Tigging tabbed dropouts

There are several ways to Tig tabbed dropouts but I’ve always wanted to emulate the type of dropout/stay connection found on my Y2K Independent Fabrication cross frame.  It’s a weird type of joinery for me to wrap my head around for some reason. I tried once on my own cross frame with Paul’s horizontal dropouts and it went just OK, but with using tabbed dropouts and larger diameter stays of the straight-gauge 4130 type…it becomes a necessity.  I put out the question to the VS forum since i knew a few people that utilize this method frequent that site.  Thanks especially to Drew and Rody for their help, among others.  Fortunately, I had gathered most of their recommendations by looking at their websites and Flickr photos because by the time anyone responded I had already given it a go.

The biggest thing I learned were that these brand of dropouts may not work the best for this type of design because they have pretty thin tabs (not plate thickness, but the width of the tab).  I don’t make my own dropouts so I’ll be looking for other options or only using this method once and awhile…because it is a time sink.

Here are some fabrication shots to show what I did.  Like they said in the forum said, it’s a difficult process and larger tabbed dropouts would be the call.  That’s why most builders that do this style of attachment make their own dropouts.


Using the Paragon yoke

I used the Paragon Machine Works chainstay yoke for the first time on Adam’s 29+ frame last month. The yoke was created to accommodate 3″ tires and allow the use of a mtn crankset with a regular chainline on a 73mm BB shell. This is a difficult task to say the least without the use of a yoke.  Possible but lots of dimpling and crankset/ring size selection is needed in my experience.  There’s a reason Surly created the Krampus with a yoke when they created this new genre.

Here are some notes that I thought is share to help others use this awesome machined yoke:

  • There are different ways to go about using it but in general, you fuse the two pieces together, bend then cut the 7/8″ x 0.035″ to length, mitering chainstays for the dropouts, join the yoke to the chainstays, miter the yoke for the BB, and then fixture it up and make some tacks. Then I tacked it to the BB before welding it all up outside the fixture but that’s just one way to do it.
  • Be sure to draw in CAD or paper accurate chainring clearances in order to find the right length to miter the yoke’s “arms” for the BB. They needed to be longer than I thought to fit a regular chainline especially if you want to run a 2x crankset or 32t single ring on the inside position of the crank. If this is taken into account there will be plenty of room for the torch and welding to the BB.
  • Once you have that settled figure out the length of the 22mm (7/8″) x 0.035″ chainstays that plug into the yoke ends. Add at least 7mm to the tire radius to clear the inside of the yoke. I’m not sure the shortest chainstay length you can get but I doubt you’d be able to get anything less than 435mm unless you sacrificed chainline and chainring size (use a 28t). Get it right as you can’t dimple this yoke!
  • Bend the 7/8″ stays to allow crankarm clearance. This is necessary! The yoke is wide and with 22mm round chainstays the bend needs to go straight back and then bend out after the crankarm distance.  I had to put a slight dimple in the stays to clear a 2x Shimano crankarm since I didn’t roll the tube enough.
  • To properly backpurge the yoke you’ll need to take an 1/8″ endmill to the inside of the yoke at a few spots and clear a path.
  • I tried both welding with 0.035″ rod and fusing without rod to attach the yoke pieces. I liked the results of fusing better and it leaves less cleanup afterwards (sanding down the bead to get a clean finish).
  • I left all the pieces loose in the fixture which was difficult to keep in one place and straight while tacking. Next time I’ll likely create the full subassembly outside of the fixture (using my chainstay mitering fixture) and then miter for the BB.  That way is more seamless IMO but better get the clearances worked out.
  • I used 0.045 rod to weld the chainstays to the yoke ends. There was a small gap because of how the yoke ends are shaped. It’s pretty air tight but i think it’s needed to get filler in there to get a strong connection.

hmm, all i can think of for now. I’d love to hear if anyone’s got some comments on their experiences with this yoke.

Process shots from Adam’s 29+

Adam’s frame is 95% done. I just need to cap the wishbone stay ends, put on a few cable guides, ream/face, and clean up before posting any final parting shots.  It was a challenging build not just because the bent tubes but because it was so interrupted by time away from the shop. I have to say, if i wanted a custom 29+ this is what I’d order too.  Maybe with internal cables though.  I’ll post geometry and more details when I post the ‘final’ shots before it goes to paint. But for now here are some process shots for the FB dorks out there.

Today in the shop

Some process photos from today in the shop. Working on my new fatbike…more details later.


A good hole saw is so so nice. No post-filing at all. That is rare for me.


Internal housing. This still gives me the heebees.


finished hole. Using a 5/16 end mill for the center, 1/4″ end mill for the sides. Helps the brass tube fit at such an extreme angle as it exits the tube.


This went surprisingly fast and smooth. The 1/4″ brass tubes are much easier to use than the 9/32″ tubes.


just drilling a breather


This was the 3rd cut on this hole saw on ‘normal’ and then BAM…a Supertherm downtube breaks off some teeth. It wasn’t even that out of round.


Facing the arbor side of an “Ultra” hole saw. This one was out of round too (aren’t they all?).


Step 1 of my new high tech way of finding the miter angle on bent tubes. SHH, super secret dude.


Step 2: after zeroing the angle finder on one of the tubes, check the angle on the other and there you have the miter angle (or actually if you subtract that from 90 you have it). Works like a charm.


Front triangle all mitered up. Seat tube welded to BB and with a fused collar in there, h20 and breather holes drilled, a good day in the shop.

S-bend seat stays

This was my first attempt at S-bends seatstays. I have built two frames with single-bend stays but those were a couple years ago and I did the single bend on my 7″ radius oak fork blade bender.  After hummin and hawing for awhile and attempting to make my own somewhat clunky 8″ radius bender out of MDF,  I decided to fork over the Franklin’s and get a properly made tool that will last for a long time and make life easier.  There are many bender options – from expensive used DiAcro’s that need customization, to new benders from ProTools and the like – but most of them need some customization to get the job done right (to fully encapsulate and support the tube so it doesn’t fold or flatten).  After looking at simplicity, cost, amount of work needed to do what I want, the Anvil was the obvious choice for me.  So for many frames to come I’ll be using my new Mr. Bender Rodriguez.  I got the 3/4″x 9.375″ mandrel for doing anything from 5/8″ stays to fork blades (yet to attempt that).  With it I can get away from mostly doing my time-intensive segmented wishbone stays to quickly bending single or S-bends in two tubing diameters (5/8″, 3/4″).  After practicing on some 4130 x 0.035″ and 0.028″ tubes, i went for it with some Nova single taper stays.  There was a little bit of flattening on the bend (5/8″ tube, 3/4″ mandrel) but almost nothing to write about.  After getting the initial bend, you flip the tube for the 2nd bend and use the phase keeper arm to make sure the bend is in the same plane as the 1st.  Then when it seemed like it’d work (comparing it to the life-sized drawing), I measured down from the tire mark on the stay/apex of the 1st curve (widest part of tire), and marked where the top of the slot needed to end up.  Then, i cut and slotted the stay for the dropout in the mill.

For now my ‘custom’ seatstay mitering fixture will work.  It’s not super pretty but i also didn’t have to file to fit at all after getting the right miter length.  The aluminum extrusion doesn’t clamp that well in the mill vise, but it’s mostly because i’m clamping so little of it due to the angle of the miter.  It all stays put after some serious force but I’m thinking that I need to fabricate an updated mount. But for now this will be my only non-Anvil fixture 🙂

Slotting and other stuff

Working when I can on Timmy’s frame.  Headed down to my Dad’s (both of them – step and real) for Dad’s day last weekend so didn’t get any shop time in for a few days.   I’ve been fighting a weird bug as well that took me out of commission for a couple of weeks.  Brutal.  No riding and only a few hours in the shop since the full body fatigue and night sweats were beating me down.  It’s the first time i have been sick in years — literally YEARS. Must be the California microbial community’s Welcoming Committee. “Welcome to California, not take THAT!”

Here’s some random shots from the build.  Got the slotting of the stays dead on…no slop at all which is sweet for brazing the stainless sliders with silver.  I’m getting used to slotting on the horizontal using the MTMFixture.  It’s SO clean and fast and keeps it all in phase with no effort at all.  I feel like I’m cheating not using the hacksaw but I still have to file to fit after i tear out the tongs.  The reason I don’t use a wider slotting saw is because I found out the hard way that not all Paragon tab dropouts are 6mm thick. Rocker dropouts are 4mm at the tab and the rest of the Paragon offerings that I’ve used (Sliders, others) are 5mm.  I also can’t find a slotting saw with as high a tooth count as these Malco’s that cut SO smooth and with no snagging.  I am going to try to find or make some spacers that I can put in between the paired saws and space them right for the dropouts, but this is really pretty easy as is.  I just measure how far i want the slot from one side (10 degree cut per the rotary table), cut one time to 12mm deep, and then move the Y-axis feed 3 or 4mm (gotta account for the blade width) to get the right sized slot.  Rip out the tong with the needle-nose pliers and start filing.  I recently ‘retired’ a couple of flat files by filing them down on the disc sander (Paterek’s idea) – one is smooth on the sides so I can file the sides of the slot without making a deeper cut; one is smooth on the faces so I can file a deeper slot without accidentally taking some off the sides.  Yes, totally unnecessary but really nice to have the option.

I also got around to drilling and sticking two internal brass tubes in the bent top tube.  It was a total beeyatch to get the second one through.  The trick I learned was to get some old brake cable housing and run it through the brass housing so you can first feed the flexible brake housing through the other end of the top tube and basically guide the brass housing up and out the hole.  Takes some doing in a bent tube because of the resistance (good luck pre-bending two to fit in a bent 31.8 top tube…) but it works.  The rear brake and rear derailleur cables will go in the left side of the top tube and out opposite sides just before the seat tube.  Then they’ll be routed externally down the underside of the seatstays.

File to Fit

It’s been a long time since I’ve ranted about framebuilding here (I do it every day in my head but rarely have time or want to spend it writing thoughts that few really care about). But hell, what good are ideas if you don’t put them out there to bounce off the interwebs?! That’s what blogging is after all, banter that may or may not be interesting to a small clique of the world.

So anyways, I have finished my 17th frame and it was a doozy.  I can’t say I enjoyed the entire build, but seeing it done is pretty satifying.  It’s almost the frame that hate built, to use the Over-Opinionated Framebuilder’s phrase.  En sum, I went through one extra top tube (cut too short, stupid SOHCAHTOA), one extra wishbone seatstay crown (burned a hole), two extra seatstay legs (slotted too short), one Ti coated drill bit (snapped at the tip), and a lot of brain cells.  The bending of tubes and the resultant geometry changes to the miter lengths and miter angles are stifling me. I am trying to use BikeCAD to get the miter angles and lengths but for bent tubes I now am just going to draw it all out life-size.  You can see some of the differences in the mitering numbers from what I had modeled in BikeCAD to what I drew out on paper in the picture below.  I’m sure many others can correctly use BikeCAD to input the right radius numbers…but I can’t.  Like i’ve said before here, I haven’t a clue how to get a given radius from the Harbor Freight tube roller unless I buy a expensive set of gauges that only help after I’ve rolled the tubes.  Ok, so I guess I could roll/check/roll/check to approach the radius I’m looking for but I haven’t tried that yet.

So what I do instead is file to fit. I am pretty sure most builders that have milling machines and expensive fixtures still file to fit a little bit but for their sakes I’m hoping they don’t do it quite as much as I do.  There was only one miter on a curved tube I got spot on and didn’t touch after the mill – the wishbone seatstay miter to the seat tube (17 degrees).  I’ll remember that number for some stupid reason for a long time. It’s stored in the same part of the brain that remembers the lyrics to those 80’s songs you really truly hate.  Of course I won’t remember how I finally figured out the right way to measure and miter the curved top tube angles.BikeCAD-notes

In the photo of the drawing only the circled numbers stayed constant. Really those are the most important since they are what control the geometry of the frame.  They’re what I set the fixture at and try not to move it more than 1mm (if i cut a tube too short).  They’re what define how the bike handles. Oops, just noticed I forgot to circle the head tube angle (but maybe that’s irrelevant as some don’t think that is an important parameter atmo).

I know over time I’ll learn to use my machinery and fixtures better, and learn to trust that ruler and my use of it so I don’t have to file to fit (as much as least).  I would like to be able to measure & mark the tube, load it in the mitering fixture on the mill, cut the tube, deburr, and load in fixture and be done with it. Paper tight miters without filing to finish.  No problem, right?

So this is what I’m going to do. No more bent tubes until I get the process for straight tubes dialed better, and when I start selling frames bent tubes will cost more because they do take a lot more time and effort.  Same goes for internal cable routing.  I do love how the frame looks without cable stops but not sure it’s worth the extra weight and effort.  Luckily for my friends I’m a non-profit framebuilder right now and willing to try anything even if it is above my pay grade. More details on this frame in post-to-come!

F17 outside

the finished product