John McAleely wrote:Looking at the bind, I've checked the quartering with my gw models jig for each axle, and it seems fine.
Love to know how you did that. I'd have said once the real axle is right through the wheels, the sprung stub axles in the GW jig can't locate in the wheels and therefore the jig can't tell you much about the state of the quartering of a fully assembled wheel set.
I've looked at the crankpins, and with some minor tweaks, satisfied myself they are at 90 degrees to the wheels/axles through their revolutions.
Taking the advice of looking at pairs of axles in turn, I've converted the chassis to an 0-4-0:
That's the way to go
... If I stop the chassis at a free running moment, and then (holding the wheels steady) pull at the rods with my tweezers, I can feel a tiny amount of give. If I let the chassis bind, and then repeat the experiment, at least one of the rods will have no give at all with an attempt at wiggling it with my tweezers. Am I right in thinking that is wrong?
Yup, that's the jam, so you can tell which side it is on and which wheels it is between. It could be the quartering is wrong, but it could also be a difference in the crank pin throw between wheels, crank pins which are not at 90% to the wheel, eccentric crank pin bushes or a difference between the axle centres and the rods. Most likely it's a combination of very small errors in all five. As the rods revolve all the little errors get combine together in different ways. The jam occurs where these add together to the point that the effective distance over the crank pin bushes differs from the distance between the holes in the rods by more than the clearance between the rods and the crank pin bushes.
So the obvious thing to do is increase the size of the holes in the rods and hence the available clearance.
It is in fact very difficult to eradicate all the possible errors, and so you do tend to get binding when the holes in the rods are a perfect sliding fit on the crankpin bushes. The system only works if you provide enough clearance for the cumulative errors, so it isn't cheating to open up the crankpin holes a bit. Just don't do it until your sure that there isn't one big error which is always going to prevent success. When your reasonably sure there are no major errors, ease the crankpin holes in the rods a little bit at a time, with a taper broach or a round file, till the bind disappears. Things should improve as you open up the holes and it should be possible to feel the improvement, it not stop and go look for that major error again. Gold stars to all those whose engineering skills are such that they need very little extra clearance
Finish the job by running it in till its silky smooth.
Second question. The chassis specifies that you need to buy extra 'long' crankpin bushes to the standard kit. This makes sense if each axle needs one. However, that means there is what looks like a lot of sideplay on the two front axles. Is that right? In the pics below, I've mounted one axle as I believe the kit intends, and one with a much shorter bush. Which should I be trying to use? (FWIW the chassis is more likely to bind with these short bushes in use).
You should file down the bushes until they just stand proud of the coupling rods, so the crank pin nut can be screwed down tight onto the bush without gripping the rod. Having them too short or too long can just bring more reasons for the rods to bind.
Will