Another alternator hits the dust and not under warranty

Discussion in 'Parallel Universe' started by C5!, Jul 11, 2011.

  1. C5!

    C5! Been here awhile

    Joined:
    Sep 23, 2007
    Oddometer:
    278
    Location:
    Venezuela & Colombia
    diagnostic of the shop: one phase looked more damaged than the other 2. He blames the regulator. but is giving me a 50% discount on this rebuilt, and will be ready tomorrow.
    #61
  2. GB

    GB . Administrator

    Joined:
    Aug 16, 2002
    Oddometer:
    61,597
    Just out of curiosity, when I got home from my work commute today, I took out my infra red thermometer and I got 47°C on the clutch cover and 85°C on the stator side.. The catalytic converter read around 75°C.
    #62
  3. flying.moto

    flying.moto Earthbound Misfit, I

    Joined:
    Nov 22, 2007
    Oddometer:
    375
    Location:
    South of Gorman
    Wow. When i built my welder about 13 years ago, the general rule was any temp of a tranformer winding over 80C is unacceptable... No wonder these are frying...
    #63
  4. C5!

    C5! Been here awhile

    Joined:
    Sep 23, 2007
    Oddometer:
    278
    Location:
    Venezuela & Colombia
    for those in the know, that is the cable he had been using. Would that be right?

    Attached Files:

    #64
  5. bmwgsrider

    bmwgsrider Been here awhile

    Joined:
    Jun 29, 2006
    Oddometer:
    307
    Location:
    Amarillo
    I have a 2011 that just turned over 10,800... is this situation also known on 2011's or do you think BMW done something to resolve the issue?
    #65
  6. Gangplank

    Gangplank Advenchaintourer

    Joined:
    Jul 20, 2003
    Oddometer:
    2,316
    Location:
    Las Vegas
    How many of these have gone? Any common thread? A few? A lot? 5?
    #66
  7. LukasM

    LukasM Long timer

    Joined:
    Nov 15, 2006
    Oddometer:
    5,461
    Location:
    On a RTW ride - currently touring the U.S.
    I would bet money that BMW hasn't changed anything, even on the 2012 models. I did hear that they are planning to at least make the stator available by itself, and not only packaged with the flywheel which triples the price and is unnecessary in most cases.


    Pretty much all of them on bikes with 50k+, often on bikes with around 30k, occasionally with significantly less.

    It's definitely a wear part in this engine, and one of the first things to go.
    #67
  8. Gangplank

    Gangplank Advenchaintourer

    Joined:
    Jul 20, 2003
    Oddometer:
    2,316
    Location:
    Las Vegas
    So when I do the 30k maintenance... Add: Stator rewind.

    Linda sounds like my 4Runner. At 90k replace timing belt & water pump. It's not needed to change the water pump but while you are there you do it because they all go before 120k.
    #68
  9. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    In my experience, about half of rotax twin stators are going by 50K miles. Germany is well aware of the problem, but there is no simple fix. The factory unit is a good one, my personal opinion, this is a lubrication problem. The stator is not getting the oil bath it needs to run within temperature tolerances.

    Theres nothing I am aware of that an operator can do will reduce the chances of this happening.

    A heavier rewind or exotic insulation would help, but finding someone qualified to do that will be a trick and a half.

    Sorry I don't have better answers.
    #69
  10. fastredbike

    fastredbike out riding

    Joined:
    Jun 9, 2006
    Oddometer:
    543
    Location:
    Moncton, New Brunswick, Canada
    this is so disappointing... too bad they couldn't at least figure a way to upgrade the bike's big brain and onboard computer to at let me know when charging has packed it in and I'm running on my battery alone...
    obviously some kind of battery monitor is going to be needed if you do much riding.

    and people wonder why I keep a KLR as a backup bike.:rofl
    #70
  11. Y E T I

    Y E T I Unpossible

    Joined:
    Apr 26, 2007
    Oddometer:
    8,332
    Location:
    San Diego
    Wow. :eek1 This statement, to me, is HUGE. Not knocking you Joel in ANY way, (I greatly respect your opinions and insight and appreciate your candor) but there are a couple of other problems with our bikes which you have downplayed. Your general opinion, as I have interpreted it, is that the perceived problems with the rear hubs and upper shock mounting bolts are due to the internet's ability to magnify the scope of problems, and not real problems.


    For you to agree there is a problem means, to me at least, that there is definitely a problem.


    Now we need to find a solution. I'm assuming BMW won't acknowlege it or solve it, so it's likely going to be up to us.
    #71
  12. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Voltage monitoring would be a nice option on the display. My 3 grand Italian motor scooter has a voltage display lol.

    The system does know the voltage, even stores under and over voltage codes. Wonder why BMW hasn't made display of voltage accessible.

    I suspect a load of aftermarket stators will hit the market soon, as well as some day supposedly BMW is going to sell this part apart from the rotor. I would happily cary a spare, once they are down around $100
    #72
  13. JRWooden

    JRWooden Long timer

    Joined:
    Jul 5, 2008
    Oddometer:
    4,643
    Location:
    The great state of confusion
    Thanks again Joel for your input and help!
    System voltage is one of the display options for the onboard computer for the BMW G650 XCountry, I've always wondered why it was not one of the selections for our bikes..... :cry
    #73
  14. LukasM

    LukasM Long timer

    Joined:
    Nov 15, 2006
    Oddometer:
    5,461
    Location:
    On a RTW ride - currently touring the U.S.
    A series regulator should fix it, or at least improve it. See the thermal images in post 22.
    #74
  15. C5!

    C5! Been here awhile

    Joined:
    Sep 23, 2007
    Oddometer:
    278
    Location:
    Venezuela & Colombia
    the voltage monitor is a nice (and necessary) feature but it is not the solution.
    I have installed a LED voltage monitor and it was useful because it allowed me to reach a safe place before dying, but it would have been better to being able to complete my 1,600 km trip between my home and my office instead of beeing stuck for 4 days right in the middle, with all the associated costs that it implies, hotels, going back and forth to fix the stator etc... carrying a stator is also not, IMHO, a solution. We need a reliable alternator which doesn't fail so early.

    I will definitely buy a new stator from this link earlier in the thread and I am thinking about this series regulator as well. I would really appreciate feedback from somebody who tried that, how it worked and if it didn't interfere with the other electrcial systems on the bike.
    #75
  16. flying.moto

    flying.moto Earthbound Misfit, I

    Joined:
    Nov 22, 2007
    Oddometer:
    375
    Location:
    South of Gorman
    I'm planning to get that series regulator as soon as when/if i get the bike I'm looking at with a full testing/writeup to follow. but then again, you might beat me to it :rofland then I will follow your lead! Good luck!
    #76
  17. JoelWisman

    JoelWisman Long timer

    Joined:
    Apr 22, 2009
    Oddometer:
    1,392
    Location:
    STL, MO, USA
    Hi Lukas. I like series regulators because they stand to improve fuel economy and I hate inefficiency.

    As to solving early stator problem, they may help or could harm.

    There are 2 common types of shunt regulators: "SCR" and "FET" Both do the same thing, the SCR type is cheaper, needs bigger heat sinks, allows voltage to move more and are electrically noisy. FET shunts are more efficient and produce cleaner power. They cost more to build and are more durable. Also they are not subject to thermal runaway.

    SCR shunt regulators are the most common. The common failure mode for an SCR shunt regulator is thermal runaway. When SCR's get too hot they usually stay closed, at least until they burn up which in our application would be pretty quick. SCR shunts are slowwww. Under load they can not switch modes faster then about 1 ms. In 5 wire applications, like ours the voltage spike during high speed engine operation causes on SCR regulator to clamp prematurely which is why our system voltage is lower at speed then at idle. This voltage behavior is not by design or desirable. It has nothing to do with the stator, it's just what happens when a manufacture specs a 5 wire SCR shunt regulator without a 6th sensing wire.

    The good thing about all shunt regulators is they are efficient at full output and hold stator voltages down nicely.

    MOSFET (Magically Obliterated, Smoke and Fire Emitting Transistor) shunt regulators are much faster acting and can even hold system voltages stable with no battery connected. When a power FET fails, it usually does so with a small mushroom cloud of smoke but fortunately they are quite reliable. FET shunt regulators are more efficient and don't get as hot. FET's nearly always fail open. This isn't strictly desirable, but at least the chances of a single FET shorting and burning out a winding on the stator is very low, unlike with SCR regulators which MAY be what is causing some of the failures on our stators. FET's don't have to be matched because their normal tolerances are super tight owing to how they are built. Slight inconsistencies in how SCR's are doped can lead to one stator winding carrying more load then the others and early stator failures.

    Series regulators are usually constructed with power SCR's because FET's can't handle as high of inverse voltage spikes. This isn't a rule and a very fine series regulator COULD be constructed using power FET's, but it would require more expensive components. Series regulators tend to be super electrically noisy. This can cause interference with components down stream and can also shorten their life. A bad battery connection or someone foolish enough to disconnect the battery while the engine is running will usually result in expensive electronics damage to the bike with a series voltage regulator. SCR shunt regulators aren't ideal for this but a whole lot better. A FET shunt regulated bike would run fine without a battery. This is of some concern on our bikes since the vibration of our engines is known for cracking the bus bar in the stock battery. Shunt, especially FET shunt regulators, but both to a great extent, are extremely efficient at transmitting the full power of the stator to the bike for charging and running loads. Series regulators convert a significant amount of energy from the stator to heat under full load conditions. I have only seen series type regulators used in industrial applications but those burned from 5 to 15% of the stators output.

    Most importantly, regulating by opening the stator output, Stator voltage is going to hit about 200 volts on our bikes at 6,000 RPM. This is a value that I measured, it is not theoretical. High voltage breaks down the stators insulation and may lead to faster failures then what we are already experiencing.

    I don't even know what is killing our stators prematurely on rotax twins. I suspect they are running too hot and that this is due to insufficient oil bathing, but I don't know this. It could be that what is happening is occasional thermal runaway and temporary shorting of one of the SCR's in the voltage regulator. Whichever is the case, will our stators hold up to high voltage better then high heat? It's going to take a bunch of people and 30,000 to 50,000 miles to find out, unfortunately.

    If a half dozen switch to series regulators AND the stator holds up for at least 30,000 miles, I'll go out and buy one because I would like the fuel savings, but I'm reluctant to lead on this one.

    P.S. the thermal picture way above is nice, but makes no sense. The hottest part, the part the cross hairs are on, should be empty! Secondly, emisivity is a huge factor when using infrared measurement devices. If the bike was dustier on the second measurement, it would look a LOT hotter. Infrared measurements are only meaningful when conducted by someone heavily trained on how to use them. I don't know who took those measurements and under what conditions, but at the least, I can't imagine why the case would be much hotter in the center then where the stator bolts to it. That makes absolutely no sense to me. In any case, at the best of times, infrared surface measurements are horribly inaccurate. I'm interested in case temperature measurements, but a simple thermal couple measurement is going to be FAR more accurate and much less prone to user error. Thermal couples are also cheap if anyone wants to buy one to take measurements.

    Wish I had something constructive to add, but I don't. Wait.... Maybe we could figure out a way to improve thermal conduction between the pole pieces and the case cover, also improve conduction between the cover and engine case. Aluminum bolts? Build up thermally conductive material between the cover and stator? Thermally conductive grease? Additives to make the engine oil more thermally conductive or synthetic which by nature is more thermally conductive? Just thinking out loud.

    Good luck, I hope someone finds a solution.
    #77
  18. C5!

    C5! Been here awhile

    Joined:
    Sep 23, 2007
    Oddometer:
    278
    Location:
    Venezuela & Colombia
    very very interesting discussion.
    I will not rush anymore about adapting a series regulator, I think there is some research to be done.
    as for the failures, i also agree about the lack of cooling. It would be interesting to get statistics about where the failed stators happened, one for my part is in the tropics where I live, and I think I noticed that some of the failures in teh US where in the South, Arizona etc... maybe there is something to investigate there.
    #78
  19. flying.moto

    flying.moto Earthbound Misfit, I

    Joined:
    Nov 22, 2007
    Oddometer:
    375
    Location:
    South of Gorman
    Very interesting indeed.

    This is what I found late last night: http://www.apriliaforum.com/forums/showthread.php?p=2598642 and i stated reading with page 18... but one guy on there did approach this with more or less scientific side, but(!) for some reason i did not see him test it for voltages "to the redline" but only for stator loads. The report he made does show quite a large voltage spiking when series switch is "off" and that is only at 2000 RPM. Check it out here

    I'm not sure if you guys found this discussion but it was an interesting read even after following this thread since these guys seem to have already tried/trying everything. It would be interesting to talk to some of them, and i'm thinking of contacting some of them too. Apparently this is how some people have been coping with similar issues on different bikes. Some seem to have experienced stator failures with FET shunting types as well. Where others have not seen or maybe have not reported any failures with series type R/R's. I personally would seriously doubt the insulation of the winding wires can't handle 200-250 volts. Just think about 220v motors, and welder transformers. The only thing that puts an "if" in that statement in my mind is the exposure to oil. Heat does not concern me as much if it is kept below 75-80C (just my personal "memory" limit from the times i dealt with buzzbox welders a lot, not a researched number). On the other side, can the series R/R handle the spikes and not loose its' circuity to the voltage? It does seem to have been paired with similar style "alternator" so it may be ok.

    And others are saying they retrofitted extra oil line with a spray nozzle to cool the stator! :eek1

    South AZ or not, this should not happen. I wonder if this engine/bike has been "hot box" tested in a desert... like cars.

    p.s. thank you Joel for measuring voltages. was 200 volts the highest you have seen? It probably goes higher closer to the redline... i would think.

    Anyway, those are my current thoughts for what they're worth.

    EDIT: I wonder if this stator on the F800 is Y or Delta connected... from how i read this Y would have more of voltage spiking capability while D would have more of current capacity (and possibly less spiking because no coil is "open" at any time?). Maybe the Harley stators they sell with Compufire regs are wound in D configuration? Correct me if I'm wrong, if someone has better information...

    EDIT2: well, it seems like Delta wound stators will not like low-rpm conditions as well as Y wound ones...
    #79
  20. JRWooden

    JRWooden Long timer

    Joined:
    Jul 5, 2008
    Oddometer:
    4,643
    Location:
    The great state of confusion
    Thanks for the education on regulators ... I'm going to have to read it twice more to understand everything you said :D

    I was scratching my head at the thermal picture as well ...
    I am wondering if the part in the cross-hairs was the bolt in th end of the crankshaft which would be hotter (maybe?) than everything else? :huh
    Not sure...

    A better oil bath would help ... I was looking at the starter gear train and wondering if it was pressure fed with oil....
    and if so could some of that be squirted on the stator but it seems the answer is no and no...

    Better thermal conductivity between the stator and the housing would help for sure ....
    Some sort of thermal grease (industrial version of what they use between computer CPUs and heat sinks) would help at least a bit... one of those fancy finned housings like they sold for the 1970's bikes that could be thermally grounded to the stator would be cool... :lol3

    The only other thing I have thought of is rewinding the stator with wire that has lower resistivity...
    The heating is result of I^2 * R heating so if we could reduce the resistance of the stator windings they would not produce as much heat....
    Either use a larger gauge wire (it it would fit) and/or select a wire alloy with lower bulk resistance, there are a few available .... might cost an arm and a leg though...
    #80