Simpler knurling tool? Get it made!

I'm not one to spend hours making fancy shit just for the hell of it. If I can design a part to achieve the required function without the need for complicated setups and machining operations, I'm all in. I'd rather spend a few hours digesting the possible concepts before reaching for the starting materials.

The previous concept looked nice enough if it was intended for sale, which this certainly isn't. I just want something that's fit for purpose. I could fabricate something like this by welding a toolholder body to a doglegged head that contains the knurling wheel subassembly.

Those chamfered faces would allow the wheel to be presented at 45 degrees but the other 2 orientations would be rather useless.

But that would still be pretty painful to make. I've got the means to weld to lengths of stock together like this. It requires the parts to be chamfered and preheated - and a decent / brutal current of 140-150A but ends up pretty agricultural looking and still requires a good bit of machining. 

Here's some 19mm (3/4") hot rolled steel, chamfered up and being preheated.

With the MIG welder, it looks like a functional weld. No prizes for looks but it's not a downright embarrasment.

But I'm not getting the happy feelings looking at this. I'd have to machine it after welding due to the heat distortion, yet this would be an unwieldy and large beast. Time to give it more thought.

This would be simpler. I could simply Loctite the head to the toolholder shaft. To alter the knurl angle, I'd simply rotate the whole tool in the toolpost. I could fit a grubscrew to lock the head to the shaft if necessary. By ensuring the face of the wheel is coplanar with the axis of the shaft, the point of contact of the wheel won't change as the angle is altered, eg for diamond knurling. And there will be minimal torque generated at the joint due to the cutting forces.

Apart from machining the relatively simple head (2 bores at 90 degrees), there's little to it. The toolholder shaft is simply a length of round bar stock turned down at one end and I can use a std boring bar holder to hold it in the toolpost.

Even The Stupid Fat Bloke should be able to manage this. Famous last words and all:

No welding, minimal machining. What could possibly go wrong? I'd rough out the stock in the mill, then machine the 13mm/14mm bores in the lathe, then cross drill the other hole in the milling machine, as this would allow me to use the probe to pick up the bores and then place the cross drilling with the required accuracy - I need the bores to be reasonably coplanar. The cylindrical external surface can simply be created by hand on the belt sander - I have no pride. Sounds like a plan....

Cut chuck of steel from some barstock:

Mill it down to size, including the step feature.

Position the bore exactly in the middle of the stock:

The HK1015 bearing looks as if it wants to be pushed in:

Some manual work on the bandsaw:

Bore the 12mm hole for the shank:

Turn down some 5/8" (19mm) round stock for the toolholder shank:

Ready for assembly:

Bearing pressed in, shank Loctited in. It all seems to go together as planned. The point of contact of the knurling wheel is coaxial with the shank body.

Fits in one of the boring bar holders:

Looks like a decent job, although the only bearing I have to hand is one of the old bearings that was replaced in the Boeni tool. And sure enough, when I "test" it with finger pressure, it wobbles visibly. Presumably that's why it was replaced.

Around 220um at the wheel end:

And ~120um at the other end:

NVM, I've ordered a couple of new INA bearings (£6 each) and will replace the knackered one before trying it out. 

And in the interests of science, I checked the Boeni tool with its new bearings and they are pretty sound - barely 10um of movement if you press hard.

That will do for now. The new bearings should arrive in the next 2 days or so (possibly Monday if I'm lucky), then we can cut some fancy knurls.