Internal cables in the twin top tubes

In and out

These are pics of my first internal cable routing foray for FULL cable housing…not just the cable inside the frame like i wrote about previously, but the housing too.  This makes more sense to me when trying to minimize water infiltrating and cable longevity. It also makes for a very clean simple looking frame.  Heavier, yes…but like most weight-related issues, there are trade-offs.

I got the inspiration from the forum’sphere (thanks to Zancanato and Wolfhound among others) and wanted to try it out for myself.  But like usual, we all adapt ideas we get to our own situation and bike’s needs so I hope to provide some further insight and feedback to other builders and customers who may want this type of cable routing on their bikes.

The holes are about 3cm long by 1cm wide and on 16mm seatstays — that looks scary taking that much out of the steel tube.  They need to be that long to be able to feed the brass tube through the narrow diameter stay. I made them as small as I could so I could just barely push it through from the back.  I didn’t pre-bend them as there’d be no use in pre-bending with such small diameter frame tubes (i tried it both ways, made no difference in outcome). It’d be a different story if I were using a larger diameter tube such as a regular 31.8 top tube.

Once I threaded the brass internal housing through, I cleaned everything up – wiping with alcohol and a clean rag – and then used brass to braze the twin top tubes to the head tube and also braze the internal housing while there.  This turned out to be the wrong brazing material to use.  Brass needs a higher temp and the thin brass internal housing melts at that temperature apparently.  I saved it by inserting a solid steel rod into the brass housing and brazing without further damage to the housing sleeve.  In hindsight, I’d use Harris 45% silver or Fillet Pro from Cycle Designs instead of brass for this kind of thing.  Silver, even these kinds, doesn’t require such a high brazing temperature as brass/bronze so wouldn’t torch the internal housing (it has a wall diameter of 0.014″!).

After getting a good fillet around the twin top tubes I moved to the seat tube plate, then stopped the torch and switched to Fillet Pro to braze the rear dropouts and the rear exit holes of the brass internal housing.  I had already tacked the entire frame so was brazing in the bike stand flipping around as needed.  There are different ideas on this but seeing some well known peeps tack before brazing the rear dropouts, I have continued to do this as well.  I use a small brush to plop in the flux as best I can inside the slotted stays/tab-style dropouts and use the brazing rod to poke more inside to get the  best coverage I can manage.  Once all fluxed I brazed these stainless sliders with the Fillet Pro (can only use silver on stainless, no brass).  I had done some practice brazing with the Fillet Pro but no stainless dropouts.  Even though it was my first time with this setup, it went really well.  I can see what is happening much better now after a couple of years of brazing.  Fillet Pro still flows a bit more like silver than brass so there’s some initial building up of a fillet inside the slotted stay but once you find it’s correct temperature range it acts a lot like brass.  Next time with large diameter stays I’d want to figure out something to stick in there to ‘block’ the silver from disappearing within, which would save some time, brazing rod…and money!  That’s the biggest issue with silver – the cost is insane.

Once brazed, I de-fluxed in the shop sink and spend a couple hours doing the last bits of filing, reaming/facing, and cleaning.  The one thing I have noticed with internal routing, at least the way I’ve done it, is that the frame has a little bit of a “resonance.”  Not the ‘ting’ you get when you flick a steel frame tube with your fingernail, but if you bump the bike with your palm — you can hear the internal housing tubes resonate (vibrate), but not rattle.  I didn’t think that would happen on this frame with the brass housing having less span between contact points (the internal brass housing hits the wall of the twin top tubes tube more than it would were it larger diameter tubing which you’d think would dull any vibration).  Having the housing inside may help dull the sound, or make it rattle, dunno yet.  Whether it bugs me on the ride…don’t know yet either.  But I’m sure some expandable foam would help dull the sound if I hate it.  I’ll report back on that one.

Lugged Cross fork for Jim’s frame

I used the Nova 7 degree offset fork crown for this fork. I dig this look and it’s still pretty light and stiff (1.9lbs/880 grams with a whopping 11″ of steerer left).  The crown is drilled out to accept a ‘fork crown cable hanger‘ to reduce brake and fork chatter when slamming on the brakes. I brazed in a threaded nut (wtf is that thing called anyhow?) for use of the 5mm bolt on the cable hanger. (BTW, the 7 degree offset means there is 48mm of offset/rake on a fork with 395mm axle to crown length — I had to do some MATH to get that dude!)  It doesn’t tell you that since forks can be made in different lengths but mostly all cross forks are 395.


The dropouts are Paragon tabs and very simple, no lawyer tabs so quick to swap wheels during a race if need. Of course, most people file them off but these come sans tabs.  I used the hacksaw method to get the slots, then fixtured it up to do some last file fitting before brazing. The brass brazing of the legs to the dropouts went nicely, I feel like I finally know what I’m doing now. I did the tacking in the fork fixture to make sure it was all aligned right and finished brazing in the bike stand so I could rotate it around easier. After a soak, the legs got emery-cloth’d smooth around the dropouts and loaded into the fixture with the fork crown now attached to the steerer.  The legs did need some reshaping to slide into the crown easier and I did that with a quick clamp in the vise or two.  I sanded the inside of the crown out too with a small 1/2″ sander roll (from my Dremel kit) and that made it nice and grippy for the brazing and cleaned off that cast finish. Once fluxed, I got to sweating the crown and legs.  Isn’t that what the cool guys call it? I’m pretty lame, so I probably say things wrongly a ton and you all just laugh.  But yes, the silver brazing is coming along. My shorelines weren’t as flooded this time except for one spot – the point. I kept trying to add just enough to get the sharp tip perfect but it ended up pooling in the little crown ‘window’ a little.  My torch heat control is much improved and I’m burning much less flux and not adding as much silver.  (I use a Victor zero tip on my mini torch for silver and a #1 for brass. Sometimes a #2 if it’s a rear dropout braze.)  Anyways, the fork turned out pretty nicely! Should look sweet on the frame all painted up.

Seatstays are a total PITA…

This slideshow requires JavaScript.

…and these are one of the easiest types you can do! More on a later post about my trials and errors on mitering the seatstays, but here’s the semi-finished product. I also added the chainstay photo-process too since I forgot to post that when I did it last weekend.

I dig the way these half-fastback seatstays turned out. Classic look, functional, no bending and still plenty of tire clearance (almost 1cm about on either side of a 35mm tire). I am learning to love fillet brazing more each time I do it (because I’m getting better at it).  I can actually push and pull the brass to my liking, and lay it down where i want it to go in the first place mostly.  Still much room for improvement, but i barely have to file anything away on these and the dropouts.  It did get a bit hot in places so you can see some burnt flux (black spots) as well as some discolored brass meaning it got too hot in a couple spots, but those are details for at this point. It’ll stand the test of time just fine.  Some more brazeons to finish before finishing this frame, and then I still gotta make a unicrown fork to go with it.  I’m set up to miter the unicrown on the mill this time with a big ole ball tipped end mill. Should be a fun weekend.

First try at internal cable routing

This slideshow requires JavaScript.

This was my first attempt at internal cable routing and it went ok.  I pretty much copied Little Fish Cycles’ blog post about this so it’s nothing huge to share. I bought the internal routing ‘kit’ is sold by Nova Cycles and it includes a brass internal tube that is 7mm outside diameter and 500mm in length. It’s just for the cable to pass through – unlike how some other builders  do internal routing by using stainless steel tubes (1/4″ x 0.016″) or 9/32 x 0.014″ brass tubing, which have a large enough inner diameter to pass the entire cable housing through (see this link for an example).  Both are viable options for internal routing, but both are both heavier and way more complicated to fabricate than using external cable stops.  The internal routing I’m using for this cross bike is not that much heavier than normal routing (weight of brass tube isn’t that much) but I am not sure I’ll use this method any longer and instead opt for the internal stainless sleeves so to run full length cable housing.  It looks cleaner (to me) and ensures that no water will ever get in to muck up the cable, whereas with the brass tube there is a possibility that water or dirt can still get in and make braking less smooth.  If it does get gunked up, at least it’s easy to clean with spraying some some WD40 in there and blowing it out with an air compressor!  I didn’t have any stainless tubing lying around and did have the Nova system in house so there you go.  It does look pretty cool when done no matter how much a pain it was to drill, file, fit, and braze in there!  The braze-ons are VERY thin so they’re easy to overheat with the torch.

The main reason not to do internal routing is that external cables are though easier to swap & maintain/clean. But you could run Gore cables for protection in wet conditions, or you can just run plastic sheaths that slide over the cables between cable stops on the top tube.  For shifters, it definitely needs to be friction free so having the ability to at least run one stretch without housing (top tube) would make them easy to clean.  The lower section from the top tube down to the rear derailleur could be enclosed and protected from the elements in housing.  But externally routed would be so easy compared to fishing/bending/brazing an internal tube down the seat-stays or chainstays from the top tube! It’s all possible, but it’s definitely more time and money to do internal cables! I estimate that this method takes 1-2 hours depending on how many cables you’re running internally where as external braze-ons take maybe 10 minutes for all of them.

Back to work

TK’s wishbone stays with the caps brazed on.

A week or two off of shop work. Moved a big truckload of stuff out to the new house in the new town and am absolutely psyched on the new shop…it’s immense…i don’t deserve it!  But I hope to grow into the space by turning ‘Pro’ with this framebuilding gig in 2013.  I think I may have to build a mini-skate ramp in there too as it rains a bunch there in winter.  I am a very lucky human. We’re pretty out there in the woods, but hey, that’s nothing new for this hermit. The new town has such a mellow relaxed vibe to it, I’m going to give it the nickname “Mayberry”. The hardware store just feels like I’m going to run into Gomer Pyle and Andy.

Anyways, still finishing TK’s Tour29er and the wishbone seat stays are finally complete! I have more braze-ons to add (this bike has probably 2 lbs of braze-ons!) for a rear pannier rack, as well as just the rear derailleur cable stops and brake hose cradles, but I think it’ll be done with another few hours in the shop.  It’s funny the things you see when taking pictures close-up.  Always room for improvement even though I’m pretty happy with how these stays turned out.  I’m now 100% sure they take too long to make when you account for all the caps and cleanup of the caps, but I like how they look still.

More pics of T’s Choad…

This slideshow requires JavaScript.

After soaking off the flux in the bathtub overnight, blowing out the water inside and out with the air compressor, using 80grit emery cloth and a wire wheel in a drill to clean up the surface rust that developed overnight on the tubes, and then 2 hours of filing…this is where I’m at.

I checked rear wheel spacing and it looks just like I wanted and drew out — about 1cm on each side of the tire.  The seat stays are being hand filed/mitered tonight and will hopefully be tacked in place by the weekend where I can then put it all together with the small downtube braces/bridges and the final seat stay bridge.  Tripper didn’t want any brazeons and just wants to zip-tie the rear derailleur and brake cable onto the twin downtubes, so it’ll look pretty clean.  Internal cable routing for hydraulic brakes…? I’m sure it’s not too hard but i worry about rattling and ease of maintenance.


…and the cradle will rock

This slideshow requires JavaScript.

Frame 11 is coming together, slowly.  It may look weird right now, but each tube has a purpose and even though this will be one heavy frame relative to a ‘normal’ two-triangle frame, it will be sweet to ride this fatty.  It’s been a slow process but mostly that was because jigging this thing up was so freaking difficult.  That’s part of the fun though, i have to admit – making what you have (brains and tooling) work for whatever situation arises!

Some things people will probably wonder about when seeing the pictures:  The plate or cradle is 0.062″ thick 4130 plate. Retrotec uses 0.04 or 0.05″ depending on the frame.  I went thicker since, well, that’s what I could get, and this will be a high stress area where the chainstays will be trying to twist.  I’ve not seen a bike with a plate for the chainstays, instead for the top tubes/seatstay junction.  The plate is MUCH wider than any Retrotec plate since it needs to fit a 4.5″ tire.  So, since the increased span, and increased torsion that’ll be applied, I chose thicker.  Also, it is cut in two pieces since I wanted to get a little angle to match the angle where the DT and CS tubes meet.  The twin downtubes are 16mm diameter x 0.9mm wall thickness and span from the head tube to the plate where they ‘turn into’ chainstays.  The chainstays are actually 16mm True Temper round S-bend seat stays.  I used a 170mm Paragon head tube, super beefy to hold all those tubes with no problem. That’s a 31.8mm 8/5/7 top tube, going back to a 34.9mm seat tube with a sleeve cut from the left over from the same tube. The reason I did this is because the seat tube is already 1.2mm up top and 0.8mm down below where the ‘legs’ take off to the BB.  Adding another 0.058″ walled tube sleeve to the top of the seat tube seemed like overkill so the extra 0.8mm was what I chose. It’s not a full sleeve, it is spliced so it will fit.  I needed a sleeve since the top tube comes in lower than the tube’s butt, so the sleeve helps me extend the effective butt length in a way.  So…why a 34.9 seat tube? I need to fit a Joplin drop-post on this frame so with a 32.4 – 31.6 USE seat post shim it’ll span the difference between the 32.5 (I.D of the seat tube) and the 31.6 diameter seat post. (Sorry if this is way too much detail, but it’ll help me when I forget what I used in a month…)  The seat tube legs are 16mm x 0.9 wall 4130 steel that are used to increase lateral twist in the bottom bracket area. The BB is 100mm instead of a normal MTB’s 73mm so this was the way T wanted to combat that twist. There will be gussets of some sort at the BB/DT joint as well but it’ll look much different from the seat tube legs.  The downtube/chainstays meet at the cradle plate with an angle.  To help braze this connection, I whittled down two short pieces of 14mm 4130 and used them as inserts to connect the DT/CS tubes.  This also helped fixture all the tubes together.

For process, I first fixtured the plate to the seat tube at the right height and tig-tacked it to the seat tube. For the twin downtubes, I did a half-fastback seat stay miter on them to braze them on the head tube.  It looks totally sweet IMO. I forgot to take a photo so i will next time.  For these I had to braze on caps before brazing them to the HT.  Once all this was done, which took a hell of a long time to miter/fit/fixture…it was time for brazing.  It took me a long time to braze this all at once.  I started at the head tube and did the front of the twin downtubes, then went to the plate where I brass-tacked everything together and let it all cool down. I added even more flux to the plate and started fillet brazing it all together IN the fixture.  Most builders don’t braze in the fixture, and I usually don’t either, but with the Access65 it’s quite easy to do so and it keeps everything in the right place.  Especially with this weird type of frame, I wanted to keep it all jigged up tight while brazing.  Anyways, so far, the only things I’d change are that I’d make an attempt at bending the downtubes and/or chainstays in the vertical plane so that it has smoother and less angular lines.  But functionally, I’m not sure that’d make a bit of difference.

This is a total prototype.  On my snow ride today (awesome, btw) I thought how funny prototypes are.  The thing about them is that you usually ride them a few times and then put them on the shelf and make another because they totally suck or weren’t what you were expecting.  We’ll see how T likes it.  I just hope I can finish it before winter ends! It must be ridden in the snow!