Carbon Brush Wear Alarm Circuit 1

[jktwn]

I am researching methods to retrofit a bunch of dc motors in the plant with indicator lights to let someone know when its time to replace brushes in the motor.

I see where one of the brush manufacturers has come out with an embedded sensor in their brushes that can be used to signal an external device. It would seem to me that one could use the armature voltage supply along with the embedded sensor as the basis for a functional circuit design.

My goal is to elimenate invasive maintenance by doing MCE testing to test for cleanliness along with an indicator showing when it's time to change.

Has anyone been down this road before that might offer suggestions?

Thanks

 

[jtkirb]

I had a hydro customer with an old style, on shaft, DC exciter. He wanted brush alarms, but in that case a limit switch placed on each brush holder was the only answer.

Not exactly easy to do.

JTK

 

[OperaHouse]

I would think a circuit that monitors the amount of electrical noise generated would be good. Unfortunantly, I that would only detect once things were worse than you would like them to be. I worked with an engineer who had a patent on a bearing monitor design. That monitor worked by a ratio of the peak noise to the average noise. In most applications he thought it was worthless because most people just waited till the average noise was enormous. I think this brush design is just an extra insulated conductor inserted in the brush to depth. When it wears down to that point, it conducts. Some other method could take years of study to perfect.

 

[waross]

It is hard to be precise without knowing the physical size of the motor, but try this on;
You could mount limit switches on the end bell frame. They may even be on the outside if space is a problem.
Use a rigid non conductig rod from the brushes to the limit switch. (Drill a small hole in the frame if you have to.)
Possiby in a multi brush holder, you may use a small hook on each brush that is held in place by the brush spring. A "Tee" on the end of the actuating rod could be arranged to be acted on by the shortest brush. The limit switch could be optical, arranged so that the light beam is broken by the target in the event of excess brush wear. That would add a measure of 'Fail Safe" security.
yours

 

[jtkirb]

waross- optical and brush dust is not a good mix.

JTK

 

[stardelta]

Brook Crompton Parkinson have a brushwear detection system available on the MK III range of Bull DC motors. You may be able to adapt it. The link provides details.

http://www.brookcrompton.com/pages/products2.asp?ID=dcMotor

 

[jktwn]

I'm looking at adding BW detection and alarm to dc motors ranging in size from 2HP to 1000HP.

We are currently measuring and recording brush lengths at each PM turnaround and will be generating statistics on wear rates. I would like to install one sensored brush in each motor if possible and worst case one sensored brush in each brush rack.

The statistics along with time between turnarounds will be used to set a depth spec for the probe tip.

 

[waross]

Good point jtkirb
I knew as soon as I posted that I should have said
(Drill a small hole in the frame if you have to and install a diaphram made of anything but silicon to keep the dust out of the switch.)
Seriously, I agree wholeheartedly that brush dust and, in a lot of plants, general airborne dust will be an issue with almost any system.
I was hoping that some-one would build on my suggestion and suggest other ways of monitoring the movement of the rod.
Limit switch, optical, induction, capacitive, alarm at wear point or continous monitoring and transmission of the actual wear of the worst-case brush in each monitored holder. There should be lots of ideas on this still out there.
wxashcra
If you have any interest in fabricating in house rather than buying off the shelf, and are interested in my suggestion I can do a sketch in auto-cad, but I don't know how to send it. Need help here fellas.
stardelta
I followed your link and found the brush monitors, but I couldn't make out the principle of operation. Can you tell us any more about them?
Thanks
yours

 

[UKpete]

I don't think switches mounted outside the motor will be possible, because the commutator covers are normally positioned over the brushgear. Microswitches would have to be mounted on the brusharm but I would fear for their reliability - you might have to test them at least every brush change, adding to the maintenance cost. I had a look at stardeltas link, but unfortunately you can't really tell what system they are using, it's on p.13 of their DC motor catalogue:

http://www.brookcrompton.com/pdf-files/DC_9715E.pdf

Good company thought, they wouldn't put anything sub-standard out.

Maybe the brush with the embedded wire. Personally I would stick with the old system (periodic inspection/replacement) on reliable motors with steady and predictable wear; the cost of a motor flashover will pay for a lot of inspections.

 

[stardelta]

The way the system operates is in effect quite simple, but both the brush holder and brush arm are specialy designed for the application. As the brush wears the arm moves forward in a arc towards the holder. The brush arm is shaped to make contact with a small flexible and insulated terminal fixed to the brush holder when the brush reaches its minimum safe working length. This terminal now becomes live and is at the same potential as the armature circuit. This voltage is fed via suitable cabling back to the control module, this can detect positive or negative voltages between 30 and 600v, and in turn activates a set of contacts which can trigger an alarm or even shut down the motor via the control gear or e/stop circuit. I found a copy of the sales brochure at work which describes the system in greater detail but left it on my desk when I left for home this afternoon! So it will be next week now before I can post any further and more detailed information.
In my own experience I can honestly say the system is not popular with end-users, I have never seen the system installed or used on any DC machine in all the years blahblahblah......, either in the shop or in the field. I share the same opinion as UKPete and would stick to regular periodic checks and inspections during downtimes, its a far more reliable method that can highlight other problems with the machine in addition to just worn brushes. If brush checks are made at regular periods, the lengths are recorded and wear rates monitored over a period of time, an approximate amount of time for the working life of the brush can be calculated. We in fact provide brush inspection services and monitoring programmes for several clients, one of which has their machines covered by an insurance policy intended to meet repair/rewind costs in the event of a failure, the insurance company INSIST that regular checks are made on the brush condition and insulation resistances. No doubt in the event of a claim they will want to see documented evidence of the fact before paying out.

 

[jktwn]

Thanks UKpete & stardelta for the suggestions. After considering your thoughts on the matter and visiting with the brush rep (he too warns about relying on wear sensors even though they sell them), I have decided to drop the alarm circuit project at least for now.

The brush munufacturer makes a brush spring with a bright orange indicator showing remaining length. There is also a company that sells nice motor covers with view windows and quick release fasteners -

http://www.sterlingcovers.com

- that would work in conjuction with the wear indicators on the brush springs.

A maintenance tech or operator could perform a running PM with a flashlight to check brush wear on DC motors on the line. Low tech but should be effective. We would still need to blow out the motors periodically - hopefully our MCE test program will be able to make that a condition-based task.

 

[BJC]

wxashcra.
Do you have a way to measure and record the load on the motor?
I have never done it but one thing I might try is to measure intergrated time and current data.
Brush wear is proportional to some combination of current and time. An hourmeter alone wouldn't be good but some combination of amps or horsepower and time might.
No soultion from me but it's something I would playaround with.

 

[UKpete]

wxashcra, certainly something practical to make the inspection easier (e.g. the indicator on the spring) will make the technician/inspector happier and cleaner if he doesn't have to pull the brush out.

If inspection is possible with the machine running (I would have thought that any windows in the comm covers would get smoked over pretty quickly though) it would have the advantage that you will see if the motor commutation is poor.

Stardelta makes a valid point - with a proper inspection it is possible to spot other problems before it's too late. Wear marks on the sides of the brush (indicating vibration, poor commutator profile), smoked contact faces on the brush (poor commutation), damage to the flexibles, poor fit in the brush box.