REVISITED - Make a perfect "MFT" with Qwas raildogs

[greymann]

As I am building my new workbenches out of 80/20 extrusions, I had planned on just finding a CNC shop to make the custom tops.  However, I found that in my area  (Asheville, NC)  the shops I could find didn't have machines up to the task.  I was a very frustrating couple of months.  

Anyway I was left to make them myself and Michaels article looked like the way to go.  I only had problems with two steps, one and two.  They were to first make the blank dead square and then to make two "exactly parallel" rows of holes on either side of the blank.

So I worked on taking out what I felt, given my skill level, was uncertainty in those two steps.

First I started with trimming one edge.  Then I mounted the Qwas raildogs to the holy rail and pushed them against the edge.  I also used one of the lr32 endstjops to register the rail from side to side.

[attachimg=#]

[attachimg=#]

I was just doing a proof of concept so just eyeballed where I wanted the holes to be positioned on the blank.  Then I drilled the first row of holes.

Then while leaving the endstop on, I reset the Qwas raildogs to insert into the two end holes of the first row.

[attachimg=#]

Once the raildogs were set the endstop is removed which allows the blank to not be perfectly squared and the process to still work.  With it reset I then drilled new holes at the end positions.  

[attachimg=#]

Repeat until you get to the far side of the blank and then drill the second row of holes.

[attachimg=#]

From then on it is Michaels brilliant method and the results look the same.

[attachimg=#]

I found that I felt in control of the process and it only took a couple hours.

Right now I only have a Whiteside 25/32 straight bit and it is too snug for the Qwas dogs but I can get some others to just fit.  I have the Festool 20mm bit on order and will wait until then to do the fullsize top.

 

[PaulMarcel]

Nice procedure; completely never thought of pushing the raildogs up against the side of it.. clever

 

[jimbouk]

Very cool thanks for posting as i was just tryining to work that out myself before i do mine!

 

[greymann]

Ok, before anyone else says it I'll admit I'm getting very close to obsessive in trying to make the perfect MFT.  Since the last post I have been trying to find ways to reduce errors.  I was sure I had errors in the process because the acid test of placing my top over an MFT top, in my case a 1080 top, was unsatisfactory.  As I looked down at the holes, they started very close and then steadily diverged until mine was long by more than a mm at the other side of the top.

After reducing all the process errors I could think of I pulled out the measuring tape to see if I could find another reason.  To start I put my true 32 tape on the lr32 rail and it was as perfect as my eyes could detect along the entire length.  There was certainly no accumulating error.

Then I looked at the boards that have the step 2 holes in Michaels method.  At this point I am doing these on jig boards not the blank that will be the final top.  Since the lr32 didn't have errors, I hoped I wouldn't see errors in these holes.  And I didn't.  I started from the right setting a 32 tick on that hole and then seeing if it stayed true at the other end.

[attachimg=#]

[attachimg=#]

So far so good.  Then I looked at the test board I had drilled and saw the same, no noticeable error.

[attachimg=#]

[attachimg=#]

That didn't leave many options for why the top didn't line up with the 1080 top.  I of course was assuming that the 1080 top was perfect.  Well just for completeness I measured from one end of the 1080 top to the other.  And found this.

[attachimg=#]

[attachimg=#]

The holes in the 1080 top were off by nearly 2 mm over its length.

So question.  Is anyone else seeing this?

I won't affect squareness as long as it is consistent.  But I have to admit to shock at having the potential of making a top more accurate than what I was using as the reference.

Dick Perry