And here are the numbers when I switched the head light on. I must confess I am not sure if my high beam was on

To explain the different oscilloscope channels and number:

CH1 was connected between the negative terminal of the battery and the red wire coming from the regulator

CH2 is the clamp meter to measure the DC current going through the red wire from the alternator

CH3-5 are the three AC phases from the alternator. Each of the three current transformers are connected directly to a schottky bridge rectifier and then a burden resistor of 100 Ohm. What the display shows is the mV output over the burden resistor so 420mV over the 100 Ohm resistor equals 4.2 mA and since I have calibrated the setup (not with 100% accuracy) with another clamp meter I know roughly what current is going through the cables.

Here is closeup of the current transformer and the clamp meter

Kamelryttarn, that is an impressive array of instrumentation.  Thanks for taking the time to dig into the system and share the results.  I wish I was smart enough to understand all this electrical stuff.  Myself, I don't trust electrons.  They're sneaky little devils.  I mean, like, have you ever seen an electron?

I guess its good to have a handle on current and wattage and voltage, and just what those values are doing under varying operating conditions, but it seems to me that the original question centered around elevated temperature resulting from the regulator dumping more energy.

My bike is currently engineless.  If it was operational, I would do a simple test to determine if the regulator ran hotter when the headlight was disconnected.  

Start the test with the regulator at ambient temperature.  Take the three 6mm bolts out of the seat, put the seat in place, and go for a ride of predetermined length & speed & time.  Pop the seat off and shoot the regulator with an IR thermometer.  

Let it cool down until the regulator is back at ambient.  Disconnect the headlight.  Then repeat the test ride (same length, speed, time).  Pop the seat off and shoot the regulator with the same IR thermometer.

Did it run hotter, the same, or cooler?  

Of course, if it runs hotter with the light disconnected, we will be faced with the question "how hot is too hot?".

RocketScienceSmurf wrote on 08/09/19 at 22:13:57:


Not sure that scenario is what actually happens - the current understanding is as follows (and it may be innaccurate).

The output from the rotor/stator is dependent only on engine rpm.....it makes low amperage at low rpm and more amperage as the RPM increases.  The rectifier/regulator converts that power from AC to DC, and then regulates the voltage and current that goes to the battery and electrical system.  Excess power is run to a resistor in the regulator/rectifier and burned off as heat.

The belief is that reducing the electrical demand increases the amount of power that needs to be disposed of as heat......and if too much power is diverted the regulator/rectifier may suffer an early demise.

  logic tells me all you really need to do is stop over engineering the subject , and cut to the chase!  Go buy a  Laser pointing infrared distance reading thermometer ,($15.00 on line)   .Start the bike with the headlight on , carefully disconnect the positive cable to the battery to stop any flow that may go to recharging it . and take a reading off the fins of the reg/rec ,  after that simply disconnect the headlight(remove it) and take the reading again,  if the temperature increases  my theory is correct, if it decreases( BUT IT WONT) start slapping LED's on your bike with careless abandon .(K.I.S.S.)
  Can you disconnect the stater from the rectifier with the bike running? If you have to ask your in trouble ! the answer of course is yes , that's how you test it ,100volts AC @ 5000rpm if it's good. 3 reading between each two leads.Sorry but your assumption that using LED's will not lead to shunting more power and more heat are incorrect
 Let's me be clear ,I never said LED's were not good, if they provide better lighting than standard lights . It's just that our regulators don't prefer the extra heat.
  Resistance causes heat .  There are no resistors INSIDE the regulator.   The shunted power is sent to the body /fins directly . The body of the reg. being Iron based conducts , but has a much higher resistance to flow than the copper,  thus the power is converted to heat on it path to ground (bolted to the steel fender).The more power sent ,the more heat dispersed , there is no need for any kind of variable resistor.

What do we know?
The regulator has fins for shedding heat.
Some regulators fail.
Why not increase the heat sinking capabilities of it?
If you're concerned about it.

@batman: Your assumption that the generator or regulator/rectifier always supplies a fixed amount of current and that somehow it must go to either a load such as a light or be shunted of as heat is incorrect.

The increase of 12.6 watts might be due to the battery. We get power from two different sources in this case and as I said I am not 100% sure I got the range setting on that channel correct. I wasn't looking to get the exact DC current but rather an indication on whether the current was higher or lower.

A chain is never stronger than its weakest link and in the case of the alternator and the regulator I would think the weaker of the two is the regulator. I have read about more cases of failed regulator than failed stator but if you install a MOSFET regulator it might very well be that you shift the balance so you fry the stator before the regulator. I believe that the stator is more difficult to replace than the regulator.

The main reason a modern MOSFET regulator creates less heat is due the extremely low internal resistance of the mosfet transistor. A good modern mosfet has an internal resistance of only a few mOhms creating a very low voltage drop and therefore little heat.

Removing the battery from the electrical system while the engine is running is NOT something I would recommend anyone to do. The battery provides some filtration of the voltage and also a kind of base line.