From
http://www.ducati-upnorth.com/tech/chargingfailure.php
"Shazaam's Tech Library - Charging System Failures
A useful technical article from guest contributor Larry Kelly of San Diego CA (aka Shazaam!).
Ducati Charging System Failures
There's been a number of questions posted recently about charging system problems. This write-up is geared towards superbikes but has general application to all late-model Ducatis. It's probably more than you wanted to know, but save it, it may come in handy ...
Ducati charging system components, are prone to failure, especially on superbikes. When regulators first started to fail on the '95 916 bikes, Ducati changed the regulator design but that didn't stop the warranty claims. The first "improved" regulator that was sent out to dealers in 1995 was marked with a green dot of paint (to distinguish them from the older version) but otherwise appeared the same. They initially thought that it was just a bad batch of regulators but the failure reports continued. Not every bike experienced this problem, but some bikes experienced multiple failures.
The second iteration actually changed the regulator design and added an aluminum backing plate that performs three functions: it shields the regulator from the direct heat of the horizontal cylinder's exhaust pipe, it acts as a heat sink to help convect away the regulator's internal heat, and it helps to better direct cooling air from the fairing duct onto the regulator.
New Voltage Regulator Design
Regulators put out a lot of heat themselves and have integral cooling fins built into their casing to help reject this heat. The additional heat sink, although not finned, was however somewhat reflective and designed large enough to shield the RR from the direct heat from the front cylinder exhaust pipe.
Note that if you fit one of these to an older bike you loose the functionality of the charging light.
The charging light is the first sign that the RR is about to fail. It will often start to flicker at above-idle rpm, sometimes requiring progressively higher rpm to extinguish it. Or, it may just come on and stay on without warning.
If this happens while you’re on the road, stop and pull the electrical lead wire (under the rubber boot) to your headlight to reduce power demands. If your battery was at full capacity before the regulator failure you probably have enough reserve capacity to run your bike for 50 miles or so if you have to limp home.
The second regulator redesign also incorporated a revised electrical connector so Ducati sold an upgrade wiring adapter kit to connect it to the bullet-type connectors of the original design. The apparent reason for this new connector style was that the original connectors would loosen and corrode so as to develop a high electrical resistance and overheat. Unfortunately the new connector style had the same problem.
Voltage Regulator Connector Overheating: Old Design/New Design
The regulator fails because there is often not enough cool air available to remove the heat generated internally by the unit and the internal electronics are damaged. At speed, outside cooling air is supplied directly to the regulator via a NACA-style duct opening in the fairing. It apparently wasn't big enough, so when the 998/748 was released, a larger duct was incorporated into the design.
Old Fairing vs. New Fairing NACA Ducts
In traffic however, there's not only no forced-air cooling, there's really no cooling air at all - the front cylinder exhaust pipe superheats the air in the lower fairing.
So the regulator still gets cooked at stoplights from its own internal heat and the nearby front cylinder exhaust header because there is such poor cooling air circulation in the lower fairing. The newer bikes have a revised fairing design with openings on the upper surface that may help convection cooling somewhat.
The best solution is to move the RR outside the fairing to the outside air stream for better cooling.
In my case, I replaced my fifth failed regulator with an Electrex unit, extended the larger gauge wiring (see below) and mounted it on the underside of the license plate holder facing the rear tire where it’s hardly noticeable.
Relocated Regulator for Better Cooling
Adding a regulator cooling fan was tried first, but the ambient temperature inside the fairing is just too hot to use as cooling air. Even though I wrapped the front header pipe in high-temperature insulation to reduce exhaust pipe heat transfer in the fairing to a minimum, the regulator surface temperature still climbed well beyond 160°F at traffic lights.
http://www.thermotec.com/products/full/11001/11001.html
So far so good. The regulator still gets really hot while stopped, so it's no wonder that they fail when inside the fairing. If it's any consolation, Honda's do the same thing.
Then there's the problem of heat-damaged stator wires.
The initial charging system design had a 350 watt single phase alternator. Here's the design basis that Ducati used for sizing the system:
Design Basis
Note that the system wasn't designed to run both headlights simultaneously.
Note also that the power rating of the fuel pump assumes a clean fuel filter. Ducati fuel injection systems use a high pressure fuel pump powered by a DC motor that draws a fairly-high current and a dirty/clogged fuel filter will causes the motor to work harder and draw a much higher current. So the in-line fuel filter inside the gas tank requires regular replacement. A pump feeding fuel through a clean filter will draw about 2 amps, but this figure can rise to 10 amps with a dirty filter. Clogged fuel filters are a prime contributor to regulator failures.
Of course the electrical system doesn't normally operate with everything turned-on at the same time. So here's a more common operating condition:
It's important to note here that, given the obvious need to charge the battery, the stator wires will have to pass 29 amps (350 watts/12 volts) continuously. So I advise you to keep your bike on a trickle-charger whenever possible to reduce the demands on your charging system. Also, avoid the temptation to replace your battery with a low-weight, lower capacity unit. The principal advantage of using a larger battery is to be able to restart repeatedly. If you regularly don't ride long enough to recharge fully between restarts, stay with a larger capacity battery. A lower amp-hour battery will need to be trickle-charged more often.
For 1999, Ducati redesigned the electrical system, going from the 350 watt design to a 500 watt three-phase system. (Three-phase alternators have three wires coming out of them, single-phase have two.) They produce an AC output that has a higher frequency than the earlier single-phase design so the regulator presumably has to do less work (i.e. less heat) conditioning the waveform and converting it to DC. The new alternator also has additional wattage available to run the lights, fuel pump, ECU and accessories but the higher output still has to pass through the regulator.
Which leads us to the second problem with the Ducati charging system ...
The wires running from the stator to the regulator are seriously undersized and their insulation becomes damaged from overheating.
The general rule-of-thumb for sizing wiring is that if it needs to carry 20 amps use #12 gauge wire; 30 amps needs #10 gauge; 40 amps needs #8 gauge. Early bikes have 350 watt/29 amp two-wire charging systems so #10 gauge wire should have been used. 1999 (and later) three-wire 500 watt alternators produce a little over 40 amps so #8 gauge wire was needed (but not used.)
So you ought to replace the wiring all the way back to the stator with a larger gauge.
These wires have to pass high amperage continuously. There’s an electrical phenomenon called I-squared-R loss. That is, if you run 30 amps through a corroded or loose connector or undersized wire having (say) a 1/2-ohm resistance, the heating effect is 30 X 30 X 1/2, or 450 watts. That’s a lot of continuous heat. This heat just conducts down the wire, cooking the insulation as it goes.
Just like you can’t put your hand on a 450 watt light bulb while it’s lit, you can’t expect a plastic connector or electrical insulation to survive radiating 450 watts of power either. A corroded or loose connector always has a higher resistance than the adjacent wires and it will heat up enough to melt connector plastic parts and adjacent insulation. That’s why it’s best to solder the wires directly together and eliminate connectors entirely.
Generally though, it is the connectors heating up that causing the insulation and conductors to melt and short out. Even if you haven't had a charging failure, next time you have a chance, check your wiring for damage.
On my 916, for example, the stator wires got so hot that the insulation became brittle and cracked near the regulator connectors. A closer examination showed that the damage extended along the wire all the way back to the engine casing. I could scrape off the softened insulation with my fingernail.
So even if the stator wire resistance measurement and the voltage output checks out OK, the output to the regulator is unreliable because the insulation between wires (that run in a common sheath) breaks down at the higher voltage levels at higher RPM.
The undersized stator wires generate the highest heat load at the points of highest resistance on both ends; connector to the regulator/rectifier (see above) and at the other end where they attach to the stator. Note in the picture below, the discolored heat-damaged insulation on the two yellow wires where they connect to the stator coil lugs.
Heat Damaged Insulation at Stator
Finally, I recommend installing an inexpensive digital voltmeter to get more information about the health of your electrical system. An on-board voltmeter will flag a charging system that is behaving erratically and intermittently prior to failure. When the charging light comes on it's too late.
Digital Voltmeter Installation
The model (Lascar EMV 1200) I use is shown here installed on a superbike: The meter's current draw is so small (3 ma) it can be left on continuously without draining your battery. You can buy it in the UK for £20:
http://www.lascarelectronics.com
To install, cut both lead wires to within an inch or so of the threaded stud. Solder a more substantial wire to each lead and run them directly to the battery terminals. Fit a piece of shrink tubing over the voltmeter's threaded stud and wires to prevent excessive bending or fatigue failure where they exit the meter.
As you can see, I mounted the meter directly below the water temperature gauge on my 916. You can run the shrink tubing covered wires down between the bracket and the instrument housing. The meter is so light-weight that no additional mounting bracket was required, the stiffness of the shrink tubing is enough to keep it from moving around. A small piece of foam placed between the bracket and the bottom of the meter will keep it facing you. If you mount it too rigidly it will see a lot of vibration that could shorten the life of the meter.
As a final note, use the above electrical load tables to see the effect that additional electrical loads will make when you make changes such as running both headlights together, added auxiliary equipment, or the use heated clothing.
The 1995–1998 superbike models were designed with a 350 watt alternator. The 1998 ST2 is unique, it was Ducati's one and only 450 watt single-phase system. For the 1999 model year 1999, Ducati redesigned the system going to a 520 watt 3-phase alternator. Make sure that your power demands during all expected operating modes don't exceed these levels. Adding additional electrical equipment or using a lightweight low-capacity battery, on pre-1999 bikes in particular, isn't very prudent."
Linear Mode
