Commutator repair - AC motor in a Miele Vacuum Cleaner

[Valden]

I saw Jerry's post and video at these links:
 
https://www.eevblog.com/forum/repair/ac-motor-question/msg5080822/#msg5080822



but decided a new post with my particular situation was appropriate.

Jerry, thanks for the video of your DeWalt vacuum motor and its sparking carbon brushes and commutator. How is your repair work going?

I have a Miele domestic vacuum cleaner that recently started doing exactly the same things. Sparking, burning smell, slowing down. It's 29 years old, a beauty and worth repairing, if only to be retired honourably to the garage. For search engines, it's a Miele vacuum cleaner model S448i (S400 series) and is called a White Pearl. It has a Miele 230V motor, not an Ametek motor.

Here's a still shot, sans sparks, at the latest state of repair and after the latest test.



My repair efforts have not yet succeeded and I'm looking for clues. Here's what I've tried.

- the carbon brushes were worn out. No, they didn't last 29 years; I'd replaced them with Miele spares about 15 years ago. Presently, one of them had become too short and due to the spring stop inside the brush holder the spring could no longer push it reliably against the commutator bars. To investigate, I bent the tabs that stopped the spring moving, to reinstate spring pressure for that brush, but the sparking continued. I now believe the heavy sparking from the initial failure event damaged a few of the commutator bars, which has proven difficult to deal with. The reply from Johansen on Jerry's thread supports this.

- Miele no longer supply replacement brushes. I found carbon brushes of the right size on eBay, though they need to be soldered into the old holders. Easy enough. I do wonder about the composition of the carbon and its suitability. I saw somewhere that Miele brushes are 'special'. Really?

- Investigation of the armature and commutator revealed several problems, in stages. The obvious ones were that there was a deep groove in the commutator from being rubbed by carbon brushes for decades, along with the commutator bars being somewhat blackened, though they didn't look too bad. Later I discovered that a few of the bars were sitting lower than the others. Further, the whole circle of 24 bars was discovered to be out of round. Tests for resistance between adjacent bars, opposite bars and between bars and the armature body are all good. The on-motor circuit board and its components (a capacitor, a heat sensor and a triac) are good. The bearings are good. The vacuum cleaner's speed control circuit board is good.

I tried several repairs. First, I put the armature in a drill press and cleaned the commutator with a strip of sand paper (not emery paper). It was very shiny but running the motor for a few seconds with the old brushes that were making good contact resulted in heavy sparking. This step showed that some of the bars were possibly low; they didn't pick up the light-brown coating that most other bars did. While waiting for the new brushes to be delivered I went to work on the commutator. With a drill press I filed down the high zones at each end of the bars, to remove the groove and so that the new brushes would not be pinched. I mounted my drill press on its side to make a crude lathe and used a jig to ensure strips of sand paper would be held steady and would touch only the high spots of the commutator as it rotated. Slowly and gradually I made the commutator and its bars level, round, shiny and on axis. I've added a lathe to my shopping list. Apart from the bars being way lower than original, it looked new. I cleaned the gaps between the bars. I held one of the old carbon brushes against the bars and it ran very smoothly over the surface.

I installed the new brushes into the old holders and partly seated these by spinning the motor in my hands with a battery drill, with help from a beautiful assistant. Everything looked good. My expectation was that the vacuum cleaner would operate, as it had with each repair test, but I was not confident that the sparking would be absent. And so it proved to be. I started the motor for brief runs at the lowest power setting, to bed the brushes gradually. As per earlier tests, after running nicely for a few seconds the heavy sparking would begin and not stop and the motor would slow down.

I'm a bit stumped.

I've thought about how a commutator needs to be perfectly round so that carbon brushes won't bounce or skip over lower bars at high revs, which would result in sparks. I've thought about how brushes get pulled slightly sideways in their holders and how, once they are worn-in, will make contact with adjacent pairs of bars. I've thought about armature and commutator wiring, inter-bar continuity and lowest paths of resistance through windings. I'm concerned that this Miele's commutator is worn down too far (and more so from my repair work) to the point where it might not be capable of functioning properly. Repairable commutators have limits. I'm looking at sourcing a replacement commutator and whether I can install it with the existing armature windings in place (to avoid balance problems), or will need to re-do the whole armature winding (and re-balance?)

The street vendors in India and Indonesia I've been watching online are not as careful as I've been with my repair. Their roughly repaired motors don't spark like my Miele motor does. They might not be showing their failures. Hmmmm.

Has anyone had success with replacing a commutator, with or without re-doing the armature winding? Motor repair guys do this for a living, so the answer must be Yes. This sounds like a fun challenge, but could be a pointless one. I don't want to declare the patient dead just yet; there are sufficient signs of life with each resuscitation attempt.

Your thoughts and clues would be appreciated. Thank you. Valden.

[johansen]

I can only conclude these impossible to repair sparking problems come from an internal shorted turn in the commutator.


Craziest repair I ever did was I ordered the wrong commutator to replace the one in my sliding compound miter saw...
I ordered the commutator for the exact same motor.. in their table saw. the helix on the end of the motor shaft was left handed..

so i cut the gear off the end of the bad motor and cut the gear off the end of the good motor and i bored out the motor to hold the new gear.. mostly on center. it has held up fine.

later i ended up burning out the brake winding, which took out the field coil winding (series ac motor).. so i ended up rewinding that. could have used some more turns, the chop saw runs at least 1000 rpm more than others do...

[Swake]

Still got 2 of these suckers around too, same generation, a red one 'cat & dog' which is full option with speed control and an extra dust filter and a green one with only an on/off button. Both about 30 years old now. They have an issue with the automatic cord retrieval now but as they're used in a static situation I don't care.

Is that picture indeed from after the clean-up? Wow, not in good shape...
It looks like the white part on the right side (is that the carbon brush holder?) is broken.

These motors can be found on aliexpress. Not saying you should buy one, not at all actually.

[Valden]

Hello J. Thanks for this reply.

Given that the sparking keeps happening, I've also been wondering if there's some electrical fault somewhere. However, the motor runs quite perfectly for a few seconds before the sparks start. With the most recent and most extensive repair effort, it ran in a series of five second bursts for a total of about 30 seconds before the burning smell and sparks became evident. Hmmm.

I'm thinking about what you've suggested. I'm imagining that a short in the armature wiring could mean that current is ready jump from a commutator bar that is close to a carbon brush before being in proper and correctly timed contact with it. Is this a known diagnostic fault that points to faulty armature wiring? I'll be looking into this. Thanks.

[Valden]

Hi Swake. Yes, they're good vacuum cleaners. I remember when we bought ours in '94 that Miele proudly advised that they were in service for an average of 14 years in Germany. It's pleasing to get nearly 30 years of good performance.

The still image I posted is after the 'lathe' work I did on the commutator and then a minute or so in total of testing, during which we nearly set the thing on fire. I quickly pulled the motor out of the body to let the heat and smoke out. The commutator looked new one minute and like it is in the photo the next. She's not a happy camper.

If I can't fix the problem with the armature/commutator I'll have another look online, including Aliexpress. My first purchase from there, of a IC to fix a car radio a few weeks ago, has resulted in a flurry of scam text messages and emails. Hmmm. The IC was good though.

[Swake]

You measured the windings and found nothing wrong. Trust yourself or redo the measurement.

On the other hand look at the picture. If that is broken as I think it is, then you have the real issue.

[johansen]

You measured the windings and found nothing wrong. Trust yourself or redo the measurement.

On the other hand look at the picture. If that is broken as I think it is, then you have the real issue.

nah, look at the commutator bar, you have a sort of ok on, followed by a string of burnt ones.

to measure a shorted turn, you need to measure the inductance, the resistance may be too close to measure.

[Doctorandus_P]

The thing has just reached it's end of it's programmed lifetime.
Not worth repairing.

In industrial applications checking and replacing carbon brushes was part of scheduled maintenance, but in home appliances they are hidden under appealing looking plastic and difficult to inspect or replace. Take for example a look at Hikoki (was Hitachi) hand tools. They have easily accessible caps just to inspect and replace brushes when needed.

Another reason it is planned obsolescence is what happens when it stops working. In some appliances, there is no built in safety at all, and when the brushes are worn and the metal spring touches the rotating commutator it is destroyed very quickly. But having a stop on the brushes does not help much. This lowers the pressure on the brushes when they are worn and they start sparking a bit and likely do this without the user noticing, and this also start damaging the commutator. And by the time you notice it the commutator is damaged too much for a cost effective repair. Some motors have hollow brushes with an extra spring and a plastic rod inside. And then, when the brush is worn out, the plastic rod is pushed out against the commutator, and then the spring pushes the brush further away from the commutator and the motor stops without damage. That is a system that works nicely. However, it costs 3 cent to implement, makes motors easy to fix, and thus factories sell less replacement vacuum cleaners. And that is 3 reasons it is almost never implemented in consumer appliances.

If the commutator is not damaged too far, you can turn it down in a lathe, clean the slots, polish it and re-assemble with new brushes, but from the picture's appearance the commutator is too far worn out for this. You can attempt to clean it up with sandpaper (rotate it with a drill or similar while doing so) but the chances for success are not very high.

[Valden]

Hello Swake. And Doctor.

The circle in your photo is showing a moulding junction in the plastic holder for the metal holder for the carbon brush. Settling of some carbon dust is making this junction look like a crack. It’s not broken. Here’s a better pic, taken today. The cause of the problem might be revealing itself over next to the brush on the left side!



I’ll do some more measurements of the commutator bars. The bar to bar resistance was consistently 0.4 Ohm, but there was one pair that was 0.2 Ohm, confirmed by several checks. This is not a big difference in absolute terms, but relatively it could be significant. I’ll check inductance as suggested by Doctor Andus. Indeed I did notice how a couple of adjacent bars did not pick up carbon deposits like the others after cleaning and brief testing. And I wondered if this was because they were sitting low and being skipped by the brushes, or because they were not experiencing the same current flow as the others. I measured opposite pairs' diameters with a micrometer and could not detect that any were low, even though it looked like they were.

The thoughts provided so far are helping. Thank you.

p.s.  This machine is worth fixing and it should be easy and cheap enough to do. It’s just a case of figuring out what the problem is. Succeeding will make me happy. This thread could help others too, so keep your ideas coming. Thanks.

[Valden]

Thanks for the input folks.

In the still pics I've posted and counting from the left, take a look at the cleaner looking commutator bars two and three. They are not darkened like bars four and onwards, but they do show signs of physical contact with the serrated edges of the new carbon brushes. So they're not sitting lower than the other bars, which fits with the careful levelling work I did. Now look at bar four. It shows clear signs of arcing, particularly along its edge that is closest to bar three. Arcing like this could occur if bar three cannot carry current as well as bar four can, or at all. Rotation is from right to left in these pictures and as the leading edge of bar four approaches the left-side brush the closing gap is being jumped by sparks.

I'll have a look to see if the trailing edge of bars on the other side of 'dead' bars are similarly pitted by sparks. We can see that the trailing edge of the carbon brush on the right shows signs of having chunks blown off by sparks. New brushes do not have neatly conforming concave wear patterns; there tends to be a gap at the brush's trailing edge.

It seems there is an anomaly in the armature windings. It could be an open circuit or poor connection involving bars two and three, or a short circuit involving bar four. Either would cause bar four to be a better path than three.

I did not have this problem when I replaced the carbon brushes in the vacuum cleaner the first time, all those years ago. It just worked perfectly.

I think my next step should be to install a new commutator and replace the armature windings, just like the skilled chaps in India do. I noticed yesterday that YouTube is a bit shit at auto-translating Hindi to English, which will add to the fun. 

[Gyro]

Given the degree of damage to that commutator segment (and the following ones), I think shorted turn on the armature is a very high possibility. The segment looks to have become so damaged that re-surfacing in a lathe would be desirable.

One other small possibility. I can't tell from the photo, but it may be that the insulation between those segments (and the damaged one in particular) has become too high and is interfering with brush contact. If so, the insulation needs to be 'undercut' to leave clear gaps between the segments. A snapped hacksaw blade is the standard tool for the job. As I say, a small possibility, but worth checking.

[SeanB]

No, got a shorted turn. Time to order a replacement motor, as that is cheap and will fix the problem. Having replaced a few vacuum cleaner motors, buying the replacement unit, which comes complete with the suction impellor as a complete unit, does work, and they last well. There are not really many different motors used, around 6 common types, including the 2 36VAC motors, that are specified for use in commercial carpet shampoo machines, where the motors have a very short and merry life, and are run on SELV voltage from a transformer, because they do run with a lot of water being passed through them.

[Kleinstein]

Most likely there is also winding damage (a short or open link, but maybe just a small gap that spark at full mains and would not be visible with a low voltage test)  and one would get a complete new motor.
It would only be if the carbon brushes failed (got to short or got jammed with dust) that is would make sense to just rework the comutator.

[Valden]

... re-surfacing in a lathe would be desirable ...

... it may be that the insulation between those segments (and the damaged one in particular) has become too high ...

Hi Chris (Gyro).

Working on this repair has lit a fire under my desire for a lathe. The problem is my old friend Scope Creep might escape from the dungeon where he's locked up and we all know what happens after that! Before the most recent test run I used my drill press lying on its side to 'turn' the commutator. I removed the high areas at either end of the groove worn by the carbon brushes and sanded away a high area on one side. The trick here was to use a jig to hold the sanding strip away from the spinning bars, so that only those that were high would touch it. Just like turning on a lathe! Gradually, the height of the bars became even again. And shiny. Once reassembled, the vacuum motor ran well for a few seconds, before starting to spark like a mad thing.

With each clean-up of the commutator I've checked and cleaned the insulation gaps between bars. Once I have it off I'll check resistance values, without doing any cleaning in the meantime. I'll then clean any gaps that show shorts, to find out how effective this step is. 

 

[Valden]

No, got a shorted turn. Time to order a replacement motor, ...

Hi Sean. Out of curiosity I've hunted for a replacement motor. Sure, there's various motors out there but I've not found a replacement for this one. Please do let me know if you're aware of a source for a Miele vacuum cleaner motor for a model S448i in 230V. It's dual stage with blow through cooling. It uses Miele carbon brushes, part number 5419810, which are unobtainable (though I think I've found a sneaky source in Germany). It's not an Ametek motor, which were commonly installed in Miele vacuums for the US market. Thanks for any clues.

p.s. When you wrote, 'No, got a shorted turn', who or which comment were you addressing? I think you're right about this. If the armature windings were all good, then new and carefully seated brushes on a clean, smooth commutator should work nicely without insane amounts of sparking.

[Valden]

Most likely there is also winding damage (a short or open link, but maybe just a small gap that spark at full mains and would not be visible with a low voltage test) ...

Grüße Kleinstein. Thanks for these thoughts.

The armature tests I've done have all been at low voltage, with a good multimeter. I see your point, that operating the motor under full mains voltage could cause undetected shorts to occur. It sure looks like this is what's happening. I should have made a video of the sparks; Jerry's video above is close enough. The insulation around my Miele motor was starting to catch on fire. Yikes.

I've searched in German on google.de and ebay.de. I've not found a replacement motor. There's Miele repair guys in Germany, but I've not seen any of them rebuilding armatures. I'm sure some do though. I'll have a go at it myself. Videos on YouTube make it look fairly straightforward.
 

[HackedFridgeMagnet]

where the motors have a very short and merry life

lol

[Valden]

Hello all. I have a question below, but first an update.

I have unwound the armature and commutator, paying attention to the winding scheme. Two wires start at opposite bars, then each loops through half of the armature slots, to end up at the other bar. If anyone is doing this I suggest paying very careful attention when unwinding the last two loops, as it is easy to get muddled as to which loop of wire goes to which commutator bar. Once the wires are freed from the slots in the armature, if they're not marked distinctively, it's difficult to tell them apart as you chip away at the lacquer that holds them to the insulation around the shaft under the commutator. So I think I have the right idea on where to start rewinding. If my rebuilt Miele blows instead of sucks, I'll know why.

The question. The diameter of the armature shaft that supports the commutator is 10.5mm. So far my searching online has found commutators with 10 or 11mm inner diameter (ID). I can imagine pressing a 10mm ID commutator onto a 10.5mm shaft, if the material of the commutator's bushing is somewhat pliable. However, the commutator I removed had a brass (or copper?) bushing for its inside surface. It was smooth on the inside and quite tight on the shaft. Very tricky to get off actually. So does anyone have experience with fitting commutators to shafts? Should I keep looking for version with a 10.5mm ID? Or is it the right idea to force a 10mm hole onto a 10.5mm shaft?

Thanks.

[johansen]

Buy a 10mm commutator ans bore it out to around 10.48mm

I have found motors with all the coils wound at the same time, from 24 separate pieces of wire.
, making for a woven basket appearance. They have fiber glass epoxy wrapped around the end turns to hold it all together. 

[Valden]

Buy a 10mm commutator ans bore it out to around 10.48mm

Now this is exactly the kind of excellent justification I need for that high quality German lathe I've been talking to my wife about.

[Valden]

Woo hoo! I've found a manufacturer with a choice of three commutators which look to be perfect. The three variants offer slightly different lengths of hooks on the bars.

The manufacturer is NIDE (Ningbo Haishu Nide International Co., Ltd) in China. The part numbers are DZQ D-RS32-31, DZQ D-RS32-31B and DZQ D-RS32-31C. Dimensions are: 28.2mm outside diameter; 10.5mm inside diameter; 34 (33.5, 32)mm liftable hook diameter; 24 bars; 17mm bar height; 20mm overall height.

Now I need to find a supplier with stock, who doesn't want a minimum order quantity of a thousand. The hunt is on!

Edit: for anyone searching for parts or techniques, there is a fella on YouTube channel Staubsaugeralbay who repairs Miele vacuum cleaners and can supply repair parts.
https://www.youtube.com/@staubsaugeralbay

[Swake]

Now this is exactly the kind of excellent justification I need for that high quality German lathe I've been talking to my wife about.

I knew this was never about the vacuum

[Valden]

It's fixed! Our 29 year old Miele S448i 'White Pearl' vacuum cleaner is back in action.

At first I replaced the carbon brushes, but the sparking and burning smell and unstable motor speed continued. The burning smell was from the sparks trying to and pretty well succeeding with setting the motor compartment insulation on fire.

Next I jury-rigged my drill-press into a red-neck lathe and turned away the commutator's out-of-roundness, while smoothing out some low bars. I also turned away the high zones either side of the carbon brushes' worn down track. I tested the motor again and found that this very nice work made no difference. I assessed that the sparking was resulting from a few bars of the commutator not conducting properly, though I could not detect the reason for this. I felt that the advice from johansen was correct; such problems tend to require rewinding.

I decided to replace the commutator and rewind the armature. Upon pulling the old commutator off the armature shaft I found that the mica material was partly perished. It crumbled into pieces rather than coming off neatly. The original commutator had a copper bushing at its centre, which had fused very effectively with the shaft. Getting it off was not easy. Think drill press, hacksaw and files.

I bought a commutator (and a spare) from China, along with new carbon brushes and new bearings from Australian based suppliers. I found a roll of 0.5mm enamelled copper wire at Jaycar. It was slightly shorter than specified. I've asked them to fix that.

Rebuilding an armature is not rocket science but care does need to be taken to get it right. A few key points are:

- pay very careful attention when unwinding the old wire and make very good notes of the coiling pattern you'll need to follow. Number the armature teeth and slots and mark the commutator bars where the copper windings start and end. Make notes of how the bars and teeth line up, so that the first coils use the right slots.

- the images of the commutators I ordered showed copper bushings in the middle. However, the items delivered did not have copper bushings. The shaft hole is mica. Consequently, the commutator is not as tight on the shaft as I intended. It'll do.

- the original commutator on the armature of a Miele S448i vacuum cleaner motor has a shaft hole to fit the armature shaft's 10.5mm bearing surface. I could not find this size of commutator, other than via a custom production run. So a friend and I had a fun afternoon using his lathe to trim the bearing surface down from 10.52mm to 10.02mm. I ordered a commutator with a 10mm internal diameter. This worked out well. Upon speaking about this problem beforehand with an electric motor rebuilding specialist in Melbourne, their preferred method is to bore out a commutator instead of turning down the shaft. I felt that I could achieve better alignment by turning down the shaft.

- perhaps the trickiest part of the job was working out how far down the shaft the new commutator should be installed, so that the new carbon brushes will ride nicely on the bars and so that the hooks don't collide with anything. A trial assembly before doing the windings was needed to check the fit. Watch out for movement of the commutator along the shaft if you strike either end with a hammer and punch. A copper bushing for the commutator's shaft hole would have helped hold it in place I think.

- do not underestimate the difficulty you will have when it comes to removing the old bearings from the armature shaft. Remove them from both ends before coiling on the new wire, as the new coils will get in the way. Total Tools have a nice looking micro bearing puller for $155. I made my own puller. Getting the new bearings onto the shaft is also fun. I needed various devices for this including a bearing puller, a drill press, a bottle jack and a cute wife.

- I found that I was not winding the loops of wire onto the armature and commutator as tightly as the dude (or robot) who did it the first time 29 years ago. Consequently, the two outermost coils clashed with the lower edge of the carbon brushes' plastic holder. I used a Dremel to remove some superfluous (I hope) plastic from the holder, to create a gap.

The total cost of parts needed was AUD$57, which is about USD$38 or 35Euro. This vacuum cleaner will now be promoted to Veteran status and will be used exclusively with the cars and in the garage, so I'm told. And for spiders. Our shiny new Miele vacuum cleaner bought at the start of this repair project is still too good to be tackling them.

Thanks for your helpful comments and suggestions earlier. Attached is a pic of the rewound armature, showing how my inexpert winding led to larger coils than desired. I'll try to do it tighter next time. I sprayed some lacquer on the windings to help hold them in place, though there's not as much as with the factory job.




 

[CaptDon]

Look up how an 'Armature Growler' works! When I see this type of commutator damage in an automotive generator or starter I instantly know there is a shorted turn (or several). Ohmic measurements are useless unless you are looking for a short to the body or shaft. Inductance measurements are more useful. Old dinosaurs like me who know even older dinosaurs who own a 'growler' are good friends to have! In about 1 minute you will know if you need to trash the armature or spend 10X the cost of a new one in man hours rebuilding the old one assuming it has become unobtainium. A burned segment coupled with out of roundness indicate armature death!! It gave its all and then some!!

[Valden]

Hi Capt.

Well if it died it's been resurrected. The tldr version of my post is - it's fixed and working. 

I saw a growler being demonstrated in a YouTube video. A cool and useful gadget. But I don't have the old friends you do, so getting on with rebuilding the armature was the way forward. It worked out well.

You're right about resistance and continuity tests being of limited value. At least my multimeter was able to confirm that the copper wire was in contact with the 24 hooks on the commutator. And that the armature core was isolated from the commutator and copper coils.

Does anyone still make growlers?