Nothing to lose really.

You could have reduced the thread diameter to 3/4 thread depth safely, to remove some metal. (even more in bad cases).
I always use a die stock, with the machine spindle locked (or lowest gear).

No machine cutting possible?
I tried die cutting 1" x 24tpi a few months ago and that wasn't fun.
Ended up screw cutting it when the tips arrived.

The die was a 'bad thread' dresser really, but I had to have a go. Got it cutting but not well.
We learn every day.

I machine cut the 24tpi threads for the 2WD unit adapters OK on the lathe (just a lot of faffing) but thread cutting is a bit arcane.
I would need to mess with the gears on the lathe for the M12 Fine and my brain is not working for that this week. ;-)

Surviving though, and that's the main purpose in life ATM. 🙃
How many drive dogs is enough?
I am thinking about how to transmit the rotational forces from the axles to the hubs themselves.
Basic premise is that the cap-nut on the thread on the outboard end of the axle will have a plate welded to it and this plate will have drive-dog that interfaces with the 5-spokes of the main hub casting. So, 1 lump or 2 Vicar?
I was thinking of a stub of 12mm bar (or similar) for this purpose.
1 dog and if it breaks you have no drive, 2 dogs, 3, 4, 5?
Bear in mind there is only one axle one nut and one plate.

Answers on a postcard. :)
If you go for more than 1 then you'll need excellent accuracy or only 1 will be driving anyway unless you coat the dogs in rubber or similar.
2 choices then ?

If you can't make it accurately make it adjustable ?
Use these
mini exhaust bobbins

2 choices then ?

If you can't make it accurately make it adjustable ?
Use these
mini exhaust bobbins

I shall endeavour to make it accurately by cheating with the wheel/hub face-down on the bench and use a transfer-punch to pop a mark in the right place, drill that out and then weld a dog in the hole. :) Poor puppy.
Last edited:
My "driven hubs" project for the trike has ground to a halt. :(
In their infinite wisdom the nice folks at Sturmey Archer (curse them) have ensured that there are no perpendicular/right-angled faces against which to trap a drive-dog. All the possible faces I have tried are at sloping angles which will have a drive-dog "climbing" up the slope. I may have to resort to making holes in the hub to drill & tap in order to affix a drive-plate. :(
I had hoped to avoid doing anything that marked the hub as non-standard in any way. Ho-Hum.
I'm going to be in the same boat at some time.
Think I will make new back plates.
But I have a plan to fix this while doing minimal intrusion(s) into the hub-shell.
Watch this space. :) I can see a way that will be aesthetically pleasing as well as minimally invasive.
But I have a plan to fix this while doing minimal intrusion(s) into the hub-shell.
Watch this space. :) I can see a way that will be aesthetically pleasing as well as minimally invasive.
It's not the backplates I am struggling with (@Dale Rider )Martin. :)
The backplates are going to "float" on the spinning axles and accompanying innards of my modified hub-shell/bearings/spacers stack.
A plate on the outer axle tube will provide an attachment point for the anti-turn plate of the backplate itself.
When braking they (the backplates and their shoes) should be naturally self-centring; when not braking i am expecting them to .....errr, well, let's see.

The bit I have been struggling with is how to transmit the drive from the half-shafts to the hub shells and thence the roadwheels.
I couldn't use a single drive-dog as it clunks & clanks and wants to climb up the hub's spoke. Tried that yesterday and dismissed it as "just awful".

So "Plan-B" is to put a drive plate on the face of the Hub-shell with 5 studs pointing inboard and these studs sit in holes drilled into the 5 arms/spokes of the hub-shell's body.
The drive plate has the axles "cap-nut" firmly welded through it's centre. When the axle rotates, the plate has to rotate and the 5-studs drive the hub shell. :)

I am not fully decided on the stud "size" but if a 12mm shaft can be driven by a single 4mm "key" in a keyway then I should imagine 5 off 4mm studs set 6mm into the hub shell spokes would be adequate to the task.

The whole point is to be minimally invasive and have the ability to return the hubs to "standard" at any time.
Presumably you are using a left hand threaded cap nut on the right side of the trike. It'll be self tightening under drive but what about on coasting? How do you stop the wheels from wanting to undo as it coasts?
Presumably you are using a left hand threaded cap nut on the right side of the trike. It'll be self tightening under drive but what about on coasting? How do you stop the wheels from wanting to undo as it coasts?
Yes LH thread on RH side.
On coasting? Hmmmm...
Well, in theory if I were coasting then the wheel would drive the drive-pins and the plate and cap-nut would try to undo themselves but in doing so would also turn the axle which would operate the ratchet of the 2WD unit and just "free-run" so the thing should not be able to undo itself (famous last words).

As we watch the offside-wheel go merrily bowling along the road at a vast rate of knots we can begin to deploy the "hopes & dreams" device for a soft landing and only a short hospital stay. ;)

Of course, if we were of an entirely cynical and untrusting nature we could shove an "R-Clip" or split-pin through the cap-nut and the axle to prevent such an occurrence. 🆒
Here was the 1-Armed "drive-dog" bandit of misfortune (I didn't pursue this because it was clearly going to be "carp".

So I decided to adopt a different approach where a plate with 5 drive pins will engage with 5 small holes drilled into the alloy spokes of the hub's shell. On the RH side the plate and cap-nut have LH threads, on the Left they will have RH threads (for obvious reasons).

Isn't it strange that using calculations and theory give a worse-aligned result than doing stuff by pure "eyeball" alone? If you look closely you can see that dividing with a compass for 5 equidistant points on a circle of known diameter resulted in a success ratio of about 2 out of 5. I can turn the plate over and have another go. Good job I did not drill the holes already.
And here it is from another angle so you can see the axle.

I'd like to say what wonderful fun I am having, but I would be lying. 🙃
Last edited:
R clip would only need to go through some few threads poking out from the nut. It only needs to keep the nut from coming off entirely. Not having to go through the nut too means no alignment issues.
Today was a day of both triumph and despair.
I sorted out the placement of the 5 holes on the drive plate for the pins for the RH wheel. This time I used a calibrated eyeball and it worked much better (IMHO).
Having drilled these out I re-assembled it onto an axle and onto a SA 90mm drum hub and proceeded to ruin bastardise adapt the hub shell.
I drilled the first hole gingerly and was relieved that it seemed to go OK and it popped through the other side of the hub-spoke.
I used another 3.25mm drill-bit as a locator peg to lock the drive plate in place and the drilled the next hole (like doing up a 5-stud mounted wheel; do one, miss-one) in the same fashion.
Once I had 2 holes and the central axle shaft as locators I used an appropriately sized transfer punch as another locator peg.
I swapped the locator peg positions around until all 5 spokes were ruined had holes in them. :D


I was going to do a second wheel but the drive plate would not co-operate and insisted on being threaded "off-square" so it had to go in the scrap bin and I will have to make another. :(
Not today I hasten to add as my quota of shed-time is all used up for the Weekend and I need to go play the part of dutiful husband.
I escaped the guard and snuck off to my workspace.
I have managed to make the other drive-plate and I even got the thread cut OK, but I had to go really old-school to cut the thread.
This little lathe is definitely too small, I may have to look for a bigger, better one. :LOL:

Tomorrow I will mark out the spots for the holes to be made in the drive plate and ruin adapt another SA90 hub shell.

The 5 tiny M4 (4mm) studs that will transfer the torque from the drive plate to the hub shell do look rather small but when I look online I see that an M4 bolt will hold a 150Kg load so maybe it isn't so bad.
After all, the worst that can happen is the wheels fall off. :)
Last edited:
Today's bit of stuff and nonsense was the ruination modification of the other SA Hub shell.

Two things became immediately apparent, and they are not good things.
If the plate or any individual drive pin is not perfectly square to the face of the hub then it will be very difficult impossible to get things to assemble together.

Some 4mm bolts are on their way and I will give it a try out. IF it looks too dodgy then I can always increase the bolt size to M6, but the same limitation will still apply, just on a bigger scale.
Last edited:
.... and in a change to our advertised programme we bring you this.

I decided that maybe (just maybe) having the entire drive reliant on 3 and 3/4 turns of M12-Fine threading might not be a good idea.
So I made a new part that has more thread traction than you can shake a stick at and it will be welded to the drive plate (or vice-versa).

and the side-view is as below. It also provides the proper spacing between the drive-plate and the inner race of the outboard bearing on a standard SA hub (2 birds, one stone (y) ).
Today I took a bit more off the end of the home made LH threaded nut so that it snugs the drive-plate nice and flat on top of the face of the hub and contacts the inner race of the hub's bearing.
I put some M4 bolts in (round the wrong way to how they will be in the finished item) so I could check just how well/badly they would fit in the holes in the hub-shell.
I was right in that any degree of misalignment makes it difficult to engage the drive pins into the hub-shell. But, with a small brass drift and a very small hammer the plate went fully home. :)
It may be VERY DIFFERENT when I try it with the bolts round the right way and their head's cut off.

To ensure the best chance of it all fitting I have index-marked the shell and the drive plate so it can hopefully always be assembled with the best matching of stud's to hole's.


Click for DIY Plans!