Starter not disengaging

[gabby]

Thought I might throw this question out there before I pull anything apart. I have a bike with a problem in that the elec starter tends to stay engaged for a while after the engine starts and the start buttom is released. There is no problem with the starter motor or electrical circuit.
The starter bendix gears and spring have been replaced with new, in fact the spring was replaced with a different after-market beast a second time. The usual problem with these is that they won't engage properly, but mine engages ok and starts the bike, but when the button is then released, it grinds fairly loudly for at least a second or even more. It doesn't sound like it's going to disintegrate, more an embarassing loud noise that attracts attention.
Actually I have two XSs that do it, and both have been rebuilt with new gears and spring and all looks ok. It's not a show-stopper, but might be good to know if it's been experienced by others. Thanks

[pamcopete]

gabby,

Do you still have the safety relays installed? The safety relay does a better job of releasing the starter relay than you can do manually.

[gabby]

Hi, yes it's fitted and seems to work. That's what makes it more puzzling. I'll try to test it against some others in case it's not being energised at the right voltage though; good point. Once the engine is spinning faster than the starter can turn it, I thought it would throw the starter gears out easily though.

[gabby]

Pamcopete, think this might be solved and I'm a bit embarassed I didn't consider this in the first place and was looking for something more complex. Thanks for pointing me in the right direction so I might post what I found in case others have this issue.
My starter lockout relay (ie safety relay) worked ok and prevented the starter engaging if the engine was running. My first mistake was to assume it was only there to stop dummies chewing teeth off the gears. Deliberate or not, it also helps disengage the starter quickly when the engine kicks into life. I took it out and did some bench testing to see what sort of voltage would pull it in. It was actuating at around 4.8V. I checked some others and found there was a huge variation in the voltage.
So with some experimentation when it was fitted to the bike, I found that if I adjusted that pull-in voltage to a lower value, it would disengage the starter much sooner. I now have it set around 3V and the gear grinding is almost insignificant. I may get it better yet but it's ok for now and doesn't wake the neighbours.
To adjust the voltage at which it operates, it's as simple as bending up the lower spring tab on the relay contact arm, so that the static spring tension is reduced. It could be tested with a variable power supply on the bench, or perhaps by trial and error on the bike. Thanks again.

[pamcopete]

gabby,

Well, I think you have broken new ground in the annals of the XS650! I just thought perhaps you had removed your safety relay, like so many do. Never thought about actually adjusting it for best results! Bravo!

The starter gear has to be going much slower than the engine in order for the engine to kick the starter gear back into place. If the starter motor continues to have power applied, even for a second, after the engine starts, then the starter motor speed will increase dramatically because it has no load on the gears, so the inertia of the starter motor and gears keeps the motor going even after the power has been removed, and the starter gear stays engaged because there is no difference in the gear speed vs the engine speed for a second or so.

The safety relay will remove power from the motor the instant that the engine speed is sufficient to generate an output from the alternator, so if adjusted properly, this will prevent the starter motor from increasing it's speed, thus facilitating the disengagement of the gear from the engine.

[xackley]

www.650motorcycles.com/SafetyRelay.jpg

The book says
yoke gap .008 in
core gap .020 in
point gap 0mm
cut out voltage 2.5V or less.

Question: is it the top or bottom unit as mounted on the bike

[xsleo]

If you don't have a variable power supply, hooking flashlight batteries in series will get you close to the right voltage. 1.5 volt battery x 2 = 3 volts.

[gabby]

Hi xackley. Looks like the safety relay shown in your link is different to mine. Mine is a similar looking thing to the old mechanical regulator, with a rounded metal cover. I've had a look in some electrical schematics and it appears US models are a little different and there may be a clutch swith in the starter lockout circuit. If that's right, then it's a different arrangement but the same principle. The relays in your picture looks sealed but if you have the information in a manual saying what the yoke gap etc should be, then it's likely you can get into it. If so, then there will be some way to mess with it but without seeing it I can't say, sorry.
It is interesting that the specs above say it should cut out at 2.5V or less. That would definitely explain why mine is much better now it operates at 3V. A bit more tweaking and it might be perfect.

[xsleo]

I have a 75 with the same problem. I pulled it off and brought it inside to test. Using two aaa batteries, soldered wires to them. hooked in series. Got 2.99 volts. Two 220 k resistors inline gave me 1.75 volts. one 2.20.
With this setup I could apply three different voltages.
With the spring off it would pull in with the 2.20 volts, with the tabs bent so the spring had enough tension to not not move around too much, it would pull in with the 2.99 volts but not the 2.20 volts.
I then put it back on the bike, it made a tiny bit of difference. I still have a 30+ year old return spring on the gears.
I'll have the right side cover off to hook in an exterior filter and cooler, and swap clutch parts around.
I'll try spreading the return spring to give it a bit more tension, if this helps I'll get a new spring.

[pamcopete]

xsleo,

Well, I actually removed my return spring on my '78/E about 10 years ago to prevent the other starter problem, grinding of the #4 gear. The theory here is that the return spring delays the gear extension just enough to cause it to not fully engage. I've inspected the gear every other year since then and there is no sign at all of grinding.

So, you might want to be careful that you don't exchange one problem for another, and of the two, grinding the #4 gear is worse.

[xsleo]

Maybe I'll just leave it the way it is.
It always engages, a bit of run on after it starts isn't that big a problem.
As far as it being a bit embarrassing to start in public, just a kick and she's running, is just the opposite, you get approving looks.

[gabby]

I'm with you on kicking it. Lots more fun and it amuses the new ones who have only ever used an electric foot.

[ronny]

OK, I'm new to all of this so pardon the ignorance. I have the opposite problem of the starter gear NOT engaging. Dose this mean the spring on the engagement gear is faulty???

[ShakerNorm]

Gabby - I have exactly the same problem as you have - so I'm going to pull mine and check it out.
Xacley - thanks for the numbers! I have access to a variable power supply at work, so this should be an easy adjustment!
Leo - which relay did you test? the upper or lower?
Ronny - the starter spring and gears are notorious for failing - so much that Mikes (and a couple others) sell a starter repair kit just to fix it! (yup - I've done mine!)

And yea - kicking it is guaranteed to get some looks of approval from (especially) the older guys ;D .... of course the young guys wonder what the heck you're doing ....
Of course that all goes out the window if the bike kicks YOU off.....

[ronny]

Ok, how do i get intouch with Mikes