Fixing a tiny hydraulic oil pump.

Flynth

Registered
Registered
Joined
Dec 27, 2017
Messages
310
Last weekend I did a repair that was quite unusual (at least for me). One doesn't see much stuff like this online so I decided to post it here thinking members may find it interesting.

There is this tiny computer controlled (4cmx12cm or roughly 1.5'' by 5'') hydraulic pump that pumps oil to operate a clutch that engages an all wheel drive on my car (it is a vw golf 2005, the awd unit is called haldex). That little device is seriously under engineered considering many (including very expensive) cars use it. The electric motor in it fails after some use and replacements are rather expensive (for an old car). So I was hoping my hobby machine shop will help me in fixing it and I wasn't disappointed.

This is the pump assembly:
Compress_20230131_082605_5304.jpg
This is the actual pump mechanism. The shaft is 4mm or 0.16in only.
Compress_20230131_082605_5494.jpg
And this is the offending electric motor:
Compress_20230131_082605_5723.jpg
12V, 35*60mm motor. (about 1.25 by a bit over 2 in).
And that's how it looked inside:
Compress_20230131_082605_5842.jpg
Oh my, carbon crud everywhere, brushes gone and after making a custom puller for the top bearing and cleaning this is how worn the commutator was:
Compress_20230131_082605_5609.jpg
It is not that visible on the photo, but the copper worn completely away on few sections showing the isolator underneath. Where it remained it was tinfoil thickness.

At this stage any "normal person" would throw the whole lot into a junk bin and buy new, or maybe order a new commutator if possible. But as it was a weekend and I thought I can fix it I gave it a go.

First I made a socket to support the top of the shaft while turning and I turned down thd remaining copper on the top and I notched the bottom with a file:

Compress_20230131_082606_6243.jpg
After cleaning it looked like this:
Compress_20230131_082606_6320.jpg
Then I made a brass ring (I had no copper round, the change in conductivity is negligible). I glued it with two part high temp epoxy and I soldered it to the bits of the previous commutator. I then turned it down to size on my mini lathe:
Compress_20230131_082606_6497.jpg
Then it was sanded and polished in preparation for slitting in this setup:
Compress_20230131_082606_6645.jpg
This was on my Chinese mini mill using a 25mm x 0.3mm 80 tooth slitting saw. That's 1in diameter, 11 thou thickness in English units.

After additional sanding and polishing to remove burrs and cleaning it looked like this:
Compress_20230131_082606_6884.jpg
Now I only had to make new brushes using a small file and spare brushes I had previously bought for my angle grinders and after some bench testing the little motor was reassembled in the pump and the pump was reinstalled into the car and it works fine :)

So I saved a couple hundred $ and spent 3 days in the shop (and under the car) :)
 
Very good work.

Have you considered finding some additional used pumps and rebuilding them at your convenience? That way, you could have a spare ready to go in case of failure or you could swap them out and do preventative maintenance before failure.
 
That motor was a mess! I've never seen on that bad. Any clue as to what was the root cause of the failure? With all the crud inside, it looks like there was a leak of hydraulic fluid into the motor. There also appears to have been a lot of heating, probably from arcing due to failed brushes. I expect that the motor used bronze bearings. Did you replace them as well?
 
Very good work.

Have you considered finding some additional used pumps and rebuilding them at your convenience? That way, you could have a spare ready to go in case of failure or you could swap them out and do preventative maintenance before failure.

Thank you :) It is a good idea, but I'm hoping my repair will last. Still, I'll keep looking for used units.

That's beautiful work! Not many of us would attempt it.
Thanks :)
@Flynth ,

Fantastic work!

I can't say enough positive about your post!
...great fix
...smart work-around (using brass when no copper in stock)
...beautiful photography
...excellent write-up

Thanks for sharing this!
Brian
:)

That motor was a mess! I've never seen on that bad. Any clue as to what was the root cause of the failure? With all the crud inside, it looks like there was a leak of hydraulic fluid into the motor. There also appears to have been a lot of heating, probably from arcing due to failed brushes. I expect that the motor used bronze bearings. Did you replace them as well?

I can only guess that the root cause was just that brushes and the commutator were consumed during use. I can't imagine one pair of brushes would grind that commutator down that far so they must have been exchanged more than once before by previous owners.

The bearings are roller bearings with R-1340hh stamped on them. These are standard metric bearings. They seemed fine (no noise at speed, rotating smoothly etc) so I didn't exchange them. I would prefer to replace the bearings and seals at the same time, but in winter this car is my main means of transport so I had to "fix" it with what I had on hand. Regarding heat build up, I'm not sure. This motor looks to be at most 15W~20W. I'm not sure such weak motor could build up lots of heat due to friction, but sparking is definitely a possibility.

All of that black crud was carbon material from the brushes. Amazingly the motor wasn't 100% dead when I started my repair. It would still turn, but the torque was very jittery (one could feel it trying to stop it by hand, visually it looked good). I have to assume the last small bit of original brushes rode slightly lower making two 0.8mm(30 thou) wide grooves below the huge gash in the commutator. Those two grooves can be seen here:
Compress_20230201_103031_1137.jpg
It is as if the whole thing moved up a bit and continued to work somewhat.

As for the idea there was oil ingress, this is interesting. When I drained the oil from the device it was black and full of what looked like carbon. I initially thought maybe someone added graphite to improve lubricity, but who would do something like that to a wet clutch? One usually does not want to increase slip of such clutch. Then I found bigger chunks of graphite (maybe 0.2 mm - 10 thou) in size which looked like very strange.

This is so strangee to me, because there is a wide seal on the motor shaft that appears to be in good condition. Although I could accept dust and oil could make it past the seal I can't imagine chunks the size of grain of sand could make it too. Also if oil slowly migrated in (and out!) I would expect the motor to be full of oil, but it wasn't. It was moist, but I did spray some wd40 on the bearings and screws so it might be WD40 not oil. So I'm not 100% sure, but I think it was just normal wear and tear and maybe someone did add some powder additive that wasn't that well ground to that oil.

The car has 190k km (117k miles) on the clock and it is 18 years old. Also, I asked on few car forums, but no one knows how often this motor runs for sure.

The car manual says "in normal driving the power is divided 90% front, 10% rear, when slip of front wheels greater than a quarter rotation is detected more power is transferred to the rear". This makes me think it has to run reasonably frequently even during normal motorway driving so it might have just lived its design life to the end.
 
I wondered if the commutator grooves were worn in by the springs that pushed the original brushes.
Perhaps the entire brushes were "consumed" in those 18 years and all that debris was the brush remnants.

Brian
 
I am having trouble understanding the design concept involved with your clutch system. Many years ago, I had an Isuzu pickup, the predecessor to the Chevy S10, that had a dry clutch that was activated by a hydraulic system similar to an automotive brake system rather than a mechanical linkage but no pump was involved there.

My experience with wet clutches was with 1960's era automatic transmissions where the clutch plates ran in a transmission oil bath. Operation was achieved by means of a piston working against springs and powered by pressurized hydraulic fluid provided by an internal pump and control valve system. I am having difficulty understanding how your little pump works. Is it simply circulating the fluid internal to the clutch or is it actually creating the pressure to operate the piston to engage the clutch? It seems rather small for the latter.

German engineering has a reputation for being top shelf but in this case, it seems to be lacking. Come to think of it, my wife used to have an Audi and I recall of number of experiences where I questioned the wisdom of their engineering.
 
that is a very impressive fix, well done! When you slit the new brass commutator did you slit it all the way through into the old insulation?
 
Back
Top