I'm planning to build a dummy load to test a stator (generator on a motorcycle).

The stator is a 3 phase star type with no center tap.
I can't decide if I should configure my 3 dump resistor as a star or delta load.

Could you please suggest which and also why?

Many thanks.

 

It seems fairly obvious to connect a delta load,
If you were to set up in star fashion you would need to create a virtual neutral.
Max.

 

Yes, that's what I initially thought but after looking at the circuit diagram I'm not sure that is how it's configured on the motorcycle.

It looks to me as if the load is also connected in a star configuration, if so then maybe I should replicate this instead?

 

Your image is a three phase alternator including the six diode full wave bridge outputting DC into a single load. So where and what exactly are you wanting to test? Testing only three phase delta you would use three loads all wired phase to phase. Testing as shown you would use a single load.

If the alternator can be removed from the bike and disassembled you may also want to do a Megger test of the stator winding.

Ron

 

Yes, sorry. I would be separating the stator from the rectifier to test just the stator, with 3 load resistors.
I put the image up just to demonstrate how it is connected in use.

It appears to me that using the bridge rectifier would have a similar effect as connecting 3 equal loads in a star configuration, would you agree?

 

It appears to me that using the bridge rectifier would have a similar effect as connecting 3 equal loads in a star configuration, would you agree?

Yes, that would be the case. I haven't a clue what motorcycle you have. Personally being old and retro type I ride a 92 Harley Electra Glide.

If you only want to test the stator then yes, I would use three loads in a delta configuration. If we label the three phases A, B, and C you would have a load on A to B, then B to C, and C to A. If I were doing this I would be concerned with the overall alternator output after the voltage regulator does its job. On my bike I pretty much just look at my battery and system voltage and know if things are working correctly. I had a Honda years ago burn up a diode in the rectifier pack. Most bikes still have a separate rectifier bridge and regulator. While I won't rule it out I have never seen a stator section fail. Rectifiers and regulators are normally the problem. I want to guess my bike has a 32 amp alternator charging system.

Ron

 

What is the reason for the test, do you suspect a problem?
Max.

 

What is the reason for the test, do you suspect a problem?
Max.

Nope, no problems.

I have 3 bikes, if any of them were to develop a problem then two of them are more likely to have Regulator/Rectifier failure whereas the other is know for the stators failing.
When testing these components, the procedures shown it manuals aren't true operational tests.
I just decided that as high wattage resistors are pretty cheap from China, but take about month to here, I'd get some in now and actually make up a dummy load. This can be an AC load for the stator or reconfigured in series as a DC load for the Reg/Rec output.

 

If you were to set up in star fashion you would need to create a virtual neutral.

Just out of interest, why would you need a virtual neutral?
It appears to work on the bike without a neutral.

I know there has to be a good reason why when typically connecting a load to a 3 phase star you normally connect as a delta but what happens if you connect star to star (with no center tap)?

 

What about using suitable nichrome replacement furnace elements, these can often be had fairly cheap for load purposes.
I am not saying you need a virtual N, in fact I recommended the Delta load.
Just that this would be the configuration if connecting a load in star.
Max.

 

The resistors were £1.00 each, that'd looked pretty affordable to me .

 

Just thought it would eliminate the Chinese shipping delay.
Max.

 

I suggest that you can do a static check if you have a supply that will run an amp or two through each winding. If you can connect to the center of the star it will be simpler to understand the results. If all three windings draw the same current at the same voltage than either all 3 are OK or all 3 have failed the same. Then check the resistance to the frame of the stator, it should be quite high.
For the bike that uses a shunt zener diode that fails and leads to burned out lights, changing to a more normal series voltage regulator can make life a lot nicer. And a simple series voltage regulator is not that hard to make.

 

That sounds like a really good idea, I hadn't thought of testing the coils by driving a current externally.
I have a bench PSU that can supply up to 2 amps, which would be good start test. I also have a fixed 12v PSU that can supply about 15 amps, which I could current limit externally. This would be good as I could then build the heat up in the stator to simulate it being in normal use.

None of the stators on bikes that I've owned have had a center tap, but I'm sure a little detective work would soon point to the weak coil.

I like series regulators in theory but I do worry about them being used on a stator that has been designed to run shunted. I would expect a series regulated stator to have better/thicker insulation to deal with the higher running voltage of being partially open circuit. Then there is the voltage spike the back EMF is going to create when opening. Do series regulators feature voltage spike suppression to deal with this?

 

I like series regulators in theory but I do worry about them being used on a stator that has been designed to run shunted. I would expect a series regulated stator to have better/thicker insulation to deal with the higher running voltage of being partially open circuit.

Don't think a series regulator would work on a PM alternator, which is what these are. Or have never seen one for a bike. Since the magnets are PM, not an excited field like a car alternator.

 

There's no reason why they wouldn't operate with series regulation, in fact it is becoming more & more popular to do so, see this thread.

TBH I'm surprised that manufacturers are still churning out motorcycles with SCR based shunt regulators when FET based series ones would now be around the same cost if produced on the same scale.

I think if I had my standard SCR shunt type R/R burn out I might consider upgrading to a FET based shunt type, several motorcycles now come with these as standard. However this type will slightly increase the load in the stator, and thus the heat it will produce.
If however I have a stator burn out I would be seriously tempted to upgrade the R/R to a series type. Many high mileage motorcyclists have done so and documented it, and to the best of my knowledge there have been no failures reported.

 

There's no reason why they wouldn't operate with series regulation, in fact it is becoming more & more popular to do so

When I see the words, "series regulator", that calls to mind one that has an electromagnet for the field/rotor. That controls the magnetism in the field, not possible with the PM field. didn't spend much time looking at the regulators in the link but it looks like it is still a shunt regulator. See this quote from your link, "It works identically to the SCR in that it shunts peak current from & directly across the stator winding." Calling them what they call them doesn't make them what they call them.

As far as load testing the stators, what do you expect that to tell you? The normal testing done on motor stators is called "highpot testing", done with a "megger", not a "load test". https://en.wikipedia.org/wiki/Hipot

 

Ah, you should have read a little more from that page.

Shunt type shorts to ground when over voltage sensed. Series type breaks the circuit when overvoltage sensed.
Yes, I'm well aware of what "highpot testing" is. It is an insulation test and not a functional test.

The whole point of this thread, and thus my requirement, is to properly load test a motorcycle stator. Often stators will test ok to the basic static tests but still be faulty. The static tests are basic, just coil to coil resistance & coil to earth resistance when not running, then the open circuit voltage when running. By rigging up a dummy load the output voltage can be measured (or viewed on an oscilloscope) whilst the stator is doing some work and heating up. Sometimes a fault will only show itself once the stator is hot.

 

I got a question for you Rich. What value resistors are you planning to use? The reason I ask is this, a motorcycle stator outputs a voltage proportional to engine speed. Just as an example my own bike, with a three phase stator outputs about 18 VAC phase to phase per 1,000 RPM. The alternator itself is a 32 Amp alternator. Did you plan on just testing one stator or the ability to switch several loads for several stators?

Ron

 

I've ordered 3off 0.5 ohms 100 watt resistors, 1 for each coil configured in a delta or a star (I haven't decided which yet, might try both).

My reasoning was that 0.5 ohms is the nominal value of each coil in a particular stator I have, though this value will obviously vary from bike to bike.
The load resistors are in a way just acting like current bugs so that I can monitor the voltage output when under high load. As long as they allow a good current flow they should work fine.

At first I wasn't sure what value to go for, so decided to aim for the maximum power transfer from the stator, and luckily 0.5 ohms is a common value for high power resistors.
Either way I don't think you can go too far wrong as the stator is designed to run at 100%, even when shorted back to itself, so it can't be harmed by this and the resistors will be useful for other stators of different ratings.

JFYI - the stator specs I'm working to,

Resistance: 0.5 Ohms
Open voltage: 75v AC max (rpm not given)
Reg/Rec output: 13.5 - 15 VDC (open circuit)
Stator rated output: 180W (so 60W / phase)
180 Watts @14v dc is 12.8 amps @ 1.08 Ohms (assuming 100% efficiency)

So I can use all three resistors together to test the stator, then reconfigure the R's in series to load the DC side and thus test the regulator, I will have to be quick about it though as they'll heat up pretty quickly, unless I series all 3 together (9.33333 amps @14v).