Starter issues

[wfooshee]

It took some digging, but I found my old Sears engine analyzer. Basically a meter with a bunch of switches to check tach, dwell, volt, and amps. It has a 160-amp scale, so I hooked it up.

Bike is cold, cranked and started perfectly. Pulled negative cable on the battery, pulled positive cable and installed the ammeter shunt on the positive, reinstalled the negative. Turned the bike on, saw a couple of amps positive, meter connection does not need to be reversed.

Hit the starter button, instantly shows 120 amps, I get a couple of turns from the starter, then it lugs and the ammeter pegs. Turn the key off, the ammeter shunt is very hot to the touch. Put the battery cables back to normal, bike fires right up.

It seems my ammeter kit is adding to the circuit, significantly. I think the shunt strip is aluminum. Granted, the kit is designed for measuring alternator output, but with the 160-amp scale and the numbers in Rad's post, I figured it had the guts for this.

Just for funsies I reversed the shunt and used it on the negative side, same result, 120 amps briefly, then the starter lugs and the meter pegs, shunt is very hot to the touch.

My test equipment may not be up to the job, but it indicates to me that I've got a huge draw from the starter, at least twice what it ought to be. I've been in touch with FJRnut from post #58 and he's sending his spare starter to test with. I'm going to wait for it to get here before I tear the fuel system out again, so I haven't done a no-load test on my starter. Stupidly, that never occurred to me while I had it out. :blink:

BTW, voltage drop at the battery is about the same, hot or cold. I see 12.8 or so when I turn the bike on, and into the eights and nines while cranking. Saw 7.9 once, cold. Still have not had the clock reset during any of this.

 

[FJRnut]

My test equipment may not be up to the job, but it indicates to me that I've got a huge draw from the starter, at least twice what it ought to be. I've been in touch with FJRnut from post #58 and he's sending his spare starter to test with. I'm going to wait for it to get here before I tear the fuel system out again, so I haven't done a no-load test on my starter. Stupidly, that never occurred to me while I had it out. :blink:

wfooshee, my lazy ass didn't get to the UPS store until yesterday- and they were closed. I promise I will get there by 3:30 tomorrow (monday). I apologize for being a lazy ass, but the price is right, right? If you want me to get it there really quick, I'll take a promise of 'the check's in the mail', otherwise it'll probably take a few days to get there.

How much of a pain is it to remove the fuel system?

 

[wfooshee]

No sweat. Can't really work on it till the weekend, anyway, as it's my daily driver and I have nothing else but feet to get me to work.

As to the fuel system, it's rather involved. Fuel tank has to come off, not just propped up but off the bike. 2 vent hoses, 2 fuel hoses, 2 electrical connectors. Tool tray comes out so you can get the air box out. Fuel rail comes off to give clearance for the airbox tubes. That's 4 electrical connectors and a vacuum tube. Air box has one sensor plug and three hoses, one of which is nearly impossible to reach underneath. Then the throttle bodies come off, which is another couple of sensor plugs and two coolant hoses to the fast idle assembly underneath, and the throttle cables, which are easier if you go ahead and remove them from the handlebar grip as well. Now that all of that is out of the way, the starter is easy. :huh: One electrical connection and two mounting bolts. That's why I didn't pull the starter today for a no-load test. B)

 

[Constant Mesh]

It's possible that the magnet problem you found may have affected the motor speed and input current. If the magnetic field is weaker than normal the running speed would increase and the input current would increase.

 

[Zorlac]

Locked rotor current is the most current a motor can draw, counter EMF in the armature increases with motor speed, therefore supply current drops with an increase in motor speed.

I think this starter is bad, period, especially since it was drawing 160A+ across the test shunt.

 

[Fred W]

Locked rotor current is the most current a motor can draw, counter EMF in the armature increases with motor speed, therefore supply current drops with an increase in motor speed.I think this starter is bad, period, especially since it was drawing 160A+ across the test shunt.

I agree, sort of. If we can trust the current reading then that is high. The only way you would draw too much current is with a faulty load device.

But... The added resistance of the shunt was apparently enough to put the starter into a mode where it would not turn over the engine. In that kind of state, where the motor is straining to turn over, it would be drawing larger than normal amounts of current. So, it's still a "chicken or the egg" question.

 

[Zorlac]

" The added resistance of the shunt was apparently enough to put the starter into a mode where it would not turn over the engine. In that kind of state, where the motor is straining to turn over, it would be drawing larger than normal amounts of current. So, it's still a "chicken or the egg" question."

So that explains why it acts the same way when the shunt isn't in series with it?

 

[Fred W]

" The added resistance of the shunt was apparently enough to put the starter into a mode where it would not turn over the engine. In that kind of state, where the motor is straining to turn over, it would be drawing larger than normal amounts of current. So, it's still a "chicken or the egg" question."
So that explains why it acts the same way when the shunt isn't in series with it?

No, he said when he removed the shunt it immediately turned over fine. This was with a cold engine.

It's only with a warm (hot) engine that the starter lugs. So you see, it doesn't explain anything...

 

[Zorlac]

" The added resistance of the shunt was apparently enough to put the starter into a mode where it would not turn over the engine. In that kind of state, where the motor is straining to turn over, it would be drawing larger than normal amounts of current. So, it's still a "chicken or the egg" question."
So that explains why it acts the same way when the shunt isn't in series with it?

No, he said when he removed the shunt it immediately turned over fine. This was with a cold engine.

It's only with a warm (hot) engine that the starter lugs. So you see, it doesn't explain anything...

I missed the cold engine part, but I still think taking the starter to a starter shop for a test could answer a lot of questions. They might even have an oven to warm it up before the test.Shorting turns on the rotor when it's warm might be what he's up against.

 

[wfooshee]

. . . but I still think taking the starter to a starter shop for a test could answer a lot of questions. They might even have an oven to warm it up before the test.

Agreed, but the shops are only open during the week, and I need the bike to get to work, it's my only transportation. When the spare arrives I'll swap, then look into test and repair of this one.

 

[swingman3]

I've been following this thread with great interest as I have had very similar experiences. When the bike is shut off while hot, it will often not start on the first try. The clock/odo will reset, and I'm pissed because I've lost my mpg calculations. If I hit the starter a couple of seconds later, it will fire right up.

I'm trying to follow your technical discussions although the troubleshooting techniques you posted have thoroughly confused me, and there still is no real answer. So may I add my thoroughly uneducated, but gut feeling idea?

My thought is that this is mechanical. When I shut the bike off, then turn the key back on to open the glove box or something, I notice that the temp gauge has actually climbed. I've attributed this to there no longer being any air or cooling water flowing through the radiator. So what if we shut the bike down, and the engine actually increases in temp for a minute or so. Enough to tighten the cylinders inside the pistons... then a few minutes later (after we're done fueling), we hit the button, the starter hits up against the now very hot and tight cylinders, and can't handle the load. It draws too much current, killing the clock and the odo. (This is the part of the theory where I run into my black hole of electrical knowledge, but you get the idea.)

Continuing on... Now, because of the force of the starter bumping the cylinders, they are broke free enough to allow the engine to start on the next try, even though they are still hot. This would explain why the bike will start right after the first failure, and also why it will start after the bike cools a little (say, after 20 minutes or so).

Does this make sense?

Swingman Sends

 

[swingman3]

Ok, I just spoke to an engineer (military-speak for a mechanic) friend of mine. Here's his thought. He says it sounds like vapor-lock. There was an issue with some Coast Guard boats a while back where they would do the EXACT SAME THING as what is happening here. They traced it to some fuel dribbling from the injector just as the the piston was starting it's compression stroke and the engine was shut down. The fuel that was in the cylinder would hit the hot steel, vaporize and expand. When the engine was started while it was still very hot, the starter didn't have enough ass to turn the engine over against the vapor in the cylinder. Later, when the engine cooled a bit, the vaporized fuel would condense back to liquid and reduce the pressure in the cylinder, allowing it to turn more easily. He said they fixed it with new injectors.

This is similar to the theory that I had (that the problem was mechanical in nature, not electrical), but the cause is different. So how about that? Any thoughts?

Swingman Sends

 

[MotorToad]

That sounds fishily like what it sounds like when it happens. It really sounds like the starter's doing its best and can't get the thing to spin.

wfooshee: let's pull the exhaust valves out and scuff them up a bit!

 

[Fred W]

Interesting theory. Sounds like we need a compression release...

 

[RadioHowie]

It is an interesting theory.....that the starter can't turn the crank because of some vapor in the cylinders, at least three of whch are NOT in compression mode....yet it can certainly hammer a set of sodium filled steel valves into an aluminum cylinder head.

I'm not buyin'.

 

[FJRnut]

. . . but I still think taking the starter to a starter shop for a test could answer a lot of questions. They might even have an oven to warm it up before the test.

Agreed, but the shops are only open during the week, and I need the bike to get to work, it's my only transportation. When the spare arrives I'll swap, then look into test and repair of this one.

Just a thought to save time and or hassle. Take my starter to your starter store. Have them check it out BEFORE you spend time R/R something on my word. I did take it to my starter rebuilding place, but like I said, they didn't heat it up.

At minimun, check them out side by side and compare ohms and all that good stuff.

Do you have a local place that can make new cables? I want to get bigger cables. Not huge, but bigger. Doesn't heat make more resistance in cables? Tucked behind the cylinders, it's very hot there. Being small I'd think they have the ability to cool rapidly, to start the motor. Does this sound reasonable to anyone else?

 

[wfooshee]

Ok, I just spoke to an engineer (military-speak for a mechanic) friend of mine. Here's his thought. He says it sounds like vapor-lock. There was an issue with some Coast Guard boats a while back where they would do the EXACT SAME THING as what is happening here. They traced it to some fuel dribbling from the injector just as the the piston was starting it's compression stroke and the engine was shut down. The fuel that was in the cylinder would hit the hot steel, vaporize and expand. When the engine was started while it was still very hot, the starter didn't have enough ass to turn the engine over against the vapor in the cylinder. Later, when the engine cooled a bit, the vaporized fuel would condense back to liquid and reduce the pressure in the cylinder, allowing it to turn more easily. He said they fixed it with new injectors.
This is similar to the theory that I had (that the problem was mechanical in nature, not electrical), but the cause is different. So how about that? Any thoughts?

Swingman Sends

I'm wondering if something along those lines may be what I'm up against, possibly. I know I have a rich cylinder, one plug fouling. Is it rich because the injector won't shut off? While the stuff was apart I swapped a couple of injector positions to see if the fouling follows the injector.

It would be interesting to see the crankshaft position while starting, see if it hangs at the same cylnder on its compression stroke every time.

 

[swingman3]

The way it only happens when the engine is very hot, it feels very mechanical to me. It also makes sense to me because the bike ALWAYS starts right up after the initial fail. That doesn't seem to be electrical. It seems that something (maybe vapor in a closed cylinder) stopped the starter in its tracks. Then, because the engine did turn a little, either a valve opened, or it cooled enough to reduce the pressure- or a combination of both- it finally turned over and fired up. I'm no wrench and I don't know what the answer is, but there seem to be more than a few people with the same issue and no one seems to know. I applaud your troubleshooting work and everyone who has contributed ideas. You all seem smarter than the average bear wrt the electrical systems on this bike.

Speaking of a compression release, I know that the portable fire pumps we use on our cutters have exactly that to release the compression in the cylinder before yanking on the starter and pulling your shoulder out of its socket.

Like I said, I don't really know. I was just throwing another possibility out there.

Swingman Sends

 

[wfooshee]

No it doesn't start right up after the initial fail. It starts right up after sitting long enough, however long it feels that needs to be. On a sunny afternoon that's longer than a cool evening.

I'm continuing my thought experiment on that, though. It ALWAYS fails to start when hot, cranks poorly and lugs to a stop. Kicking the button might get another turn out of it, and it'll hang again.

I would think it would stop by chance about half the time where the cylinder with a drippy injector is ready for the exhaust stroke. (Actually about 3/4 of the time not on that cylinder's compression stroke, but on one of the others.) Since the guilty cylinder would have exhaust coming up, it would shove accumulated vapor out the back, alleviating the problem, and allowing good cranking and good starting. That never happens.

Just thinking out loud, not realy accomplishing anything useful.

 

[Fred W]

I think we already have the evidence that shoots a hole in the drippy injector theory:

Once the bike is running and has warmed fully up to temperature, if you stop the engine and then immediately try to start it again, it does not turn over fast enough, correct? At least that is what I have inferred your testing protocol was thus far.

Since this is before the alleged leaky injector has had any time to do its drippy deed, that rules that theory out.

Another thing that occurred to me last night; What is the resistance of the meter shunt that you used to attempt to measure the current?

Usually ammeter shunts are very low resistances, usually a fraction of an ohm, so as to upset the circuit under test as little as possible, but of a very precise tolerance. My point being that adding just that tiny amount of resistance in series with the motor current mimics the failure mode, even when cranking it stone cold. So your problem may very well be something in the starter circuit increasing in resistance when heated up by as little as that much.

Or, getting back to the chicken and egg, it may be completely normal for the starter circuit resistance to increase by that very small amount when hot and your battery just cannot supply enough current, or the starter is too wimpy to turn with the restricted current.