of the mortise has been BY FAR the hardest part of both my saw builds.

Jim, thanks for the templating advice. I think I am doing something similar has far as using the plate/back assembly as a template to mark out the holes and the mortise. I would not have the to drill the hole in the handle a touch undersized to start - that's a good idea that I'll try.

[Klaus Kretschmar]

Hi Chris,

sorry to hear about those issues. I do know them however! Quite often it's indeed a misalignment between blade slot and spine mortise that causes a bending of the blade heel. But in your case it seems to have another reason since you widened the spine mortise in a way that it has space in every direction. Given that the blade slot was sawed that carefully with a good saw, I can't believe that the blade slot causes the problem. My guess is that the slot in the spine might be slightly misaligned. It's absolutely necessary that the slot is 100% parallel to the sides of the spine. If there is a little misalignment (what hardly can be seen) the blade WILL bend. To test it, you can clamp the spine with inserted blade in the front vise in the way that the heel of the spine and the blade can be seen from above. Now the blade has to be exactly parallel to the jaws of the vise.

Cheers
Klaus

[David Weaver]

As Klaus has mentioned that, I should revise what I said about how I do the slots. (as he mentioned, it's important that they are parallel).

I do cut a mortise first, but I cut it a bit undersized. then saw the slot, and I float or pare the sides of the mortise to size to fit the plate with spine attached last.

My vintage saws have the mortise touching the spine, and the reason I do things in that order is so I can get a flush fit of spine and mortise on all sides to mimic the vintage saws. I doubt I could tell the difference in use if the cheeks have good contact to the plate, though. I never saw anything that roughly with a light plated dovetail saws, but as george often tells us pikers to do - copy something until you can make perfect copies before freelancing.

[Chris Griggs]

So, seeing as that handle was more or less going to be junked at this point I went back and started messing with it again, figuring I might as well learn something from it.

I trimmed the sides of the mortise more, probably to the point to where you can fit a dime in there - blade still warped. So I took another very close look at the kerf. Setting a marking gauge to the edge of one side of the kerf I moved it around the kerfs edges. While it was indeed in line on the visible part of the kerf where I had originally struck and sawed to the gauge line I all of the sudden noticed the kerf where the plate enters the handle inside the back of the mortis. It was noticeably off center. How that happened I cant say, but it is clear that somewhere in the end of the cut my saw wandered. Not enough to cause bend when fitting the mortise to the kerf (using the process Dave describe above), just enough to bend the blade when I tightened the nuts. To test this I put my crappy Ryoba saw in the original kerf and twisted it sideways to widen it where it looked like it was pushing the blade - tightened the nuts down and long at last the bend was gone. Of course, at this point the mortise in my handle looks like it was fitted with a dull screwdriver and the kerf looks like it was cut with butter knife so heres hoping the next turns out better.

Thanks a bunch for the help everyone. I don't mind making mistakes and having to redo things, but not knowing what is wrong and being able to learn from the mistake, and thus avoid it on latter attempts, makes me crazy.


And Klaus, I've been meaning to mention. I took your advice and filled the pores on my sash saw handle by sanding in shellac. It feels and looks a fair bit nicer - I'm not sure how well the difference will show up in pics, but whenever I get around to snapping some shots I'll repost some more pics to that previous thread - that was really great advice.

Thanks again all!

[Mike Allen1010]

Hey Chris,

I absolutely know the frustration you're talking about with the "mysterious" bend in the plate when you tighten saw nuts!

I'm glad you found the cause -- in my experience it's always misalignment of the kerf in the tote for the saw plate. For me, this is particularly a problem with thin plates, .015"-.020". I've built a couple dovetail saws with .015' thick plates -- sounded good in Theory -- thinner plate = thinner kerf = easier/more accurate sawing. In actuality it was brutally difficult getting everything straight!

For what it's worth, here's what works for me (please don't ask how I know this!):

1) Make sure the sides of your tote blank are absolutely parallel (Calipers). if they're not, when you reference you're marking gauge off one side of the tote to lay out the kerf for the plate, the layout line won't be directly aligned down the middle of the tote.

2) Which leads directly to -- reference your marking gauge off both sides of the tote when laying out the line for the plate kerf. This way you'll be sure to find the absolute middle of the tote and if the sides of the tote aren't absolutely parallel you'll be able to tell and split the difference when you saw the kerf.

3) This problem is the reason that I saw the kerf for the plate first, and then insert the saw plate/back into the tote and mark the location of the saw back Mortice directly off the saw back in itself. This way, if the kerf for the plate is even minutely misaligned (which is all it takes), you can line up the mortice for the saw back in the same orientation as the plate. It may mean that the saw\plate is 1/32" off from being perfectly aligned with the tote, but I would rather have that than a bend in the plate.

4) To saw the kerf in the tote, I clamp the tote to my bench top and then lay the saw I'm going to use to cut the kerf on top of a spacer/quide block that is precisely thicknessed to align the saw teeth in the middle of the plate. Using the guide block helps me ensure that I saw the kerf for the plate exactly parallel to the side of the tote.

If all else fails -- playing cards make great shims!

Good luck with saw. I'm looking forward to seeing the pictures!

All the best, Mike

[Chris Griggs]

Hi Mike - I knew I'd get some good input from you.

Yes, I have to believe the thin plate is what is making it that much harder. There is no way I was more accurate with my first saw, and I think the thicker plate just made it so it wasn't a problem.

Your order of operations makes a lot of sense - some imperfection seems more likely to occur than not and I like that you method corrects for that. Like I said I've been preferring to cut the mortise first, but this will definitely make me reconsider how I'm doing things. Thanks for the very detailed response - your advice is always very helpful.

I'll shoot you an email when I get it licked and have some pics to post.

Thanks again to Dave, Klaus, Ron, et al. as well!


P.S. Mike, how's that chest coming? You get it started yet? I'm looking forward to seeing it.

[Mike Allen1010]

P.S. Mike, how's that chest coming? You get it started yet? I'm looking forward to seeing it.

Chris, thanks for sending me pictures/dimensions for your chest - without your help I'd still be staring at a blank sheet of paper!

Plans are now done. For me the hardest part -- I couldn't suck more at design! Masters like George with an "eye" for form, shape and proportion are amazing to me! It took me two hours to draw templates for the ogee bracket feet and they still look like they were done by a kindergartner under the influence of hallucinogenic drugs! (Okay, that is probably not a PC metaphor - my bad!).

I'm using curly maple with a really cool shimmery figure. It looked good at the lumber yard, but I just spent three hours surface planing the glued up carcass sides. I can report the following:

1) I will not be going to the gym until this project is finished as planing this Maple is WORK! Surface planing just part of the carcass produced a trashcan full shavings and about a gallon of sweat! I'm certain I will need numerous medicinal beers to appropriately rehydrate!

2) With regard to my ability to sharpen/use hand planes that can surface this curly maple without tear out; there is sharp and then there is "Curly Maple Sharp". If I were a baseball manager I would be desperately looking to the bullpen for some kind of high angle/infill smoother relief. Guys like Ron Breese and David Weaver make those infill planes sound really appealing about now!

3) In the back of my mind all I can think is, if I spend all this time with this curly maple -- I better not screw up the finish! I put my odds of ultimately achieving a successful outcome at 50/50.

Chris, I appreciate the moral support! I look forward to your pictures of your dovetail saw. Don't hold your breath on me producing any thing photo worthy.

All the best,

Mike

[Chris Griggs]

Mike, glad to hear you're progressing - I'm especially looking forward to those feet. Getting those proportioned right was a bear for me. All I can say about the curly maple is next time you buy look for the word "soft" before the word maple ;) That should make things a fair bit easier.

For the finish, well I don't know squat about finishing, but what I do know is that BLO looks mighty nice on just about everything I've put it on, and is darn near impossible to screw up. Good luck with the project - really looking forward to seeing it.

[Chris Griggs]

Took vacation day today and, of course, used some of the time to take another crack at the saw handle. I had prepared a blank earlier this week, so I spent 3-4 hours this morning doing the letting in. Really really took my time. I used the procedure that Mike mentioned above. First, cutting the kerf, putting the plate in the handle with the back sitting on top and used it's location to mark out the mortise. It worked very well. There was still a lot of delicate fitting to do, but I finally go a pretty tight mortise, nicely aligned saw nuts and a plate that stays straight with the nuts tightened down. Proceeded to rough out the blank, and now I can get to the fun part, that is shaping. The shaping will take me several hours, but all in all I find it to be a lot easier than the letting in of the handle, and again, it's a lot more fun.

Thanks for all the helps guys. It's amazing how much more precision the thinner plates require. This was really challenging for me. It feels really good to have a blank I'm happy with, and I learned a lot in the process. Will post a new thread when I complete the saw...probably next weekend.
245189 245188

[Jessica Pierce-LaRose]

Looks good so far!

A few times when I've needed to make some "need to get it spot on, and only one shot" saw cuts, besides marking with a knife, I've taken a scraper as thick as my kerf, and turned a hook on the corner, and used that to "kerf in" the line in all directions to help start the saw. The little things that stewmac sells (http://www.stewmac.com/shop/Fretting_supplies/Saws_and_slots/Fret_Slot_Cleaning_Tool.html) to clean up fret slots work well too, and area little easier to hold than a scraper working sideways. Same concept as that Glen Drake kerf-starter, I suppose. I wonder if this would help in this application?

[Chris Griggs]

Thanks Joshua. I have one of LVs new freehand inlay groove cutting stringing tools that I sometimes use for that purpose. I thought of using it here but decided that sticking with just a gauge line would be more precise in this case.