Britbike forum Forums Projects started by Britbike forum members Members bike projects No perceptible vertical play

You obviously can’t exert much downward force, so that’s where the ball joint remover/installer would come in handy. Maybe a thrust radial needle roller in the mix. Basically a C clamp on steroids. Hopefully never to be used on ball joints. Sworn off that kind of torture.


E630D569-F056-4BCD-986A-418D330248B8 by First Last, on Flickr

If your mill has any backlash in the mechanism that moves the spindle (which mine does), you could bring it down to assemble the two pieces and keep applying downward force as you gently sledge-hammered directly over the crankpin. Assuming there was enough room, and operator dexterity, to do that without whacking the side of the spindle.

How did you centre that? Indicated on the shafts, or the flywheel rim?

Using that backlash, I did some tapping with a medium sized brass hammer. I’d be worried that anything heavier would cause a bolt of lightning to descend from the heavens. I always have this eerie feeling that my brother ( previous owner) is watching me.

The nuts do appear to be the correct BSA parts. Yes the nut cracking route is an option. We had to do that with some Kubota blade spindle nuts a while back.
Another possibility is that they have been Loctited, although it isn't readily visible. If the rattle gun doesn't undo them, I'll try heat + rattle gun next. The heat gun should be sufficient to bring it up to Loctite release temperature.

I doubt I'll have time to get to it before the weekend, so any advice before then is most welcome.

That’s where the farting around came in. Some of it is a little fuzzy at this point, but I do have some photos stored on another device. IIRC, most was done using an indicator mounted in the quill and run around the circumference of the flywheel. Previously the drive side was held in the lathe with a collet and indicated to make sure everything was more or less true. While in the lathe, you can also check the centre (or make a new one) on the flywheel end of the mainshaft. If all is good, you can use the centre as an additional way to check if the flywheel is centred. I’ve since purchased a coaxial indicator which might help, but would have to confirm that.
The Japanese chuck has a mounting plate that doesn’t jive with anything I have. The plate is removed and the chuck sits squarely on the bed.

I have three air impacts, none of which are pulled out anymore. A 110V AC Brauer (Harbor Freight) 1/2" has spoiled me. Haven't found the bolt/nut it wouldn't handle yet.

I was trying to remove a stubborn one. Tried my air impacts along with the neighbours and no joy. Bought a 1/2” drive electric one and it did the job. Electric better? Of just because it has less miles on it? Who knows… it worked

No standard for measuring them…. air supply makes a difference. and seldom set to meet the requirements of the gun, yet the guy installing the wheels on the car swears by the accuracy of those torque sticks. Whatdayamean you want me to use a torque wrench.

Thanks for the suggestion. A quick search seems to indicate that my 2135QTiMax is roughly equivalent to the 1" electric impact wrenches available locally,, which is a bit of a bummer.

The 1 3/16" AF impact socket* arrived yesterday, and I managed to do battle with the nuts today.
The rattle gun did loosen them, turned to the highest torque setting.+

There wasn't any sign of Loctite in the threads when I checked earlier, but there were certainly signs of copious application once the nuts started to undo.
There appears to be a little flattening of the outer threads of the nuts, so replacements seem to be in order.

[*] The crankpin and nuts are BSCY, but the 12-point 1 3/16" AF socket is a very nice fit. A 31mm 6-point socket would be a little looser, but being 6-point would have quite good contact patches. British Standard ("Whitworth") impact sockets make hens' teeth and rocking horse droppings seem positively commonplace.

[+] Now that I know they were Loctited, heat would certainly have helped break the bond.


Update. B/M crankpin nuts (31-0298) seem to be made of unobtainium at the moment as well. Luckily, closer inspection and a quick cleanup with a 3/4" 20tpi tap shows they're in fairly good shape.

Fixing the tools to work on the bike.

I bought a very used Alpha Tru crankshaft truing jig ages ago, which unfortunately had a couple of chunks out of the knife-edged rollers.

After a couple of false starts, I machined a mandrel which would take the rollers as somewhere between a tight sliding fit and light press fit, and warmed the rollers with a heart gun to help them slide on. The roller furthest from the chuck overhangs the flank of the mandrel by 0.030" to allow the nut and washer to apply clamping force to hold the rollers in place.

It doesn't really look terribly deep, but it took 0.245" off the diameter to clean the nicks out entirely.

Here is one of the rollers with its ball bearing pressed back in, sitting on the stepped bolt prior to reassembling the jig.

Hi All
The BSA crankpin is drilled through to the bearing surfaces
One side of the crankpin is keyed to the flywheel to keep the oilways lined up
If it was me? I would split the crank to clean out the oilways and to see what is causing the misalignment

John

Schadenfreude

I get 0.16mm runout on the timing side and 0.12mm on the drive side, so I'm hoping it was just not aligned properly when it was assembled.

Time for the next dumb question.

As can be seen by an earlier image, the drive side roller bearing is at best a tight sliding fit in the crankcase. Short of Magenetoman's approach of plating the outer race, what are the options for ensuring an interference fit?
The bearing is 25x62x17mm (NF305W or NJ305).

One option might be to use a 1652 (1 1/8" x 2 1/2" x 5/8") or 1654 (1 1/4" x 2 1/2" x 5/8") with a sleeve for the mainshaft, but that involves taking the housing out by approx .057" for 2.4409" to 2.498".

Readings from the housing and the new bearing:
housing, half way down - 2.4404, 2.4409, 2.4409
housing, bottomed - 2.4390, 2.4382, 2.4370

So it's slightly oval as well as belled out.
The poor old thing had led a hard life.

Bearing outer race - 2.4404, 2.4405
which is slightly undersize.

Fortunately, Chris Johnson was facing he same problem with his B44, which also has a 62mm o.d.bearing.

The Speedy-Sleeve he used is an SKF 99243

The primary chain case flange cleaned up eventually, with about 0.065" taken off

The magneto holding down strap may have loosened off at some stage.

Alas, we managed to get to 0.05mm runout with the nuts tightened to 30 ft.lb., but after tightening to 140, even with a packing piece between the flywheels, it's gone back to 0.12mm on the drive side and 0.16mm on the gear side. It may be a key sitting slightly proud, like MM's Ariel.

It looks like cyborg faffing about on the mill or an MM press jig are written in the stars.


This was all going to be so straightfoward...

How is it out if you haven't already described it? When you make your marks on the flywheels do they show the counterweight side too close together? A wedge usually will work on that type of problem. A twist usually responds to skilled hammer blows. These are two common and relatively easy problems to cure. The wedge between the counterweights can take a disconcerting amount of force from a hammer to get anything to move.

I've found the keyed BSA crankpin arrangement to be sensitive to condition or presence of the key. I once tried to true a M21 crank without the key because i supposed it wasn't needed since BSA deleted it in later models. It wouldn't true up even with much effort. So I disassembled it, added the key and began all over again. It then cooperated fairly well.

IThe flywheels are "pinched" opposite the crankpin, so the shafts read high when the crankpin is at BDC.
There doesn't seem to be any twist, which is one reason I'd prefer not to dismantle the assembly.

We opened up the gap in the press using some aluminium plate to protect the flywheel cheeks, and a wedge between them. he crankpin nuts were tightened to 30 ft.lb. at that stage.
The rough check used a digital caliper to measure the outside faces, which gave a pretty fair indication of what would show up on the truing stand. Before starting, it was 3.002" at the crankpin, and 2.978" opposite. we got it to 2.992", and 0.05mm (0.002") runout.

My working hypothesis is that the key is throwing a spanner in the works when the crankpin nuts are tightened further.
It was within spec after tightening to 60 ft.lb, and had gone out a bit at 100. At 140, it was right back to the same figures as before doing battle.


I've started machining the 8" dia x 1" steel for the MM press jig, but that can wait if there's an easier approach.

That jig might not help with your current problem. The jig makes it a lot faster to get the two halves aligned and gently pressed together. But, after you remove it from the jig you have to count on doing final tweaks to the alignment as you successively tighten, check, and tighten some more. Your crankshaft goes off the rails as you tighten, which points to a different problem.

Stated differently, each half of the crankshaft is oblivious of the location of the other half. If you had a set of flywheels with "perfect" bores and also a perfect crankpin, but they were initially misaligned such the.runout was even 1.00 mm, pressing them together would leave the runout still 1.00 mm. Once the crankpin and flywheel tapers are nicely engaged there isn't any incentive for them to become misaligned. Unless there is some other factor involved, such as a protruding key that is encountered as the halves were pressed together.

I would be going after that last 2 or 3 thou with the crankpin nuts almost fully tight knowing that any movement from truing will seat the tapers more, effectively loosening the nuts, and that any tightening of the nuts will tend to mess with your truing. The desired goal is to get the nuts fully tight and the shafts true with no need for that one last whack to bring it in.

Here's another way to spread a pinched together crank that's less violent and allows you to measure as you go. Make yourself a custom machinist's jack like these to force the flywheels wider apart. These ones are 3/4" thread, so they're fairly large but super easy to make on a lathe. Expect to go wider than needed when truing because the crank will spring back a lot.


Good advice.
If all this effort doesn't produce any success.......... it might be time to fully disassemble the crank for inspection.

Possible problems that influence truing?
Cracked key slot in flywheel
Worn or stretched tapers
Loctited tapers (Butchery)
Debris packed between pin flange and flywheel face
Beat up flywheel face
Poorly fitting key
Missing key

The press and wedge idea works fine as well.
As mentioned before, at this stage I would be working with the nuts nearly fully tight and aiming to get the flywheel width within a thou or two all the way around. It looks like you got it down from .024" pinched to .010" pinched.