Intermountain F Unit Tweak

[u18b]

A friend contacted me about some noisy Intermountain  F7s he had.

He had several units that were very VERY quiet, but two were noisy.
He could not quiet them down.

I've looked at them and discovered an adjustment.

Most of you are aware of my Bearing Block Magic tweak. 
http://trainweb.org/passengercars/Bearingblock.pdf

However, that tweak does not work on the InterMountain F units.  In fact, it does not work on any loco with any kind of true universal joint since the joint would not be able to hold the worm shaft rigid without the inner bearing block.

Back to IM.  I looked at the two F7s, and they really were noisy.

I discovered that IM uses a unique drive train.  The worm shaft has a doughnut with spikes on it.  The inside of the flywheel has one too.  And then, they use a cylinder with slots to go between the two.

The bottom line is, on the noisy units I had, the excess noise could be traced to too much distance between the doughnuts so that the cylinder rattled around.

The solution (similar to the bearing block magic technique) was to fine tune the parts to take out almost all lateral rattling.

In the case of this loco,  I carefully moved the doughnut toward the motor (outward on the shaft) until the excess noise was gone.  The trick is to move the doughnut JUST enough to take up ALMOST all of the slack- but leave a tiny tiny amount of wiggle left.  If not, a slight bind will be placed on the drive train and noise will be back.

Here are stock parts.  The cylinder is sitting to the side for clarity.
http://www.trainboard.com/railimages/data/500/DSC04444.JPG

Now, here is the doughnut in its stock placement.  Note that it comes about even with the wormshaft.
http://www.trainboard.com/railimages/data/500/a_IM_F_DSC04462.JPG

And here is the adjusted doughnut- moved outward a bit.  Note the slight hollow on the inside as it has been moved outward on the shaft slightly.
http://www.trainboard.com/railimages/data/500/b_IM_F_DSC04456.JPG

You may have perfectly quiet IM units.  If so, ignore this post.

But if you have one or more that are too noisy, give this tweak a try.

And by the way, these really are gorgeous locomotives.

Ron Bearden

[bryan9]

THANK YOU for taking the time to investigate this and report your findings to us!! Very much appreciated!!!

[RWCJr]

Yes, thank you for the information, and the photos are really workbench useful.

[Bob Horn]

Ron, next time you get over here I think I will tie you down till I learn ALL of your tricks. Very good!!! Bob.

[PAL_Houston]

+1, +1
Good stuff.  Thanks.

[crencs]

Ron,

Thanks for the tip.  It work perfectly on a C&O B unit that I have.  Even my wife would say "whats wrong with that engine?" when it ran.   It was destine to be a show case unit only or eventually going up on eBay.  I can finally enjoy running my ABBA consist!    Thanks again!

Craig

[spookshow]

Great stuff, Ron. Mind if I steal your photos for the encyclopedia? 

Cheers,
-Mark

[peteski]

Excellent detective work Ron and a very simple fix!  Nice!
I hope that these doughnuts are made of plastic stronger than those infamous Bachmann gears used to use. Otherwise with their thin cross-section they could be very prone to splitting.

[u18b]

Thanks for the kind words, guys.

This is a kind of simple fix- which is the best kind.

Yes, Mark, go for it.

I have been thinking and wondering.    IM units are generally very high quality and quiet.
So what makes these few different?  Is it merely a QC issue?

I'm guessing based on what I see that the wonderful folks in the factory would simply place the doughnut on the table and press the worm into it.  The worm would hit the table and stop - thus the worm shaft is even with the doughnut.

So where would variation come in?  It is possible there is fractional variation (micrometers) in how far the worm is pressed onto its shaft.  Same with the flywheels on the motor.   The luck of the draw would be when a flywheel happens to be fractionally closer to the motor and a worm is fractionally away from the motor.  These two nominal differences just barely add up to a space big enough for the cylinder to rattle.

However.....

One other QC issue that has occurred to me is that there is a little nub of flash on the cylinder from casting.  It would be where it was originally attached to a sprue.  You can just see it on the left side of the cylinder in my photo.

Well obviously, the nub can be small or significant.  The bigger it is, the more wobbling force will be placed on the cylinder as it spins.

If this is true, then the solution is obvious too.  An Xacto knife to smooth the area will allow it to spin more true.

By the way- the Rio Grande unit I worked on did not need the nub filed to be quiet.

Peteski,
I have no experience with this.  However, a modeler named John on the Atlas board once talked about how much he hated the doughnut design.  I'm thinking he found something went wrong with time and slipping occured.  Whether this was from shrinkage/expansion (whatever) or from splitting, I don't remember.

Ron

[u18b]

Bob,
I do also hope to see you and Bruce and the guys again.

Ron

[u18b]

Hey Spook,
I just saw your updated encyclopedia page.

Looks good.

But I noticed above the tweak, you complain about frying decoders.

You do mention Digitrax.... but you might want to add more clarification.

The headlight solder pads on the Digitrax decoder are RIGHT up against the front edge.  So that is where the shorting happens. (your solution is tape- good).
However, TCS designed their board with the pads moved BACK.
In fact, they also made the board shorter in the front.
http://www.tcsdcc.com/public_html/Customer_Content/Installation_Pictures/N_Scale/Intermountain/F3A_B/Decoder%20Installed.jpg

Looks like they thought ahead.

Here is the whole install page.
http://www.tcsdcc.com/public_html/Customer_Content/Installation_Pictures/N_Scale/Intermountain/F3A_B/F3A_B.html

Ron