Good morning. I have a question, I am currently busy with a project. Now the project is based on the maintenance and inspection of carbon brushes on a DC machine, I ma trying to find a way so that we can move from manually monitoring of the brushes and brush holders of the machine, any suggestion of what to look into when designing the project.

Thank you family.

Mod: Created a new thread

 

Good morning. I have a question, I am currently busy with a project. Now the project is based on the maintenance and inspection of carbon brushes on a DC machine, I ma trying to find a way so that we can move from manually monitoring of the brushes and brush holders of the machine, any suggestion of what to look into when designing the project.

Thank you family.

Mod: Created a new thread

Hi,

If the motor stops turning, the brushes are shot.

You can get an idea sooner if the RPM's start to decrease under load. You'd have to measure when they were new and measure as they age so you know what to look for.

 

It seems that the intent is to learn the condition of the brushes prior to failure, and detect the start of being "worn out". Certainly the brush noise voltage versus load will change a bit, but seeing that will require quite a bit of instrumentation. Very possible but not simple.

 

We typically replace on early failure indications on critical motors with encoder feedback by looking for lag-band errors (position tracking error actual vs calculated) above some threshold and electrical energy effort calculated vs actual tracking. That requires systems with some sort of data-logging and monitor capabilities, it's not something that can be added quickly on old systems.

 

A calendar is often the best. If you’ve been inspecting them on a pm program, adjust the inspection period accordingly, dictated by the operating environment and observed consumption. Lack of visual inspection will eventually catch up

 

The data plot shown in post #4 is an excellent diagnostic tool. But, as stated, " it's not something that can be added quickly on old systems."
Detecting a change in the electrical noise from the brushes would be simpler. That is based on my guess that the brushes are commutator brushes, not slip ring brushes. Brush noise amplitude versus motor load current should show a change as the brushes wear out. That will require some added instrumentation but no mechanical additions or revisions.
What size of motor is being studied? and what sort of operation? If this is a servo-motor application that is different, and also if it is a reversing application.
OR, is this a student research project??

 

I concur with @GetDeviceInfo , a regular PM program is usually the best way and once regularly observed eliminates any problem.

 

It seems that the intent is to learn the condition of the brushes prior to failure, and detect the start of being "worn out". Certainly the brush noise voltage versus load will change a bit, but seeing that will require quite a bit of instrumentation. Very possible but not simple.

Hi,

Supposedly the resistance presented by the brushes increases as they wear down. That could be partly because of the way the spring has to extend out farther and farther. I do not know offhand what it takes to measure the rpms with load though, but I believe that is one way to determine the state of the brushes.

 

Good morning.

I am still confused; using sensors to monitor brush wear condition.

What is the best way of monitoring a wear condition of a system that consist of 12 brushes of an ac system, 84 brushes of a DC system, the DC system is driven by the AC system.

 

Are all of the brushes on a single machine?? Is the system new? or has it been in use for a while? Does anybody have experience with "the system"?
In some instances, where the factors affecting wear are consistent, brush inspection is based on hours of operation.
The scheme presented in post #4 seems to be applied to some type of servo system, where the motor portion is given velocity commands and an encoder measures speed. In a servo system with an encoder and computer control that could be added at a reasonable cost because it would all be in software.
In a non-servo system a different scheme would be required.
One scheme would be a probe contacting the rear end of each brush so that as the brush wears down the probe would lose contact with it. For a physically large device that would be reasonably possible, for a smaller motor it would be difficult.

How big are the systems with the brushes, and what voltage range is present on those brushes??

 

I am still confused; using sensors to monitor brush wear condition.
What is the best way of monitoring a wear condition of a system that consist of 12 brushes of an ac system, 84 brushes of a DC system, the DC system is driven by the AC system.

What type of system is this? If brushes used in an AC system, is this a wound rotor induction motor etc? what form does it take?
If the DC devices are driven, are these generators?