A statement has come up in discussion, "Running a 2-way radio directly from a turbo will take out the bearings in short order".

Has anyone else heard this?
Has it been proven? If so, by whom and how?
Has it been dis-proven? If so, by whom and how?

Is the scenario of a turbo feeding a 32 to 12 volt convertor (railroad grade) and then a radio any different?

Thanks in advance for any consideration, thought or reply............mld

Pete, look at the radio converter to see how many amps it requires. Now look at the generator and see its amperage rating. If the appliance requires more amps than the generator can provide, you have an electrical problem, not a mechanical one.
The only way the bearings can go bad is if you fail to lubricate them or the rotating parts are out of balance and they get beaten to death.

Thanks Chuck,
Perhaps I should have added that for our theoretical discussion that all components are operating within their normal ratings for capacity.............mld

I can only think of a couple of possibilities... fairly distant ones for transistorized modern radio equipment.

If an older radio was powered via genamotor voltage conversion, the relatively high load imposed by transmitter operation might stress the bearings in the relatively low-voltage turbogenerator. One way to mitigate this, as noted, would be to have the supply taken from a comparatively large battery of appropriate cells, and the batteries charged or floated off the turbogenerator.

The other prospective issue would be stray or induced current grounding through the bearings. This would require a certain amount of stupidity, or poor RF antenna design or coupling, but under some circumstances you'd want a good 'third brush' in the generator to be certain the bearings weren't being used as a continually-arcing return path -- we had this happen with exercise treadmills designed by people who really should have known better, and the bearings can degrade dramatically with surprising speed.

.

Just to emphasize what Overmod already said, the electrical field around a two-way radio antenna and leads can be quite impressive. A quick "field" test to tell whether a marine two-way radio was transmitting or not was to tape a 12-volt bulb to the antenna. If the bulb glowed when you transmitted you at least knew that you had output. If that circuit was not properly contained or the equipment was not properly grounded it is not difficult to see a situation where stray current could cause arc damage to bearings.

PC

Not a turbo expert here but wouldn’t you avoid arc grounding through the bearings if you ran isolated negative wires not bonded to the locomotive itself? Wouldn’t you possibly have the same problem with lighting grounded through the locomotive? Incandescent bulbs can draw some amps. With an isolated ground, seems to me that you’d need an isolated ground plane for the antenna.

Just curious....

BINGO... When a mechanical bearing (ball, roller, spherical, bushing, etc...) is intentionally or unintentionally included in an electrical grounding/return circuit the metal surfaces in the bearing may/will "arc over" and cause pitting in the bearing surfaces. This leads to early failure of the bearings.

See here;

https://www.youtube.com/watch?v=KibT-PEMHUU

With the tight tolerances in a "ball bearing" of about a few thousandths of an inch as little as 6 volts of "stray" voltage will easily cause arcing in the bearing. Which results in pitting of the metal surfaces (like a whole bunch of little arc welding strikes) and this is pretty much independent of the hardness of the bearing surfaces.

Normal air (at average humidity and temperature) breaks down and allows an "electric arc" at about 25,000 volts per inch. That converts to about 25 volts at one thousandth of an inch. With ball bearings with "nominal" tolerances of a few thousandths of an inch or less and some dirty oil with an few metal flakes in it as little as 5 or 10 volts across a ball bearing will cause arcing.

The laws of Physics apply from spacecraft to diesel loco's to steam loco's to ox carts...

Electricity, everybody's been studying it for years, but nobody knows how the heck it works.....

This sounds like the same phenomenon seen in three phase induction motors driven with a VFD, Variable Frequency Drive. This is far from my area of expertise, and so I will likely get some of the terminology wrong, but as I understand it, the high frequency the triacs switch at, up to 10 Khz, creates electrical noise that can induce a voltage in the rotor that must go to ground through the bearings. There are several manufacturers that make shaft grounding brushes to remedy this problem, AEGIS makes one that looks like a circular wire brush that bolts to the end bell with the wires rubbing the shaft, while others are carbon brush holders. It sounds like a powerful radio signal can produce the same effect in the generator armature. The solution sounds like a grounding brush or wiper that runs on the generator shaft.

I have another possibility that might have caused bearing failure, if the insulation was failing in field coils and the armature the bearings would provide an alternate path to the brushes. this would not be detectable under normal service but the intermittent loads of a radio transmitting would create the potential to arc thru the bad insulation.

MANY thanks to all for the thoughtful discussion & ideas already shared. If you'll pardon the pun it has sparked a few more questions that I'd like to ask, but hopefully not too many at once.

First off jayrod is wondering the same thing I am - would a full 2 wire (ground isloated from chassis) system eliminate the probability of arcing?

Overmod, you'd mentioned a "3rd brush". Do you mean a way of grounding the shaft similar to what Dennis Storzek mentioned?

Again, all postings are appreciated................mld

That is exactly what it does; the idea being to provide a MUCH more conductive path for current around the bearings than through them. This might be done with a carbon brush bearing on a machined part of the shaft, with the same general maintenance as for stoning commutators.

If anyone is familiar with spark EDM (or discussions of model-railroading track gleaming!) they will be familiar with the mechanisms involved in microarcing. The spark generates a plasma, which at the point of generation can be over 35,000 degrees, well above the melting or vaporization temperature of any physical material. As this heat is absorbed in atmosphere or lubricant the result is rapid expansion away from the arc, the same phenomenon that generates lightning, and the combination of heat and shock can ablate or cause damage much as cavitation in diesel engines or 'sonoluminescence' does.

Note that the pitting is cumulative; if it pits a surface to deformation, the new contact area can pit just as readily. Since in general load transfer from the [pitted' areas to so-far-undamaged metal will likely occur, and electrical make-and-break at appreciable current will be preferentially directed through areas of greater conductivity, this can result in fairly rapid and detrimental wear to rotating-element bearings.

A potential cure might be to substitute ceramic bearings in the turboalternator -- that's essentially a one-time expense if there are no other 'issues' with use of that type of bearing on a locomotive. To my knowledge their wear is comparable and there is less tendency to fretting damage when idle.

The term "grounding" is ambiguous in this context (as it often is)! Whether the bell housing is electrically-connected in any way to the electrical outputs of the turbo is irrelevant to the solution for this sort of problem.

To be clear, and Mr. Ellsworth should confirm, one is tryiing to establish a low-resistance conductive path from the shaft to the bell housing in which the bearing sits, not from the shaft to any of the electrical connections on the turbo.

If I understand the problem, it is one of RF energy near the conductive materials of the bearings. If you electrically short the conductive materials, the RF energy cannot induce a voltage. If there is no voltage, then there cannot be a spark. The solution is to significantly reduce the resistance between the components. The reduced resistance limits the possible induced voltage. The brush Mr. Ellsworth recommends can provide this reduced resistance when carefully installed and maintained.

It may also be prudent to guarantee that there is a reliable low-electrical-resistance path from the bearing housing with the brush and the other bearing housings. This would help to reduce any induced voltages seen at the other bearings in the turbo.

Regards,
Bob Milhaupt

More curiosity on my part.... Can an average railroad radio power output induce a high enough current to actually cause arcing in the bearings? I ain’t smart enough to calculate that. I can see that if your electrical line length resonates with the transmitted frequency that you can induce voltage but maybe not inducing enough current at, say, 50 watts or less to cause arcing. I find this conversation interesting for some reason.

Bob M. wrote:


Thanks, Bob for stating what I meant in this case. I was remiss due to laziness in writing and you laid it out well. Thank You.

Along the same lines I'll paraphrase Paul D's question and part of Kelly's reply. If the power connections are isolated from the chassis of the locomotive (both wires running to and from each load and solid connections on each end) - If the ground plane of the antenna is isolated from the chassis of the locomotive (a separate sheet of metal) - If the insulation of all windings is in good shape:
Is there a leakage path would cause a difference of potential between the rotor and the end bell with enough current to create/sustain arcs long enough to cause damage?

For the sake of discussion please consider the scenario of a steam locomotive with a wooden cab so an isolated ground plane for the antenna is easily attainable, wiring run in conduit, a modern solid-state DC-DC convertor, modern 50 watt VHF radio. The one wild card is whether or not the negative (ground) is carried through the convertor to the load so please consider both possibilities.

This is very interesting and I appreciate everyone sticking with it..........mld

edited for clarity.........mld