Questions about worms and worm gears

[Jim Starbuck]

Looking for some insight from the collective here.
In my adventures with home brew mechanisms and turnback drives I’ve encountered something that is puzzling me.
The 90 ton Whitcomb has turnback with a single shaft and four mod 2 worms each directly driving an axle gear. In one direction it runs perfectly smooth and almost silent while turning the opposite direction the worm wants to climb out of the gear. The axles are all costrained with clearance but very little play. I’m trying to understand why the difference when turning the opposite way. To me the tapers in the teeth would want to push away from each other the same in either direction. What am I missing?

[Chris333]

The lay shaft could be shifting. I usually try to mount the worms to the 2 outer piece of frame with just a piece of paper for clearance. This way it won't rock back and forth with direction change.

Just guessing 

[Chris333]

This is what I'm trying to say. There will end up being another worm in the center here, but. See how the 2 outer worms are right against the frame It cannot shift more than about ~.003"

[Jim Starbuck]

Thanks Chris.
What I’ve got is very similar. My outboard worms new to be inward some so I have a slim sleeve over the shaft that runs  against the ball bearings pressed into the printed chassis to hold end play. Just by feel I’d guess there’s.003 to .005.

[garethashenden]

A lot of model manufacturers play fast and loose with gear design, so you may have inadvertently inherited their sloppiness. Are the axle gears you're using straight cut or helical? For good gear design they should be helical to match the worm, something that is rarely seen in N scale as it costs more to tool it properly.

Does it run well without the axles?

[Chris333]

Also there should be a .004" gap in the mesh of the gears. It is hard to measure when so small. When doing the CAD work you can measure the dia of  both gears and measure them together. Then add .004" when you do the shaft hole spacing. But yeah you can just eyeball this too     Hold one gear still and rock the other back and forth to see if it's tight or loose or just right.

[Jim Starbuck]

I think I got it cured.
There were a few things going on at once. I found I didn’t need the spacer tube on one end. I was actually running it too tight end to end. Also when I started thinking end play (thanks to Chris) I also looked at the motor. It’s a double shaft and I made a flywheel for the inbound end. When the motor changed direction and the slack ran out the opposite direction the flywheel would touch one of the wires. I don’t think it was making electrical contact but it did make friction.
As far as gear mesh I was sneaking up on it. This is a version of the printed chassis.

[ Guests cannot view attachments ]

The idea was to have the brass axle bushings bear at the top of the circular slots and not bottom the axle gear into the worm. I would run it, watch, then take a couple swipes with a round file to close up the gear mesh where noted. It’s good to know it takes .004”.
The axle gears are helical and the parts are reasonably precise.
This is one of those things that you can design by the numbers but it just takes a feel when it’s just right.
Thank you for the input gentlemen.

[Chris333]

Looks like this link doesn't work, but it did let you know the shaft to shaft centerline of 2 gears and it let you type in the gap you wanted.

https://www.ultrascale.uk/node/15