Why don't ball bearings conduct electricity?

[peteski]

If BB's were the preferred for performance why not use a wiper on the back of the wheels like Kato Mikes?

Humbly Submitted,
The S.

Kato Mikes do not use wipers for conducting electricity.  The brass bearings still conduct electricity between the axle and the chassis.  What you are probably thinking of is the brass inserts in the back of the initial design on the driver wheels.  Those had a metal hub and tire but the spokes were plastic.  The brass inserts were press-fit onto the hub and they provided the electric connection between the hub and the tire. But those were attached to the wheel  and rotated with the wheel.  The next design eliminated those by making the tire and hub into a single metal part.

[mmagliaro]

I guess you won't even try a conductive lubricant.    And I thought you were an experimenter.   

No... I will try it.  But it's not something that could be a long-term solution, even if it works.
Lubricant will get dirty or work its way out of the bearings, sooner or later, and I will have to keep
applying a special-purpose material (this lubricant) into the bearings to keep the contact working. 
I really don't want to do that.

In answer to some of the other observations:
I did clean out the bearings with alcohol and acetone to see if that improved their behavior.  It
did get the resistance down from near infinity (they were not conducting at all!)  to about 30-40 ohms,
somewhat intermittent as the bearing rotates, but did not solve the problem.

As for how ball bearings work...  the whole idea as I understood it is that the steel balls bear all the
contact between the two races.  The balls cannot "just touch" the two races unless the bearing is
just spinnign free.  Under load, they are bearing
all the weight that is being placed on the bearing, so at every moment, there has to be a ball or two
that are touching with a good amount of pressure on them.

Hmmmm.... that may be part of the secret.  I was just testing these holding them in my hand.  With an axle through them, and a little weight on the axle, they might act quite differently.

[peteski]

No... I will try it.  But it's not something that could be a long-term solution, even if it works.
Lubricant will get dirty or work its way out of the bearings, sooner or later, and I will have to keep
applying a special-purpose material (this lubricant) into the bearings to keep the contact working. 
I really don't want to do that.

In answer to some of the other observations:
I did clean out the bearings with alcohol and acetone to see if that improved their behavior.  It
did get the resistance down from near infinity (they were not conducting at all!)  to about 30-40 ohms,
somewhat intermittent as the bearing rotates, but did not solve the problem.

As for how ball bearings work...  the whole idea as I understood it is that the steel balls bear all the
contact between the two races.  The balls cannot "just touch" the two races unless the bearing is
just spinnign free.  Under load, they are bearing
all the weight that is being placed on the bearing, so at every moment, there has to be a ball or two
that are touching with a good amount of pressure on them.

Hmmmm.... that may be part of the secret.  I was just testing these holding them in my hand.  With an axle through them, and a little weight on the axle, they might act quite differently.

What you said makes sense - the contact points between the balls and the races will have to support all the weight (and thrust if there is any) of the item suspended by the bearing.

As far as lubricants go, Neloube is a dry lubricant which should stay where you applied, and be easily reapplied if and when needed.

[mmagliaro]

I disassembled one of the bearings, and tested every part with an ohmmeter, including the tiny steel
balls.  Everything, including the balls, reads zero ohms, so it is not oxidation or the material that the bearing is made of.  I took another bearing, boiled it in water for a few minutes, then cleaned it in alcohol.
It still tested 20 ohms at best, like the others.  I then dunked it in Neolube, let it dry,
and it tested exactly the same - 20 ohms.

By the way, you know those Bachrus engine-testing rollers?  I have a set, and they work by letting
the engine run on top of a bunch of small bearing races very similar to what I am using.
Their bearings show a similar resistance - about 30 ohms at best.  And yet, those Bachrus thingies work!

[Chris333]

If they did conduct wouldn't you then worry about arching and pitting the smooth bearings.

[peteski]

Well, 20 ohms is low enough to only slightly affect the speed of the engine.  If your engine is consuming 100mA the 20 ohms will lower the voltage by 2V.  If you have 3 bearings in parallel (as in a 0-6-0 engine) then their total resistance is 6.7 ohms.  So 6.7 ohms on both sides of the engine is 13.4 ohms.  That would only drop 1.34V at 100mA.

[u18b]

Hi Max.

I gather from your post that you have not actually done a project.

So here are a couple of my thoughts.
1.  You only need one bearing making contact at a given moment for the motor/decoder to work.  Right?
2.  Wouldn't contact increase when the loco is moving?  It seems to me that if there is a problem, it would be when the loco is stopped more than when moving.
3.  As someone said, wouldn't contact increase when the weight of the loco is on the bearings?

As a comparison, here is a project I never publiushed.
I installed micro-bearings in Bachmann Amfleet cars.   This GREATLY reduced the friction so that my F40PH went from pulling TWO to pulling 8 with room to spare.

Watch the video.  First part is boring.  But at the 4:20 mark, I pull a train with modified cars. 
Notice all the lights are on (except first car, which is a Kato car).

/>
So my point-  I never noticed diminished electrical conductivity.

[mmagliaro]

All:
To be honest, I don't like any of this.   Tolerating fluctuations of 1-2 volts, hoping that more than
1 wheel is touching at any moment so the resistance is lowered, or just living with a 20 ohm
resistance in the wheels.

But I suppose that's the way it has to be.  Either that, or I give up ball bearings, which I really don't want to do.
I was really hoping to see a dramatic improvement in engine smoothness from them, and I really want to
build this engine with them to find out.

Sooooo.....   I guess I will use them and hope for the best.   I will be tugging an all-wheels-live tender behind this thing, so I do have a "safety net".

Chris, as for pitting the metal surfaces in the bearings from arcing... I did wonder about that.   I am just
hoping that given such low voltages and loads, and given that these are stainless steel, they can
stand up to it without any burning or pitting.

[up1950s]

I think a capacitor of unknown value bridging both races might stop the arcing . I'll let Pete chime in on viability and values .

[u18b]

I disassembled one of the bearings, and tested every part with an ohmmeter, including the tiny steel
balls.  Everything, including the balls, reads zero ohms, so it is not oxidation or the material that the bearing is made of.  I took another bearing, boiled it in water for a few minutes, then cleaned it in alcohol.
It still tested 20 ohms at best, like the others.  I then dunked it in Neolube, let it dry,
and it tested exactly the same - 20 ohms.

By the way, you know those Bachrus engine-testing rollers?  I have a set, and they work by letting
the engine run on top of a bunch of small bearing races very similar to what I am using.
Their bearings show a similar resistance - about 30 ohms at best.  And yet, those Bachrus thingies work!

Max,
Forgive me if you already addressed this.

But how are you measuring again?

From inside race to outside race?

Have you tried putting the bearing on a loco shaft and then measuring the outside race (casing) to the loco shaft?

It probably won't be any different.  But I was just wondering.

[nscaler711]

Hi Max.

I gather from your post that you have not actually done a project.

So here are a couple of my thoughts.
1.  You only need one bearing making contact at a given moment for the motor/decoder to work.  Right?
2.  Wouldn't contact increase when the loco is moving?  It seems to me that if there is a problem, it would be when the loco is stopped more than when moving.
3.  As someone said, wouldn't contact increase when the weight of the loco is on the bearings?

As a comparison, here is a project I never publiushed.
I installed micro-bearings in Bachmann Amfleet cars.   This GREATLY reduced the friction so that my F40PH went from pulling TWO to pulling 8 with room to spare.

Watch the video.  First part is boring.  But at the 4:20 mark, I pull a train with modified cars. 
Notice all the lights are on (except first car, which is a Kato car).

/>
So my point-  I never noticed diminished electrical conductivity.

THAT WAS YOU?!?!
Dude, I have to know what you used!! I've got 5 of the newer versions that have been driving me insane....
At least do a write up!! If I have to ill send you a coach to do the write up!!
PLEASE!!

Ok I'll stop...

[peteski]

I think a capacitor of unknown value bridging both races might stop the arcing . I'll let Pete chime in on viability and values .

OK Richie, think about this.  In order to install a capacitor to suppress arcing between the races (one is moving and one is stationary) you would hard wire one of the cap's lead to the stationary side but where would the other lead connect? The other lead would have to be attached to an electrical pickup wiper which would be attached to the stationary side of the mechanism (where the cap is and would rub against the moving part (like the inner race, axle or the wheel) to complete the suppression circuit.  That kind of makes this whole exercise of passing electricity through the bearing  pointless.  Made as well use the electric wiper to just pick up the current from the moving part of the mechanism.

I think that Max is over-thinking this, but until he decides to do some experiments, we can only speculate as to how well will the bearings conduct electricity.  I think it will work (if only judging by the performance of the ball-bearing test stand Max mentioned.

[garethashenden]

In an ideal world ball bearings do not conduct electricity. Some machines are set up with circuits through the bearing such that when the bearing is nearing the end of it's life the circuit is completed leading to some form of notification that the bearings need replacement.

The fundamental problem with any pickup-through-bearing system is that if you are getting electricity then the bearing isn't functioning properly. This applies to the more common practice of using bushings as well. There should be a film of oil between the bushing and the shaft and the shaft should ride on that. The fact that we can pass electricity reliably through this shows that there is metal on metal contact, and that causes wear.

[mmyers]

Needle bearings might have greater contact area.

[peteski]

In an ideal world ball bearings do not conduct electricity.

The fundamental problem with any pickup-through-bearing system is that if you are getting electricity then the bearing isn't functioning properly. This applies to the more common practice of using bushings as well. There should be a film of oil between the bushing and the shaft and the shaft should ride on that. The fact that we can pass electricity reliably through this shows that there is metal on metal contact, and that causes wear.

Hmmm...  That means that all the models which do not use electric pickup wipers rubbing against the metal part of the wheel  have incorrectly designed bearings?  That would also include the axle-point bearing/pickup system too since the end of the axle is the only bearing supporting the wheelset.   We are talking a vast majority of N scale model locomotives here which aren't properly designed! 

While your point makes perfect sense, in the real world of N scale model design it seems like most models do operate in a bearing-failure mode on purpose.

I think that the what you describe does occur in bearings with constant speed (or maybe a narrow range of varying speeds) of the parts supported by bearing .  In those cases the lubricant's viscosity can be chosen to create a film between the moving surfaces.

In model locos the operating speeds range from zero to few hundred rpms (for most of the drive train) and the lubricants used are too thin/light to create a film between the moving surfaces.  The only exception might be the fastest-rotating part of the mechanism (the motor and worm shafts). These bearings at the higher rpms might actually operate in the ideal mode where film forms between the moving metal surfaces.