Starting has always been a gamble for me; not motor-wise, that's always been fine, but its a big bike to bump- and at my age not a good idea to try! This means that in the absence of a kick-start you are reliant on the starter motor and its chain of components.
I'm guessing we've all been there... operate the starter and nothing happens... well usually something happens but it doesn't result in engine turnover or motor start. In my experience one of three faults can screw up the starting system - and I've had all three. I caution you be wary if the same symptoms present themselves- its not always the same cause!
OK scenario 1- you operate the key and get a loud buzzing or rattling sound. This is probably a weak battery- the battery has enough power to operate the solenoid maybe even the lights alone, but when the solenoid activates, it connects the battery directly to the starter. The starter draws a huge current and so promptly soaks up everything the battery has to give, its voltage collapses and the solenoid disengages. This lets the battery voltage recover and so the process repeats- fairly rapidly and the obvious buzzing results. If this is your problem then a pair of jump leads should sort it immediately.
A poor battery may simply be an old battery but if the fault recurs and you find yourself going through batteries like chocolates, check the rectifier/regulator- Mine was overcharging and cooking the batteries so even a replacement didn't fix it for long.
Scenario 2 you operate the starter, there's a loud clunk but the starter doesn't turn. Jump leads don't help. For this you have two potential causes.
Cause 1- the solenoid has failed and isn't connecting to the starter. As the starter isn't getting power, its obviously not turning. You might find that if you turn your headlights on, then the bulb doesn't dim appreciably as you push the start button. It should virtually flicker out if the starter is taking power. You can test the solenoid by shorting across the two solenoid terminals with a screwdriver or equivalent The terminals I mean are on top of the solenoid under a rubber flap and have chunky, thick wires going to them. There is a reason for that, they are intended to handle a whacking current so be prepared for a flash when you short the terminals- and for whatever you are using to get really hot! However this bypasses the solenoid, so if that's your problem the starter should now turn. Incidentally this test works regardless of the position of the ignition key, but if you want to start the bike this way obviously you will need to have it turned on.
Cause 2 the starter is taking power, but its not turning. There are a few explanations for this but the most likely is that its either shorting out internally or something is physically jammed- either in the starter motor or the bike engine. Both explanations would mean that the starter will soak up large amounts of current to no effect. The headlight will dim whilst trying to start as all the current is sucked into the starter. You may notice the chunky leads described above that go from the solenoid are getting really hot so don't hold "start" down for too long. Obviously shorting the solenoid or using jump leads can't help in this case.
If you are "lucky" then maybe its just that you've left the bike in gear (duuuh!), or maybe the starter clutch or it's drive chain is jammed somehow. Check the gears and if that doesn't work try to loosen whatever is jammed by putting the bike into top and rocking it forwards and backwards on the wheels to shake everything up a bit. Put it back into neutral and try again. That might free the jam and let the motor turn... But its a short term fix. You will need to investigate further. Of course, if that doesn't help then you probably have a jammed or shorted starter motor and there's no alternative but to take it out and fix or replace it.
Well I've had all three of these and considering I do a restrained mileage I think that's not really very fair. Certainly its not the sort of legendary reliability we've come to expect from Japan.
The first time I had the problem it was a poor battery. It happened soon after I'd acquired the bike and I just thought the PO had probably never replaced the battery. However, when I found I was running through batteries like chocolate bars I realised it was more serious. I had a duff regulator rectifier that was overcharging and brewing up the batteries like teapots. Swapping that fixed the problem for a couple of years. I then had a duff solenoid- again changing that cured the problem, and now in its third incarnation; I have a stuck or otherwise buggered starter motor. No choice now, got to remove it, strip it and see if I can fix it! This is my experience of doing just that!
Removing the starter motor
Before we start, a warning; access to the starter is fiddly- so chant your favourite mantra before you begin. If you want an easier life then remove the fuel tank, carbs and any engine bars fitted as well as the saddle. I managed without removing the tank or carbs and I left my Renntec engine bars in place- although I did have to remove the front bolt and lever the bars open to release the alternator cover!
You have to remove the alternator cover (left hand side) and its held on by 9 fastenings. 8 of these are 8mm bolts and one is an Allen cap screw. All of the bolts are the same length except two which are longer. The longer bolts go on the lower front of the case.
The Allen screw foxed me for a bit- it looks like an oil level screw, but alternators don't usually run in oil so I wasn't expecting any in the case. Well this one does! Be prepared and get a drain tray.
My camera battery failed during this strip and many of the pics I took didn't save but essentially removing the cover is like this....
Drain oil
Remove the alternator cover screws
Remove the 4 bolts holding the chain sprocket cover to give room to move the alternator cover
Tap the cover with a soft mallet or use a wooden bar to tap the front lower section where there is an accessible lug.
The cover is stuck on firmly with gasket goo and retained by the alternator magnets as well so it can be tricky to remove.
If you have engine bars remove the front mounting nut and loosen any others you can get to on the left of the bike. Lever the bar away from the motor so that you can fineagle to cover over the alternator and off the crankcase. It will remain attached by some electrical leads passing through a rubber block. This should just push out of the cover, but again its a very firm fit.
You should now see two cross head screws that hold on the starter chain tensioner. Photograph the tensioner in position as its tricky to see how it goes back- these are some of my lost photos so I'm a bit stymied there! Hope I can figure out how to refit it. Anyway remove these 2 screws. I recommend an impact driver as they are in quite tight and their heads are made of cheese. They will strip easily rather than come out.
![]() |
| Chain guide is made of 2 parts and held by 2 screws. |
Remove the two sections of the tensioner and you will have enough slack to slip the starter chain off the starter sprocket at the top.
Next remove the thick wire terminal attatched at the top of the starter (10mm flanged brass nut).
Move to the right hand side of the bike and remove the two 10mm bolts that hold the starter attached to the crankcase. You will need a 10mm socket on a short wobble bar extension. Note that the battery/motor earth cable attaches at the rearmost bolt as well. Don't forget to reconnect this when rebuilding.
Now you have the pure joy of trying to remove the starter from its home when all exit routes are smaller than the starter in at least one dimension. It is necessary to push the starter backwards (ie out of the lhs case) whilst lifting it as far as possible to let it slide back. Once its out of the case you then have to try twisting and shimmying it in all possible orientations until it comes out. There is one- and as far as I know only one route by which the starter can come out, I found it by (considerable) trial and error and I can't remember what it was- sorry I can't be more help. I despair of refitting the motor!
Anyway, all being good you will end up with this.... I tried to rotate the motor by hand but couldn't. I took this to mean that it was siezed but I think it might simply be down to the internal gearing*.
* In fact after the rebuild I can turn the started by hand but it does take quite a bit of effort. In any event I also connected the battery straight across the motor (positive to terminal, negative to case) and achieved only a bright flash- no indication that the motor had any intention of turning so I was satisfied that the motor really is the cause of my problems this time around.
In the picture below the motor is positioned as it would be seen from the front of the bike and left or right in what follows refers to thsi view. The mounting points are on the left and the gearing/drive section (which rotates the motor) is on the right. The motor consists of a central section containing permanent magnets through which the motor armature passes. The armature doesn't extend right across the motor; as shown here its lh end is supported by an oilite bush in the lh endcap (which also contains the brushgear and motor mounting lugs). It s right hand end passes through a second bush fixed in an internal plate and not visible until the case is opened. The end of the armature projects beyond this plate to engage with the internal gearing system which in turn transmits drive to the final drive shaft contained with its own bearing in the second endcap (on the right as shown below). The drive is then transmitted to the drive chain (and thus the motor) by means of a sprocket on the end of the final drive shaft. There are O rings between both endcaps and the body section and also on the final drive shaft to seal its entry into the motor crankcase.
Stripping the motor
Remove the two long bolts that hold the unit together (8mm )![]() |
| Removing the through bolts |
![]() |
| Anti-rotation pin in groove, Thick grease surrounding the gears. Note splined end of armature in centre. |
![]() |
| Solenoid terminal. Note O ring between end cap and motor body |
![]() |
| Both terminal nuts and three insulation washers, the widest goes immediately below the bottom terminal nut. |
![]() |
| Lhs end cap removed, brushgear visible inside along with much carbon dust. Commutator now revealed. Note O ring on the central section |
Press the terminal inwards and you can pop the brush plate out as the stud goes in.
![]() |
| Easing out the brush plate |
![]() |
| Brush plate as removed |
![]() |
| RHS end cap cleaned up with oilite bush in centre. |
![]() |
| Armature as released. The copper commutator contacts the brushes in the rhs end cap and the splined gear at the other end drives the planetary gear mechanism to rotate the motor's final drive shaft |
![]() |
| Removing the ring gear, planetaries are off already. |
![]() |
| Armature shaft- the three spacer or thrust washers are shown. Don't loose these. |
![]() |
| RHS end cap supported on a steel cup whilst the splined end of the final drive shaft is pressed downwards |
![]() |
| Press in action |
![]() |
| Oil seal popped out. Note the bearing in the centre retained by a wire clip. |
![]() |
| Clip removed, note open side of bearing was positioned towards the oil seal... from where no oil was ever going to come! |
![]() |
| Pressing out the bearing |
Bearing safely removed.
![]() |
| Success, bearing out. |
This completes the disassembly of the motor, I have to say that I'm disappointed. I didn't really find anything that could have accounted for the failure to start. Certainly the bearing was in wrongly and did run a little rough, but this is no explanation. The best I can suggest is that the smudged excessive carbon deposited across the armature was shorting out the internal segments or jamming the armature's rotation. I was encouraged that others have apparently found that often a clean is all thats necessary. Anyway I cleaned everything thoroughly, and just to be sure that I hadn't got a dud armature I checked the electrical conductivities as suggested in the Haynes manual. These showed
1. Continuity between all commutator segments and all other commutator segments.
2. No continuity between commutator segments and armature shaft or metal plate stack.
3. Positive post brush connected to positive post and no continuity with brush base plate.
4. Negative brush in good contact with brush base plate.
I found that all commutator segments were indeed connected to all other commutator segments with a resistance of app. 0.3ohm between side by side or opposite segments. Given this, maybe all I needed was to clean it up? In any event its worth changing the final drive bearing, although the oilite bushes look fine. I just soaked them in engine oil for 48rs in an optimistic attempt to recharge their oil content.
I cleaned everything prior to rebuilding, but here I got my first shock. The kit supplied (SMU9115) was clearly wrong. The bearing included was obviously too small. This motor requires a 6003Z bearing which is relatively easily obtained. As I still wanted to use the rest of the kit and needed to order an oil seal anyway (surprisingly not included even though you can't change the bearing without changing the oil seal) I went ahead and ordered both to supplement the kit. These are 17, 35, 10 double shielded bearing 6003 and 20, 35, 7 nitrile oilseal with garter spring.
![]() |
| Old and new bearing. That removed from my starter is at the top, the smaller bearing included in the kit is below. |
Rebuilding the starter motor
Firstly clean the commutator with scotchbrite pad- don't use emery as the grit will stay in the workings and gradually wear things away.![]() |
| Commutator cleaned |
![]() |
| Orientation marks on mounting end cap, linear moldings and small semicircular indentations |
![]() |
| Orientation marks on gear end cap, striations and larger squarer indentations |
Feed the stud through the cap making sure that it seats correctly so that the rubber bushing surrounds the contact. This is all that stops your motor shorting out directly to the case. I would have expected something a little more positive like a nylon top-hat washer or equivalent, this does seem a flimsy feature for such an important function.
![]() |
| Insulation collar pushed up into cap and surrounding the stud. |
![]() |
| New brush plate showing the order of assembly of the various washers. Note the extra O ring . |
![]() |
| Rubber O ring added above collar on the positive stud. |
The motor body is sealed by 2 O rings, in the original Kawasaki system these are different thicknesses, but in the kit they are identical and both are 1.9/1.7 sq section O rings. I fitted a ring to the brush cap, inserted the armature and then fitted the body section, noting the alignment. Check that the smaller thrust washers/shims are in place on the armature shaft before fitting it into the endcap; you need to lift the armature up towards the brushes as you insert it so that they are pushed back into their housings and can slip over the commutator.
![]() |
| Brush cap back in place |
![]() |
| Adding the three thrust washers or shims. |
... and follow that this with the internal plate and oilite bush. Note the orientation pip that must fit into the cutout
![]() |
| Internal support plate with bush added, note orientation pip at 1 O'clock. |
![]() |
| Bearing and clip in place. ote that this bearing is double sealed and doesn't need lubrication |
You can then press the drive shaft into the bearing supporting it with a suitable steel tube bearing on the inner race only. This is a bit awkward. You can use a deep socket as long as its not too chamfered on the inside. You need to press it in only as far as is necessary to expose the circlip groove so keep checking as you go or the shaft will nip up too tightly to turn easily. You can feed the circlip down the shaft...
... and clip it home in the groove.
Turning the end cap over I inserted the counter-rotation pin in its cut out (appx 10 O' clock in the picture below) and dropped in the planetary gears.
The ring gear can then be orientated and slipped home over the pin, meshing with the internal gears.
Not shown here, the next step is to add grease to the gearing (I used a couple of teaspoons-worth of Castrol LM), add the O ring to the body and refit the gearing end cap. Note hold the cap with gears-facing upwards and feed the motor body with O ring downwards onto it, or the gears and pin will all drop out in a greasy mess! Rotate the cap to align the orientation marks and mesh the armature gear as you fit it. Then refit the two long retaining bolts and put a new O ring in the drive shaft end groove.
![]() |
| Bolts refitted, new O ring |
With great trepidation I tried connecting the motor. It didn't exactly burst into life, I suspect because the brushes weren't properly seated but after an abortive twitch or two it span smoothly and reproducibly. I'll go ahead and refit it to the bike, hoping that this has cured the problem.













































