Another alternator hits the dust and not under warranty

I'm wondering if that will do any good or if the case is even bonded to the ground side of the circuitry ???
I reckon I"m going to have to get the ohm-meter out and check continuity from the ground wire to the heatsink...
They sure didn't do anything at the factory to enhance the grounding I can tell you that much..........

I also went to the Haynes manual and only found the wiring diagram. It shows two wires going to the battery return. this seems to imply the current is at least some times large enough to require two wires. ..and likely good connections on those two wires.

Battery connections are prone to corrosion. Tomorrow I'm going to check my battery connections...Haynes says they are the heavy brown(1.25) wires. The wiring diagram says there should be another very heavy (6) brown wire from the battery going to frame which acts as the starter return path, spark plug return etc.

In the wiring diagram there is no indication that the Regulator is chase grounded so JRWooden findings seem to confirm there is no chassie ground there.

In my manual it shows two @ 1.5 wires -- I'll "assume" (yes I know) that the 1.5 means 1.5mm wire diameter...
1.5mm is roughly equal to 14 AWG and for the length of the run back to the battery 2 conductors @ 14 AWG or even two @ 1.25mm would be adequate ....

I'm not sure where all the ground wires get bonded together ... I don't think it is AT the negative terminal,
but wherever it is... as new as these bikes are, I would not think that this would be the problem....

... and ... I'll have lost all faith in German Engineering if I find out they were relying on a ground path being available thru the heatsink of the regulator .... so I'm not sure where that leaves us, but I think I'd rather be over here:

http://www.advrider.com/forums/showthread.php?t=163155

I do still have a factory wiring diagram, as well as the fact that I have pulled one of our R/R's apart that took impact damage from a fork lift. There is ABSOLUTELY NO electrical connection between anything in the R/R and the anodized case. ZIP ZILCH NADA NONE. Removing anodization serves absolutely no purpose but to waist time. It also does not particularly lower the temperature of the R/R to grind on it and use thermal paste because it's bonded to a part of the frame that gets hotter then the regulator. The regulator sheds heat by radiation and some convection. Further, I have yet to see or hear of a single regulator for the F8 electrically fail, so not sure why we would care anyway.

As to adding more ground wires or improving the ground path, (which is from the R/R to about mid harness, spliced to the brown wire between main ground from battery and brown wire that grounds engine near starter) why?

From a stator current perspective, if this did anything at all, the only thing it could do is increase stator current by lowering load resistance. If the theory is that a poor ground is causing voltage to climb in the stator, thats true. If excessive voltage causing partial discharges and flashover of the stator is our issue then a series regulator should kill the stator much quicker.

I suspect all the theories on other forums that regulator wiring is the cause of stator failures is 100% wrong.

The regulator portion of the R/R can cause stator failure by shorting one winding, but even this is super unlikely cause usually this type of failure would blow the diodes into outer-space long before the stator felt it.

The regulator reducing load resistance by pulling the voltage too low could cause problems but only if the stator / rotor is designed wrong and can't operate at full power output.

Open circuit connections to the regulator could cause it IF the stator can't handle open circuit voltages and certainly this is a hard thing for stators to cope with unless designed for it but with this would go flickering lights. Further, theres a lot of power going through these circuits. If theres a bad connection anywhere in the stator, R/R or R/R to bike wiring, that bad connection is really likely to burn up and produce somewhere between a cloud of smoke and a fire.

Interesting theories these other forums are putting forward but I'd bet a shiny penny that they are 100% wrong about this one.

I think that if we are trying to blame the failure on a mechanical connection...ie...the ground path..... Then perhaps the people with failed stators....could have a look at the wires from the R/R to the battery ground.......( seems like this is the path...rather than a ground of the R/R itself).... I will not dispute the aging of the winding.....etc....since I have no clue in that regard... But.. If the wire/s from the R/R to battery neg and subsequently chassis gnd....is somewhat corroded.... It could cause the stator/RR not being able to regulate. There are several similar issues on other bikes..... That can be fixed by soldering....rather than crimp connectors. A multimeter might not tell the whole story...ohming the connection....as vibrations....and the tune of up to 400 watt of energy dissipation could make a mess..... But tracing the Chassis gnd path....and solder it...will at the very least eliminate one suspect....imho....

Erling

A multi meter is the answer to finding these problems, but not in ohms mode.

Set the multi meter in volts DC mode (manually range it up to at least 200 volts), use a piercing probe on the ground wire near the regulator, alligator the other probe to the battery ground post.

Put the meter in min/max mode and go for a ride. Since your measuring in series with the the ground path, any resistance will be super easily and accurately detectable regardless of if its at the R/R connection, mid harness splice battery terminal or anywhere else thats in series.

Eyeballing and cleaning connections doesn't do anything if it's a mid harness break causing problems but series measurements capture all and tell all.

If your really interested in resistance measurements then you need at least a 2 channel scope and to measure current at the same time but thats just for entertainment value.

An average voltage drop in the entire ground series at these points of over 0.4 volts is bad. A high peak indicates an intermittent bad connection.


While your at it, measure from the battery ground post and starter engine ground. Grab bare metal right on the starter. Crank it over with meter in volts and min/max mode. 0.8 volts peak is acceptable. over 1.0 volts and there is a bad connection for sure.

All of this assumes a min/max sample rate of at least 1khz doing 1-100 m/s average peaks.

Now do the same between + battery post and starter + post. Here up to 1.0 volt is ok,as the starter cranks over 1.3 is bad for sure. On this side you can't start the meter recording until the bike is already cranking, i.e. until the solenoid has closed the circuit.

Same for positive output of R/R and same figures as with ground.

In any case, the regulator reaction time is measured in microseconds. Open circuits or bad connections will unload the stator causing voltage to climb, just EXACTLY as a series regulator does almost continuously. In the case of the series regulator eventually causing problems through premature aging MIGHT occur, but nothing quick. The F8GS regulator is 5 wire. That means any bad connections will equal undercharging. If a connection were bad enough to maybe cause stator failure years down the, the bike wouldn't charge at all and run dead every few hours. If the connection were just marginal then it would never hurt the stator but would likely start a fire.


Everything I have ever been trained or learned across 3 professions tells me the theory of stator failure of a PM alternator with a 5 wire regulator is absolutely 100% impossible.

This is true of many other bikes including one I now own, but it never leads to stator failure, it leads to undercharging, melted connections were they are going bad and fires.

This is not the F800 stator failure issue.

Agree, I seen Aprilias which had exactly the problem and after soldering the connectors all problems were gone.

Steve

Thanks Joel ... looking at the wiring diagram it did not seem the R/R was intended to be a ground point, glad to know that is confirmed (and my faith in German engineering now less in question ) ! Perhaps it's a problem on some other bikes where the R/R is intended to be a ground point but that's not the case in the F800GS design ....

For those that have not seen it... here'e the wiring diagram ...
the blue box at the top is the ZFE,
note that the pos. & neg. leads from the regulator are doubled up...

I agree with part of that...... The undercharging...... But how do you them propose that the stator yield the "non used energy" to chassis ground..... When said path is......should we say.... not good.??? Excess power is grounded....or shunted to chassis ground...... if you diminish that path..... I agree that it happens on the DC side......after the regulating..... But it is still energy.....and's gotta go somewhere...... So in short.... If the regulated now DC's path to ground ( excess energy) is not good..... Then it could back up and the stator cannot get rid of the energy...... I find this a very likely suspect......again...imho.....

Erling

It's been about 20 years since I was in a classroom studying fundamental laws so I may not be able to explain this as well as is possible.

This is a super complex subject, and when approached from the standpoint of energy, theres no way to simplify it.

Taken to extreme, if you were to spin the rotor around a stator that was open circuit, zero energy would build up. Voltage would build up, but voltage is "potential" energy, not energy. In electricity energy is defined by "potential energy" (P) also known as "voltage" multiplied by "intensity of current (I), also know as "amperage".

In the case of the open circuit stator, there is no path for current to flow, therefore the math would be "x voltage" multiplied by "0 amps". Anything times zero is zero.

open circuit voltage is perfectly linear and is established by the magnetic field density multiplied by the rate of change, which in our case is the number of poles on a single phase multiplied by engine RPM.


Now it COULD be that the k7 platform has such a lousy stator that it isn't insulated against the maximum open circuit voltage possible at RPM MAX and the stator develops an arc right through the insulation. I highly doubt this, and even if it was the case, no re-winder would ever use magnet wire with this low of insulation yet rewinds have failed and rapidly.

This is the only way for voltage to suddenly kill a stator, it generates an arc which burns the insulation at one point which then creates a short that burns out the stator. At 6,000 RPM the F8 stator has an open circuit voltage of about 200 volts. At redline this voltage will be about 284 volts.

The inner layer of the insulation used on the k7 stator magnet wire is normally assumed to have a breakdown voltage of 2,700 volts. I don't know how much more the outer layer will add, but it's moot cause open circuit voltage is never going to get with an order of magnitude of what it would take to cause flashover.

Partial discharge aging of the stator is a wholly different subject, but it does long term damage. bad connections from the R/R are not going to cause any long term damage that a series regulator wouldn't continuously cause, and theres no way this could continuously happen without burning the bad connection completely up long before even slight long term aging occurred.

Open circuit stator = zero current flow = zero energy = zero stator heating (beyond small amounts of heating from hysteresis and inductive reactance and eddy currents all of which are small enough to be irrelevant.

High resistance connection = higher voltage, lower current, less heating of the stator.

Actually, creating a poor connection, i.e. adding resistance between the stator and R/R would lower stator temperatures. It would also lower output. You don't want higher resistance after the R/R cause this will directly lower charging voltage.

Does any of this make since? If not, here is the technical explanation http://dspace.mit.edu/bitstream/handle/1721.1/33569/63516405.pdf?sequence=1

Joel.... You knowledge is great.....and I will agree with most of your statement above...... With the exception of the overall conclusion...... If it was as simple as you stated...... Well... Not really simple........ But The thesis as you so elegant explained...Not kidding here... has one simple piece perplexing me.... Theory is great....and while for the most part is or can be part of a resolution...... But when things go wrong electronically.....and surges...spikes.... voids...Call it what one chooses.... can wreak havok... A firm ground is needed....and I am of the opinion that a poor connection on the DC side can cause issues over time upstream on the ac side.... I have seen this on other bikes...... Not Prillas...for the record.. but others..... But in perhaps to not run this any further as to bestow on others lengthy nights of boring read...... I will just rescind myself to check my connections on my charging system.....solder all connectors.....and be done with that...... And when my stator fails..... I will owe you a cup of the finest espresso.....



Erling

alright, I will relent :) Way way far out are things such as "every frequency in the spectrum is present in an electrical arc" and "resonance induced spikes" and "ringing"

I don't personally believe this could ever hurt the stator, but that is an educated hunch based on some education and experience. There are phenomenon that I am only mildly familiar with that I could be guessing wrong about the significance of.

Mostly I simply believe that a bad connection in this powerful of a circuit will make it's self obvious pretty fast, but the consequences at the bad connection can be quite bad so checking those connections is ALWAYS a good idea as is the positive side.

Thanks for the complement by the way. I am in awe of your knowledge on many subjects so your complement is flattering :)

Absolutely Joel......And well deserved it is.... I was sort of devastated when I red that you sold your gs...... But good it is that you still hang here with your willingness to explain. I used to work on air defense missile systems in Denmark.... half plummer and the other half chasing electrons...( since captive energy was led in the form of both cabled, and waveguided) I don't mean to set forth that experience as a shield.....as it has Zip to do with the stator..... I learned a lot in those years.....One thing stand out..... I cant count how many times...multimillion Kroner piece of equipment was left inoperable because of a bad ground. Several times.... It took way too long to find.... Because it did not make any electronic sense..... That said..... Rather than modifying the charging system with extra wires....etc.... I fully agree that checking and verifying that the existing connections are sound. Perhaps rather than a fix...... It just gives me the warm and fuzzy feeling......Until someone finds the smoking gun.....

Erling

This has been a very good discussion. However it is clear that we are hampered (or in some cases freed) by a lack of knowledge of precisely how the BMW/Rotax/Denso charging system is designed.

We do not know the design of the stator, materials, winding, where/how saturation / self limiting occurs. What are the parameters of the eddy currents?
We do not have a schematic of the regular or rectifier, know the the electronic components used, so can not in any reasonable way predict or eliminate failure modes and consequences.

Use of multimeters, while valuable, are a crude tool. By their very design they filter out all but the very lowest frequencies. There is no way to see the complex capacitive or inductive reactances (why generators work).
Most poor connections, especially those due to corrosion conduct, differently at different frequencies. EtcEtc

We just do not know.

We need BMW/Rotax/Denso to stand up and address this problem.

Having not put my bike back together here's what I can add:

The OEM R/R (on my bike anyway) is mfg. Shindengen - it has the corporate logo stamped on the back,
along with P/N: SH541G-12

If you google the P/N it is used on many bikes - Goldwings and other larger Hondas, and etc...

The corporate web page is not as helpful as I'd like ...

http://www.shindengen.co.jp/product_e/electro/catalog.html

http://www.shindengen.co.jp/product_e/electro/reg.html


but my guess would be that the schematic for our unit is the second one down on the page ...
I gotta get to work so need to knock off on this nice diversion for now ....... and earn some cash to put in my reserve fund for a new stator, a replacement fuel pump, and etc

I suspect you are waiting for a horse that isn't going to come with BMW, and Rotax and Denso has nothing to do with this. BMW designed this engine including the charging system from the ground up except for parts they bought of the shelf. Rotax just put it together to their specifications and occasionally not to lol. Plenty of stators fail on plenty of bikes, my guess would be if BMW addresses it at all it will be in a silent rolling production change.

I have a an F8 coming by for other reasons. I have a 4 channel 200MHZ DSO 2.5GS/S 400pS resolution, A reluctance meter and a 30MHZ 3 phase harmonics analyzer as well as a 2MHZ octopus and I'm quite proficient in all of their use. What is it you wish to see?

Somewhere I have video of the output waveform on one leg at a resolution of 4.2nS a division and the transient induced by unhooking the R/R at 3,000 RPM but I can tell you right now the transient was milk and toast.

A while back someone asked what temperature the oil might reach ....
From this post (data gathered in real time with GS-911) the answer is that 250F would not be extraordinary:

http://www.advrider.com/forums/showpost.php?p=15438749

so ... I would guess that the heater windings <errrrrrrrr STATOR windings> could easily see something a lot higher than that....

A mechanic told me that there was a car that was constantly in trouble with the starter, and that every year the car took a starter again, then began to suspect the exhaust pipe passing just 5cm from the starter and began to cover it with heat wrap and the car never had problems again. In old Ténérés also had problems with heating the oil and then they would put heat wrap in the exhaust pipe that passes under the engine and did get extremely hot sump oil. So I think it will not hurt to put heat wrap into the exhaust systems of the F800. Someone other than the Cowboys have already applied heat wrap? And what was the outcome?

Cesar:

I had proposed a "group purchase" of the insulating material and stainless steel ties to do this...
It would cost $20 or less per install if I could buy in bulk.

To see what if any the affect the insulation had on the condition I think we'd have to instrument a bike with a thermocouple in the stator and take some before and after measurements. (This was Joel's idea originally I think, and is likely the best (only?) way to do it properly). I may have some free time on my hands the later part of next month and try to do this....
I have a GS-911 on order which I think would allow me to monitor oil temperature in real time ... that coupled with stator temperature would be interesting to view.

While exhaust heat may not be the "root cause", if insulation does significantly reduce the stator temperature it does follow that the life would be extended.

does anyone know -- IS IT LIKELY I WILL NEED A NEW SIDE COVER GASKET ... IF SO I'LL GO AHEAD AND GET ONE ON ORDER...

It would be interesting to compare the stator design of the twins to the stator of other BMWs.....

Random