Stuck -- half blind gaps

[Aram]

Maybe someone can key on on my mistake. A few weeks ago, I was building drawers are getting good results. Lately I have been consistently getting a gap on my half blinds, between the end of the tail board and the... I wish there was a standard word for it. Ian Kirby calls it the "end wall" so I'll go with that. Here's a practice set.





Here's my method. Sorry for the length, but maybe there's a clue here. I make the tail board first. I cut the pin board, shoot the ends, and scribe the baseline and end walls with a marking gauge. (As it happens, I am using one gauge for both; in any case, I always use the same gauge to mark the tail boards and the pin-board end wall). Lately, I have been going very lightly on the end grain marking, to make sure I don't get a deep, thick groove that could throw off my chisel work. By the way, I always check, and the ends of all pieces are square in both directions. I use Derek's blue tape substitute for the 140 trick, 3 layers of tape for easy registration -- and the tail board lines up perfectly with the end-wall scribe line. I've been marking and sawing well lately, so that doesn't seem to be a problem.

The tail piece gets clamped flat on the workbench with straight board across it as a guide for chiseling the baseline. I alternate chopping from the inside face and end grain, until I am <1/8" from the end wall. Next I clamp it firmly (2 clamps) to a backer board, clamp the pair into my face vise, and add a 3rd clamp across the width to keep wood from splitting at the edges. I chisel, alternating down and in. I'll go in a few times with a fishtail to clean up corners and remove waste from the pins. I do this until there is almost no wood between where I am and the end-wall scribe line. Only then do I chisel to the scribe line. I'm not removing much wood at this point. I check and recheck that the end wall is square, not tapered. (I've had that problem too, but that pushes the whole tail board out). My chisel is maybe half the width of the wall. It takes 2 (or 3 if I overlap) cuts to complete each side. I keep my chisels honed. Through dovetails have been fine.

The shoulders are tight enough, so it's not like the whole tail board is shifted out.



The good news is, I can do this consistently. 


Anyone want to armchair QB this one?

Thanks in advance.

[Hank Knight]

Aram,

I had the same problem when I started making half blind dovetails. I never could figure out where my technique was wrong. I suspect that I was just a little sloppy in several aspects of the process and that produced the gaps. I built a jig to help me with the job and it solved my problem completely. Others have lambasted me for using a "crutch," but I don't care. It works for me and gives me perfect half blinds every time. Here's the process:

First I select a short offcut the thickness I want for the end wall of the pin board. It's laying on the jig in the photo below behind the holdfasts to the left rear of the workpiece (the black thing is a chisel edge guard). I lay it flat on the bench and butt the end of the pin board against it. I strike a line with my marking knife, registered flat on the cutoff, to establish the end wall (this will make more sense when you look the jig photos below). I set my cutting gauge from this mark for the depth of the tails. Mark the tails, saw and chisel them. Then mark the pin board from the tail board, being careful that the two boards are square - Dereks blue tape trick will work here.

Now the jig comes into play. Here it is:

DovetailJig002Medium by Hank Knight, on Flickr

As you can see, it consists of a base board with a cross piece that is carefully milled at 90 degrees to serve as a clamp and a vertical chopping guide for the chisel. I applied some PSA sandpaper to the bottom of the cross piece to keep it from slipping on the workpiece.

After sawing the pins, I clamp the pin board in the jig, carefully aligning the cross piece to the base line. I've found that registering a chisel in the knife line and butting the cross piece against it helps with the alignment. Clamp it down securely with the screw clamps. My jig in the photos has cam lever clamps, but they are not as convenient as they look. I've since switched to screw knobs. Springs on the bolt shafts help with loosening the clamp. Chop the pin sockets until you get near the base line and the end wall, then chop the base line, using the cross clamp to keep the chisel vertical, similar to what you're doing now.

DovetailJig007Medium by Hank Knight, on Flickr

When I approach the final depth of the sockets (the end line I marked from the offcut at the beginning), I take the offcut and place it flat on the jig to use as a reference surface for paring the ends of the sockets. This insures that the end wall of the pin board is a consistent thickness for it's entire width. It has completely eliminated the gaps I was getting like yours.

DovetailJig008Medium by Hank Knight, on Flickr

This jig insures consistency in cutting both the base line and the end wall on the pin board. You can also use it to chop the base line on the tail piece if you choose. I don't have much trouble with the tail piece, so I only use it for the pin board. It works well for me and gives me perfect half-blind dovetails every time.

Hope this helps.

Hank

[titanxt]

Aram,

I had the same problem when I started making half blind dovetails. I never could figure out where my technique was wrong. I suspect that I was just a little sloppy in several aspects of the process and that produced the gaps. I built a jig to help me with the job and it solved my problem completely. Others have lambasted me for using a "crutch," but I don't care. It works for me and gives me perfect half blinds every time. Here's the process:

First I select a short offcut the thickness I want for the end wall of the pin board. It's laying on the jig in the photo below behind the holdfasts to the left rear of the workpiece (the black thing is a chisel edge guard). I lay it flat on the bench and butt the end of the pin board against it. I strike a line with my marking knife, registered flat on the cutoff, to establish the end wall (this will make more sense when you look the jig photos below). I set my cutting gauge from this mark for the depth of the tails. Mark the tails, saw and chisel them. Then mark the pin board from the tail board, being careful that the two boards are square - Dereks blue tape trick will work here.

Now the jig comes into play. Here it is:

DovetailJig002Medium by Hank Knight, on Flickr

As you can see, it consists of a base board with a cross piece that is carefully milled at 90 degrees to serve as a clamp and a vertical chopping guide for the chisel. I applied some PSA sandpaper to the bottom of the cross piece to keep it from slipping on the workpiece.

After sawing the pins, I clamp the pin board in the jig, carefully aligning the cross piece to the base line. I've found that registering a chisel in the knife line and butting the cross piece against it helps with the alignment. Clamp it down securely with the screw clamps. My jig in the photos has cam lever clamps, but they are not as convenient as they look. I've since switched to screw knobs. Springs on the bolt shafts help with loosening the clamp. Chop the pin sockets until you get near the base line and the end wall, then chop the base line, using the cross clamp to keep the chisel vertical, similar to what you're doing now.

DovetailJig007Medium by Hank Knight, on Flickr

When I approach the final depth of the sockets (the end line I marked from the offcut at the beginning), I take the offcut and place it flat on the jig to use as a reference surface for paring the ends of the sockets. This insures that the end wall of the pin board is a consistent thickness for it's entire width. It has completely eliminated the gaps I was getting like yours.

DovetailJig008Medium by Hank Knight, on Flickr

This jig insures consistency in cutting both the base line and the end wall on the pin board. You can also use it to chop the base line on the tail piece if you choose. I don't have much trouble with the tail piece, so I only use it for the pin board.  It works well for me and gives me perfect half-blind dovetails every time.

Hope this helps.

Hank

Hank, that is an awesome jig! I will have to give it a try!

[Handplanesandmore]

"Crutches!"

So you know: James Krenov used this kind of "crutch" to chop his dovetails too.

Simon

[Derek Cohen]

Lately I have been consistently getting a gap on my half blinds, between the end of the tail board and the... I wish there was a standard word for it. Ian Kirby calls it the "end wall" so I'll go with that. 

Hi Aram

The reason for the gap is one of three possibilities:

1. The baseline has been marked a tad further away than it should have. This is due to the set up of the gauge (nothing to do with blue tape and transferring marks). For example, the pins have been marked deeper than the tails.

2. The second possibility is that there is an interference inside the socket, such as waste that has not been fully removed at a corner. This will prevent the joint closing up. Undercutting (slightly) the socket at the baseline helps prevent this (Undercutting here is OK since the section is end grain to end grain, and not structural).

3. The third possibility is that you have inadvertently pushed back the baseline when chiseling. This occurs when you place the chisel in the knife line too soon. "Too soon" is when there is more than 1mm of waste in front of the baseline. Hank's suggestion of a fence to pare against is one way of controlling this (incidentally, the jig is one I had published in FWW magazine about 15 or more years ago). The method I use today is to undercut the baseline with a chisel wall. This prevents the chisel being pushed back.







Regards from Perth

Derek

[Aram]

When I approach the final depth of the sockets (the end line I marked from the offcut at the beginning), I take the offcut and place it flat on the jig to use as a reference surface for paring the ends of the sockets. This insures that the end wall of the pin board is a consistent thickness for it's entire width. It has completely eliminated the gaps I was getting like yours.

DovetailJig008Medium by Hank Knight, on Flickr

Thank you. This is clever.

Hi Aram

The reason for the gap is one of three possibilities:

1. The baseline has been marked a tad further away than it should have. This is due to the set up of the gauge (nothing to do with blue tape and transferring marks). For example, the pins have been marked deeper than the tails.

2. The second possibility is that there is an interference inside the socket, such as waste that has not been fully removed at a corner. This will prevent the joint closing up. Undercutting (slightly) the socket at the baseline helps prevent this (Undercutting here is OK since the section is end grain to end grain, and not structural).

3. The third possibility is that you have inadvertently pushed back the baseline when chiseling. This occurs when you place the chisel in the knife line too soon. "Too soon" is when there is more than 1mm of waste in front of the baseline. Hank's suggestion of a fence to pare against is one way of controlling this (incidentally, the jig is one I had published in FWW magazine about 15 or more years ago). The method I use today is to undercut the baseline with a chisel wall. This prevents the chisel being pushed back.

Derek

Thanks, Derek. I spent a lot of time systematically eliminating 1 and 2. I'm fairly sure it's not those. In any case, I used a double square to measure  the depth to the end wall, as well as the length of the tails. The first is deeper. I am cutting past the scribe line for sure. What bothers me is, and I might not have been clear about this above, I crept up on it. When I dropped the chisel into the scribe line, there was not much wood left to pare. Much less than 1 mm. What bugs me more is that I made some really tight half-blinds a couple of months ago! Whee. 
 

I've read your undercut description before. Question. Are you doing this on the end grain cuts that I am having trouble with?

Thanks,

[Hank Knight]

Looks like the old saying is true: There's nothing new under the sun. I was proud of myself for coming up with my half blind dovetail jig. I put it together from separate ideas I'd seen for using a backer block for maintaining vertical for chopping and a spacer for horizontal for paring. Thought it was my design. Cudos to Derek for coming up with it before I did. He said, "[I]ncidentally, the jig is one I had published in FWW magazine about 15 or more years ago." Had I known I was copying Derek's jig, I would have credited him with the idea. My apologies, Derek.

Hank

[Rob Young]

Looks like the old saying is true: There's nothing new under the sun. I was proud of myself for coming up with my half blind dovetail jig. I put it together from separate ideas I'd seen for using a backer block for maintaining vertical for chopping and a spacer for horizontal for paring. Thought it was my design. Cudos to Derek for coming up with it before I did. He said, "[I]ncidentally, the jig is one I had published in FWW magazine about 15 or more years ago." Had I known I was copying Derek's jig, I would have credited him with the idea. My apologies, Derek.

Hank

The little horizontal paring block, seen David Charlesworth use that in an "old" video from Lie-Nielsen.  Don't remember if there was a vertical paring/backer block. I think the ideas are also hinted at in some old "Woodworker" articles from Hayworth.  But without the fancy springs and screws and stuff.  Just some "G cramps" and away you go.

If the gap is small enough and the front reveal lip thick enough (probably not the case in the original photo), a variation on "bishoping" might solve the gap problem.  Extra wet glue and give it a smack and clamp to force the swelled fibers to close up the gap.  Then a quick plane to take out the ghost that might show up in the drawer front.  But as thin as the reveal is in the drawer in question, that is probably NOT a good method.

[Handplanesandmore]

Thought it was my design. Cudos to Derek for coming up with it before I did. He said, "[I]ncidentally, the jig is one I had published in FWW magazine about 15 or more years ago." Had I known I was copying Derek's jig, I would have credited him with the idea. My apologies, Derek.

Hank

It was your design as you had not known about the same jig others people had come up with. It was not an act of copying by any definition.

It is not uncommon that in the research field, many people or teams are working on the same idea without each other knowing.

Simon

[Mike Brady]

I use that same paring jig that Hank Knight shows.  My spacer stock is almost always masonite in either 1/8" or maybe 1/4".  Set the marking gauge from that stock.  

This type of paring jig is almost fool-proof, and when you are investing the time and materials in dovetail construction, I think that anything that eliminates errors is worth evaluating.  Not all work for everyone.  For example, driving a card scraper into the end grain to cut the web left behind by the saw has not worked well for me.  There are tools made for this.  Ron Bontz makes a nice one.  Might have to give that a try.