Category Archives: What I Screwed Up

A close look at mistakes that I make, and how you can avoid them.

What I Screwed Up This Week – Molding Miscalculation

Welcome back for another educational issue of What I Screwed Up This Week!  This time, we take a look at a molding miscalculation from the current blanket chest project.

The Problem:

Panel Molding 2

My plan was to have the panels of the blanket chest set in 3/8″ from the front of the frame.  This would allow a 1/8″ shoulder at the top to provide a visual break between the molding and the frame.  However, there was one problem.  The above drawing reflected my thinking.  I visualized the panel as a flat surface, rather than having a raised bevel.

Panel Molding 3

The reality was far different.  The sloping bevel of the panel raised the molding upwards.  While the drawing above is exaggerated, the oversight is clear.  The shoulder at the top of the molding becomes too small, giving no real definition to the transition.

The Solution:

Panel Molding 4

I played with several ways to rectify the problem.  However, the simplest turned out to be the best, and easiest to implement.  I simply trimmed down the upper edge of the molding to recreate the shoulder.  Surprisingly, it had very little effect on the appearance of the cove – much less than I had anticipated.

Blanket Chest 30

To remove this wood, I took a page from Matthew Bickford’s blog, Musings from Big Pink, and made a small sticking board.  This allowed the small piece of molding to be held in place simply by the friction of planing it.  The taper of the side support reduced the area of contact for the plane, and acted as a sort of depth stop (or at least reference) once the first piece had been planed to the desired thickness.

Blanket Chest 31

A few strokes with the plane were all that were necessary to reduce the height of the molding enough to create a viable shoulder at the top.  Here, you see a piece coming to final thickness.

Lesson Learned:

Angles can change everything.  Even a small slope or curve can drastically affect how pieces interact with each other.  Be sure to do mock-ups of any such areas, and determine if you need to make any modifications to your design before you paint yourself into the proverbial corner.

What I Screwed Up This Week – March 1, 2011

Welcome back for another edition of What I Screwed Up This Week!  This week, we look at a difficulty with chisels.

The Problem:

Roubo Bookstand 8
While I was paring the hinge knuckles for the Roubo bookstand I built back in February , I ran into trouble keeping a 45 degree angle consistently for each knuckle.  The edge angles were easy since I had a marked line on the side of the board to follow.  However, I lost this reference on the inside knuckles.

The Solution:

I took my sliding bevel, set it to 45 degrees, and then used it as a guide to help me line up my cuts.  This was better, but still allowed for variable results.  I was still freehanding, and the gap between the chisel and the bevel necessitated by my hand made this approach assisted eyeballing at best.

I would have gotten better results if I had cut a piece of wood to the 45 degree angle and used it as a guide, at least for the last critical cuts.  This is a trick that I once saw Tommy MacDonald use on his old podcast, and I didn’t remember it until after the fact.  It solves both problems at once, but is really limited to lighter cuts due to the hand position required.  A conbination of the two techniques would probably yield the best results.

Lesson Learned:

Even though they’re called hand tools, don’t be afraid to come up with guides or jigs to help with alignment.  The old-timers were smart enough not to work harder than they had to.  Once again, this is an area where going back and reading some of the historical texts can give you lots of insight.  Oh yes, and don’t forget to keep those chisels sharp.  Nobody gets good cuts with a dull blade.

What I Screwed Up This Week – December 12, 2010

Welcome back for another edition of What I Screwed Up This Week!  Once again, we’ll be looking at a problem with building the turning saw.

The Problem:

Screwup 121210

After affixing my templates to the hickory, I went ahead and cut the pieces out without taking into account that I needed to bore pin holes (bottom circles) and chop mortises (upper circles).  This left me in a situation where the pieces would not be properly supported during these operations.

The Solution:

Turning Saw 4&5I used cabinetmaker’s handscrews to hold the pieces in alignment during the drilling and mortising.  Handscrews can, when properly adjusted, provide excellent holding power for irregularly-shaped objects, and can act as a flat reference surface when performing operations like drilling.

Lesson Learned:

Whenever possible, plan to do operations like mortising or drilling BEFORE cutting the curves on parts.  It’s often possible to establish flat reference surfaces parallel or at right angles to the area that needs to be worked.  Then, proceed with shaping the piece after the machining or joinery is finished.

What I Screwed Up This Week – December 6, 2010

Welcome back for another edition of What I Screwed Up This Week.  This week’s snafu happened while I was cutting out the cheeks and stretcher for my Gramercy-style turning saw.

The Problem:

As I was cutting out one of the cheeks, an endgrain offcut fell off in two pieces.  This was unexpected, and I stopped to investigate.  I found that a check from the end of the hickory board extended through the end of the piece I was cutting:
Turning Saw 10

The Solution:

The only solution in a case like this, where the wood will be under tension and require no weak areas, was to discard the piece and make another one from sound wood further up the board.

Lesson Learned:

Be very careful when laying out pieces near the end of a board.  Checks from drying can easily go undetected and can quickly ruin a piece.  In some cases, they don’t appear on the surface, but cause the interior fibers to separate.  It’s always a good idea to cut off the very end of the board to expose fresh surface, which can help to reveal any checking.

What I Screwed Up This Week – November 28, 2010

Well, I hope everyone in the States had a very happy Thanksgiving.  As usual, I ate too much and had to work the weekend.  However, I still managed to get in some shop time.  That, of course, leads us to this edition of What I Screwed Up This Week.  This one actually happened a few days ago, but I just now worked it in to the lineup.  It actually relates to stock preparation that I did for the frame of the Wing Chun dummy.

The Problem:

As I was ripping 2×10 stock into 4” widths for the frame components, reaction wood caused one of the 8’ pieces to bow inwards.  This is always a prime cause of kickback, but fortunately, my SawStop has a very good riving knife, so I was in no immediate danger.

The Solution:

Wing Chun Dummy - Prepping the Stand Stock 1Interestingly, the piece was only bowing on the left-hand side, away from the fence.  The fence-side part of the wood remained perfectly straight.  Such things can occur as a result of varying stresses on different sides of the log caused by unequal growth.  Due to the fact that I still had a solid fence reference, I decided to continue with the cut, but remained well out of the way.  The cut went just fine until the last inch, when the built-up stress of the curving piece pressing against the far end straight piece caused the two parts to separate rather violently, sending the left-hand piece almost a foot across the table.  You can see the result in the photo to the left.

Lesson Learned:

Always be alert for reaction wood.  A piece such as the one above that spans the pith of the tree is a prime candidate for this.  If you notice bowing in a piece as you cut, stop immediately and asses the situation.  If you don’t have a riving knife, DO NOT CONTINUE THE CUT.  Even if you have one, don’t continue if the wood is warping away from the fence.  In retrospect, I underestimated the force of the bowing action of the wood, and had an unexpected occurrence at the end.  I probably should not have continued the cut either.  Remember that if wood shows signs of movement, it’s not going to stop until the energy is expended.  Also, never forget that tablesaws don’t have a very forgiving nature.  When in doubt, shut it down.

What I Screwed Up this Week, October 29, 2010

This is the last installment for the outfeed table.  I’m not sure what it says about my woodworking skills that I got three different screw-up articles from one project, but at least you get to learn from my mistakes.

The Problem:


The outfeed table was intended to push up flush against the back of the tablesaw, but, as the photo shows, the leg of the table lined up perfectly with the slightly-splayed leg of the saw and prevented a flush fit.  Tthe table’s overhang is kept back from the edge of the saw top about two inches.  This isn’t enough to cause problems, but is not what I intended.

The Solution:

At this point, there really isn’t one.  If I had recognized this problem sooner, I could have left the overhang of the outfeed table longer, so it could reach over the rail all the way to the table.  As it is, I’ll just have to live with it.

Lesson Learned:

Never forget to consider how your piece will interact with its surroundings.  While you can’t predict everything, repeated checks and test fittings before important glue-ups or other one-way steps will greatly reduce your grief.

That’s it for this project, but plenty more are on the way in the near future, so stay tuned to find out What I Screwed Up.

What I Screwed Up This Week – October 20, 2010

I mentioned that the outfeed table should provide plenty of food for thought for this column, and this proved true. The first problem occurred while assembling the plywood frame for the table top.
The Problem:
As I drove the first pocket screw to assemble the frame, it went through the bottom of the hole, and split the end of the plywood apart.  I was puzzled – the screw had worked just fine when I did some test joints in pine.  Obviously the torque setting was too heavy for plywood, and had spilt the plys apart rather than stopping when the screw bottomed out..
The Solution:
I backed the screw out, glued and clamped the split plys back together again, and redrove the screw with a correct torque setting.
Lesson Learned:

Always make a test run of joinery techniques using scraps of the same material as your project.  This will ensure that everything works correctly, and you don’t get surprised on that first connection like I did.

The second problem was noticed the morning after I had attached the frame to the top.

The Problem:

The plywood top had developed a pronounced bow during the night.  This was despite the fact that the support frame had been screwed into place. 

I had tried to save some money by using a B-C exterior plywood. I reasoned that only the legs would be visible, an the frame would provide plenty of rigidity.  I was wrong – the plywood was not stable enough for this type of light support, and bowed 3/4″ in the middle.

The Solution:

There was only one thing to do – replace the plywood top.  Accordingly, I went back to the lumberyard and got a sheet of cabinet-grade maple.  This stayed obligingly flat and allowed the project to continue.

Lesson Learned:

Construction plywood is for just that – construction.  It rarely has the degree of flatness required for furniture.  In construction, a certain amount of warp is acceptable since the plywood sheets will be firmly nailed to a sturdy frame such as a stud wall or roof rafters.  In furniture, rather than being attached to a structure, plywood often IS the structure and flatness and stability are paramount.  A sheet of cabinet-grade maple was only $11 more.  Spend it.

That’s it for this episode.  Stay tuned for more goodies in the future.

What I Screwed Up This Week – October 13, 2010

As I mentioned last time, there were two errors that I made during the construction of the sauce packet bin.  The first had to do with planning for installation.  If you missed that one, look back at my post on October 7.

The Problem:

WISU 10/12/10 1The one covered in this episode was also the result of a failure to plan.  (I think I’m starting to see a pattern here.) In this case, it had to so with trimming out the pivot blocks.  When I cut them out, I made them the same height as the bin ends.  Normally this wouldn’t be a problem, as I usually cover plywood edges with edge banding.  However, this time I was planning on using 1/4″ stain-grade birch plywood.  As you can see from the trim strip perched on top of the pivot block in the photo, this presented a problem.

The Solution:

WISU 10/12/10 2This was a simple enough fix.  I removed the bin from the shelf and trimmed 1/4″ from the top of the pivot blocks.  This allowed the trim to nestle into place quite nicely.  Why not the front as well?  If you remember from the first installment, I had planned for the front panel of to overlap the pivot blocks, and it was made from the same plywood.

Luckily, the entire assembly was secured with some screws, so re-installation was not a problem.  The trim was then cut to fit and glued in place and we were ready to paint.

Lesson Learned:

Never forget that components have thickness.  This may seem obvious, but can be easy to overlook.  I’ve caught myself almost making this mistake on bookshelves – calculate the space between the shelves, but forget to allow for the thickness of the shelves themselves.  Small things like veneer, edge banding, and even film finishes like varnish or paint all add a thickness that can be at least felt with the fingers, and can influence the fit of components.  As a last step in any design, at least do a layout of certain key components as a crude sketch and make sure that all your measurements add up.

Later this week I’ll be starting on the outfeed table for my tablesaw, and I’m sure this project will yield more grist for our error mill.  Stay tuned!

What I Screwed Up this Week – October 7, 2010

This is the first article in the “What I Screwed Up This Week” series, and my in-progess sauce packet bin provides a couple of good first example.  For full details of construction, please refer to the related posts blog posts on the bin itself.

What constitutes a “good” screw-up?  For the purposes of this series, a good screwup:

  1. Is not safety related.
  2. Is not terribly difficult to fix, just enough so to make it memorable and,
  3. Provides a good lesson for our readers.

The Problem:

This week’s  problem actually began during the design phase.  I designed the bin as a rectangular box sandwiched between two pivot blocks.  The front panel was designed wider so that would overlap and hide the pivot mechanism.  Unfortunately, I forgot to take one thing into consideration – getting the assembly into place.
WISU 10/7/2010
The cabinet side had an oversized face-frame that prevented sliding the bin assembly straight in – it had to be angled.  Unfortunately, the total assembly was too wide to slide in at an angle with the pivot blocks in place, and they HAD to be in place.  Hmm…  The only way to do it was if the right pivot block was rotated 180 degrees outwards to angle into place, then rotated back into position.  The only way to rotate this way was to (yep, you guessed it) cut the “ears” off the front.  So, out came the router and off came the ears. Now, I’ve got to work out a different trim scheme.

I realize pictures of this would make everything clearer, but there was no way to get the photographs I needed in the space where I was working.

Lesson Learned:

It’s not enough to make sure a component will fit a space – you have to plan for how you’ll get the item INTO that space.

The classic example of this is the entertainment center that, when completed, won’t fit out the shop door.  In my case, I neglected an obvious restriction in the form of a face frame that should have indicated a different approach to the problem.  Perhaps something like attaching the front after the rest of the frame was in place.

Working in cramped quarters is never easy.  Narrow-focus consciousness tends to cause us to “not see the forest for the trees”, and important details are easily overlooked.  Whenever possible, make a mockup of your design to allow you to test for various problems that might occur during installation.  In my case, taking a measure of the width of the area overlooked the face frame, and caused a major revision.  If I had simply held the measurement and come straight out with the measuring sticks, I would have run into the face frame and been alerted to the problem.

This revision and installation revealed another problem, one that required the bin assemly to be pulled back out.  However, we’ll save that one for next time

Do you have thoughts on the subject?  If so, please leave a comment below.  They say that two heads is better than one, and I hope we’ve got more than that out there.