Morning all. Running a 50A alternator with (c. 3k miles regulator and brushes).

Yesterday alternator stopped charging.

Battery is 13.1v at rest, drops to around 12.9 with key on, starts fine atm. 

On a 170 mile run yesterday voltmeter started showing around 11.8 at highway speeds. Got home and battery showed 12.4 at rest.

This morning I cleaned and tightened the alternator red and blue connections. Removed regulator: brushes clean and practically as-new length with free movement.

9 times out of 10 the battery charge light isn’t lit at key on (this is a new thing today).
If it is lit it goes out when starter pressed never to return. When its on its the usual nice and bright.

Voltage measured at both battery terminal and at alternator terminal - no change across the range - tickover to 3k.

I suspect a faulty charge globe or connection, any more impressive thoughts?

Is there a way to simulate the exciter circuit (3 ohm bulb +ive connect to the alternator blue and earth to engine?) so I can rule confirm its not an alternator issue. 

I have a working 30A alt I could I could swap in but would prefer to do EITHER that job or the lunch box dismantle. Where would you guys put your money?

That said, all my plans for Biaritz and Monte Carlo have been cancelled for this season so I could do both jobs if required. Then maybe Ill get round to doing the LED lunchbox backlights Ive been thinking of.

TIA, Stu.

Charge light some times not coming on at key on [and kill switch in on position] seems to possibly suggest dodgy bulb connections etc.

On a non ABS K alternator change is very easy. If its my holiday I think pack in the spare, seems you are getting charging but at least you have the solution on board. Just make sure you have Allen key that gets at the third bolt so you don't have to remove the starter although you do need to get at it from the left side. For good measure remove that bolt  and copper paste it and reinstall before you travel.

Regulator can go faulty internally, change there might before you travel might be a good idea.

You probably can eliminate the bulb itself as a failure point since you mention that it works 1 time out of 10.  Do you measure any voltage, relative to ground, at the exciter input of the alternator (i.e. blue wire) with (1) the ignition on, and (2) kill switch in normal position?  If not, then you can suspect the bulb holder and its connections.

You could temporarily attach a resistor of 150 Ohms 5 Watts to the alternator.  One side of the resistor would go to the field coil terminal on the alternator and the other side would go to a source of +12V on the bike.  If you start seeing the battery charge, then the charge circuit on the motorcycle is broken.

As an aside my dad years ago had an alternator failure on his automatic car. Could not get it sorted in time and heading off on holidays, 100 miles pulling a caravan. So he bought another battery and changed it en route. Had a charger to keep it charged and got the alternator replaced some days later.

Stu: further to Rob's suggestion, pull the back off the instrument panel and check the bulbholder. One of two things:

1. The bulb contact wires have fractured/worn thin/corroded (which idiot designed bulbs like that anyway?)
2. One or both copper tabs on the flexi-pcb has folder backwards rather than be poking down into the bulbholder's socket. It's dead easy to miss this when putting the bulbholder back in. What happens is

- you pull the bulbholder out and as you do so, the flexi-pcb flexes far enough so that the copper tab(s) clear the socket
- the copper tab(s) move into a position that is just slightly more than 90 degrees to the socket
- the flexi-pcb snaps back down but the copper tabs don't go back into the socket - that over-angle causes them to fold under the main flexi-pcb
- there's still enough copper exposed to make a connection with the bulbholder's contacts (most of the time)
- eventually they either stop making contact or they fracture away from the main flexi-pcb.

Seen a few like that. Actually, seen a lot like that.

Thanks guys. Since post I have fired up an LED +ive to the blue wire terminal and neg to engine so looks like I’m heading for the charge light wiring/connector.

Hmmm. Getting interesting in a K Kinda way. I put in a spare cluster, charge light good and strong at key on. Like the other cluster the charge light was an instant off after starting. So went back to a decent fake charge light rig with a 3w bulb, 1 terminal to a switched live and the other to the blue terminal. Exactly the same behaviour (light off after start on until revs over 1400 but and no increase in charge rate) so maybe  the cluster or the field circuit is still an issue but I guess something else to.

When the new regulator was added a few Thousand miles ago the voltmeter would show 14.4 at pretty much anything over tickover.

Anyone have a view on how long the light should be on before the field is excited, or is it purely about revs rather than duration? Putting this Q out there while I go back to do a regulator test.

You can fire up the K, red charge light on and leave it tick over like that for hours...eventually you completely discharge the battery.

It must be revved to get that light to go out.

92KK 84WW Olaf wrote:You can fire up the K, red charge light on and leave it tick over like that for hours...eventually you completely discharge the battery.

It must be revved to get that light to go out.

Yeah got that, but is a momentary (after starting) illumination sufficient to excite the alternator? 

I’ve cleaned the batt neg, no change.

I think I may have an exciter circuit/bulb issue too but Im going to find a cheap bosch rectifier regulator and install that before going back to my working 32A alternator (cos I really dont want to faff around with that LH alternator bolt unless I have to.

This is a schematic of the K-bike alternator:


When the key is turned on, but the engine is not turning, power is applied to one side of the charge light. The diodes (the little ones) block this current from flowing into the stator windings, through the bulb. It can however, pass into the voltage regulator, and from there into the field winding (magnetizing the core of the rotor, thereby getting the alternator ready to do its job). This is why the light comes on before the engine starts. 

Once the engine starts spinning, the alternator starts working. The big diodes provide current to the battery and the little ones provide it to the voltage regulator. Since the lamp is connected between them, it has about 0 volts across it, and it goes out. Everything is normal at this point. 

Having your charge light go out moments after the engine catches is not good. That would mean the output at the small diodes tracks the output of the battery terminal voltage.  Can you test the voltage at the D+ (blue wire) input to the alternator.  My expectation would be that the voltage will vary by engine RPM.  If it doesn't then, it might be an indication of a bad diode board.  Can you get access to the diode board to test the diodes?

Thanks Rob, one for tomorrow morning. Great explanation. Thanks
S

The output voltage at the +ive alternator terminal and at the D+ terminal (without the blue wire connected) and at the blue wire all show pretty much the same DC voltage as the battery and unchanging with revs.

Further reading also suggest that ANY AC voltage at the battery above zero was another indicator of a failed diode. Although not a constant value, continuous positive readings returned.

My new regulator arrives in a couple of days, so I’ll test that before pulling the alternator if required.

You don't want to know this Stu: you could have pulled the one off the 33amp alternator for testing

Bugger 

Is that a temporary thing for testing only, as different part numbers?

Stu

Dai wrote:You don't want to know this Stu: you could have pulled the one off the 33amp alternator for testing

Dai, on that same note part numbers for the diode board are the same across 30A and 50A, so is that a straight swap too? As I note are the regulators (DOH! just like you said, I blame lockdown).

Cheers, Stu

Yes. Apologies for the lack of reply - a pile of stainless steel and a welding torch had my attention. You may find that removing the diode board is a bit of a biatch. I did try once and stopped after a few minutes, but then again I wasn't in the mood for fighting an alternator that day.

If you quit at it then I’m quite discouraged. 

I’m not going to try until the regulator arrives tomorrow, as I don’t want to take the alternator off unless I have to. With only a couple of K miles on my current reg, it seems diode failure is more likely. 

Recon is the obvious but expensive option, I don’t want to go back to the smaller amp as I altered my wiring to suit the 50 and it’s nice and tidy. So I’ll probs have to take a punt on a Motorworks used one. Shame, as you folks have got the bike running pretty sweet this year.

If the diodes are accessible, you should perform tests in-situ to verify whether the diode board is defective or not.  Use the Diode function on your multimeter if it has one (most modern multimeters do) and see whether one or more of the diodes are shorted or open.

OK. Not the regulator. So, alternator out. I’ve got the 3 screws out of the diode plate but it doesn’t want to shift. Anyone have a ‘technique’ or a view on how robust I can be so the I can get to the diodes which I’m hoping to find on the underside?
TIA, Stu

[UPDATE] I can’t shift the diode plate but using a second hand alternator from motorworks plus a new regulator has got me back to charging at 14.4. So I’m back on the road, which is really all that matters, but will see what services the local alternator folks can offer to try and get my dead one usable as a backup.

Thanks for all the feedback.

bb

Bugger. I knew I had to reply to something here. You can get the diode board off but you'll need a fairly hefty soldering iron to desolder the windings from the board. A not-Bosch-finest-hour design.

Thanks Dai, I’m currently dismanting my old working 33A to try and mimic the diode and rotor test I have performed on the 50. Weirdly all the tests on the 50 seem to pass so I wanted to compare them on a known working alt.

Not sure what I’ll find as the two models are so different but I must be able to learn something, and I have more doco on the 33 to reference.

I’ve got all the bolts out, now trying to wrestle the rectifier assembly from the rest so I can get to the stator wires.

Also trying to cross reference the contradictory Clymer guidance on the diode test which says on the one hand: 
to test for low resistance and reverse test to get infinite resistance vs

to test for low resistance  and reverse test to get a second reading which is ‘the opposite of the first’.

Probs just poor language but quite important.

Watch this space.