Suggestions:
1. Try to shave insulation (maybe mica) between commutator lamellae, if the insulation is higher than the commutator lamella's copper.
2. If 1. is o.k. then there is a higher current being drawn. This may indicate a possible inter-turn shorts and it may need a dc machine repairer to examine/test it.

Another possibility is that the brush rigging may have shifted. Check that the rigging alignment marks are lined up. If not, then your rigging is probably not set right. Another thing that needs to be determined is where the arcing is coming from. Leading edge arcing is a different beast than trailing edge marking. Also, are you experiencing any brush pattern marking on the commutator? There are many different reasons that cause sparking mixed with a brush wear pattern. This page "http://www.joliet-equipment.com/dc_motors-generators_maintenance_(3).htm" has a good troubleshooting chart at the bottom. Hope this helps

wbd:

If you have changed the brushes and seated them as best as your mounts will allow, you need to look very closely at the commutator. There are three main possibilities:

(1) Damage to the surface of the commutator. Check for pitting and/or mechanical wear.

(2) The commutator is "out of round". What I mean by this is that if you were to mount your generator shaft on a lathe, you could determine that the commutator is not exactly rounded any more, but has worn in a slightly (or sometimes not so slightly) elliptical pattern. When this happens the distances can exceed the tolerances of the mounts and springs enough to cause arcing.

(3) A combination of (1) and (2).

Unless the surface damage to the commutator is extremely slight, or extremely severe, the best solution is to turn the commutator on a lathe. This will remove most surface irregularities and pitting, and will bring the commutator back into "true". This will ensure that the brushes are making consistant, constant contact throughout the rotation. All this assumes that the commutator is of a design that CAN be turned: not all are. If not, the shaft assembly will have to be reworked or replaced.

Hope this helps!

Suggestion/Question to HUGO June 19, 2001: Please, is the web link correct?

jbartos, it seems to work OK; cut-and-paste the whole url.  The end of the link as posted is 'missing':

http://www.joliet-equipment.com/dc_motors-generators_maintenance_(3).htm

[Good article]

Controlman is right that commutator surface irregularities can and will affect commutation (ie. cause sparking) as implied by jbartos in an earlier post with respect to high mica. However, in many cases the cause is much deeper. Please refer to huko's recommended web site as it offers excellent on troubleshooting DC motors.   

P.S.: electricpete, this is a good site to add to your web page.

Update on the arcing:

The commutator was stoned in place to a good clean surface and then undercut.  A dial indicator was placed on it and the runout is only 0.010 (per the website recommended by Huko, max allowable is 0.030).  Brushes seated and unit run with 160A @ 125Vdc and no arcing.

However, when unit comes down for annual PM inspection, I intend to send out exciter for inspection, test and cleaning.  The commutator is worn about 1/4" from new but it is about 80+ years old.

Thanks for all the help :)

We do a lot of on site brush changes. Whenever we change a set of brushes we use sand paper inbetween the comm and the brushes to get as close to 100% contact between the brushes and the comm. We then use a sand stone comm smother to remove the old film that was developed by the previous set of brushes. The film that is developed by the brushes is used by the brushes for lubrication and better conductivity. When a new set of brushes are installed a new film must be developed using the carbon from the new brushes. The old film must be removed.

Once the old film is removed a new film must be developed. Burnishing the comm with hard wood will begin the process. While the motor is running apply a film of hydrogen peroxide to the comm with a piece of canvas rolled up and attachted to a stick. This begins the process and allows better operation of the brushes.

All of the above information is good, but usually when a motor is already running good and all you have done is changed out the brushes the film is a good place to look and how well the brushes are seated.

Well, since it hasn't been mentioned (I throw it in only as an aside)....

Did you use the RIGHT brushes?  NOT the ones that were in there before you changed them, but the ones the manufacturer recommends?

Seen this one too many times to discount it completely.

Don, the guy who still doesn't know what he's doing

Suggestion: What about the generator load? Has it change in any way?