Ignition problem, intermittent, iBeat logs of good and bad runs

Thanks Seymore, I'll check out the supposed shielded cable. I've swapped stator/pickup coil and didn't notice it being shielded in any way besides the plastic. Are you sure the diagram is representing an electrical shield or the plastic protective 'shield'? On second look the lighting coil wires that go directly to the regulator don't show the 'shield' yet they share the same plastic shield as the pickup coil wires. I'll investigate.

If I'm unable to identify the culprit using my 'donor' cycle parts, I'll definitely get in touch with the national tech rep.

Not yet, I'm a working man so it will have to wait until Saturday.

I did check out the leads from the pickup stator, there is no electrical shielding whatsoever, on either my 08 or my buddies 09.

Ok guys, more to report. I've gone back to my ECU and stator, confirmed the problem still exists. I also replaced the connector for the pickup coil to eliminate it. I've swapped fuel tanks to rule out the fuel pump.

During my numerous runs to test, I've discovered that the problem only manifests itself when I use the e-start. If I bump start, no problem, consistently.

Riddle me that one Batman.

Next I'll swap the throttle body to see if the MAQS is the culprit.

What type battery are you using?
Does e-start sound normal or slow?
Have you tried replacing the battery?
Did you check all the battery cable connections on both end?

A battery with a bad or intermittent cell or connection could cause the problem even though the data log shows normal voltage. The effect from the high current draw during starting might cause a unusual voltage fluctuation that could confuse the ECU.

Just some wild guesses to your odd riddle.

This is like CSI for motorcycles.

Wonder who the killer will turn out to be?

Is the problem consistent with the use of e-start, or do you occasionally run with no problems?

Currently the only electrical component that has yet to be swapped is the flywheel, the dash and wiring harness. Couldn't find a puller in town, headed up to see George Sunday.

Sometimes it will run OK with use of e-start but less and less frequently. So far, every bump and kick start has run good. We'll see if the flywheel is the culprit, if not, then it must be a combination of parts...but it's getting tiresome all this swapping then a few runs around the neighborhood...

Gordon, I'll be at George's about 9:30am or so Sunday, if you want to check out the PC V + AT.

I'll try and be there. You can ride my bike and see what you think too.

Resolved problem, thank you Seymore

Seymore, thank you for causing me to think about the battery again. I had replaced the battery with a spare and then the new LiFePO4 early on in my troubleshooting last month or the month before and found neither made any difference.

I figured in for a penny in for pound, I might as well try a 4th battery, the one from my buddies donor cycle, since I've just about run out of parts to swap!

What do you know, I'm unable to reproduce the problem. I rode and stopped and started the engine using the estart 15 times, no problems.

My mistake was assuming two things, that my spare and my newly built LiFePO4 were good or still good, particularly the LiFePO4 tested very well under load when new but since the problem remained, I swapped back to my spare and when this didn't fix the problem I went back to the LIfePO4 battery. Well now the LifePO4 battery sags horribly, down to 9V, under load, it seems to spin the motor just as fast as my spare, but certainly not as fast as my buddy's battery.

The low voltage sag caused by the starter motor must have been just below the tolerance for the ECU when it acquired the signal from all the sensors, screwing everything up.

When bump started the sag does not occur of course, hence all is well.

I will buy a new battery! I also reconnected the PC V + AT and they are functioning normally too.

I don't believe there is any need to swap the flywheel now, I was struggling to believe that could be the problem but I had thought I had run out of options. I certainly know the bike a bunch better than when this all started. Thanks for all the ideas along the way.

And everyone should note, the HT coil should measure 20k ohms, not 5k ohms like the service manual states.

Awesome news Brett!

I want to commend Brett for methodically diagnosing his bike step by step without the premature brand bashing that typically goes on!

PS I hope you can make it up to the Pacific NW next year for a husky ride...that goes for everyone else too!

David

That's right. I tried the two local shops for a flywheel puller today, Ron Bishop had one that was close but no cigar and Moto Forza didn't have one. George said come up in the morning and use his puller, so I had nothing to lose by trying my donor cycle's battery this afternoon. 4th battery was the charm.

I somewhat surprised the ECU doesn't through a code or error when it encounters a startup voltage that is clearly too low for it's liking. Sure would be easy to set a flag if voltage drops below a threshold.

The 09 ECU definitely behaved differently but not right, better than the 08, maybe it's more tolerant of borderline low voltage.

Now I get to ride tomorrow instead of wrench and troubleshoot.

Glad to hear the good news and I'm happy to have have been of help. It was great to see iBeat used to help troubleshoot a problem. I don't think you could have afforded the cost of bike shop going though this learning curve at your expense. This was a good learning event. Thanks for sharing the details.

If I remember correctly you built you battery yourself from individual cells. Do you know what the AHr rating is and what the cranking amps or battery impedance is?

I installed a 2.5 AHr TurnTech LiFePO4 battery and found it was very marginal at starting the 450 especially when cold. The problem wasn't so much with the starter RPM as it was the bike acted like EFI was have trouble supplying fuel to start the bike quickly. I order a 5AHr battery but haven't had the opportunity to test it, I have the stock battery installed now. It maybe that the smaller LiFePO4 batteries are not suitable for Husqvarna's EFI.

Husky's EFI is different than Suzuki, Kawasaki Honda and probably Yamaha. Husky powers everything from the battery and uses a low pole count magneto or alternator to charge the battery after the engine has some high RPM. The other guys use a high pole count magneto (more efficient at low speed) to quickly charge a capacitor during the kick start to power the EFI system.

I post long ago on TT, if your battery fails a long ways from civilization, your screwed unless your buddy on his KTM is nice enough to give you his battery. (tell them they'll ride better with less weight)

Because of this I did developed a compact capacitor device that weighs about 3 ounces that I carry with me in case I have a battery failure. With the device installed, you kick start the bike about 15 times with the ignition off, then turn on the ignition and quickly kick start the bike within 5 seconds. There is enough power stored to run the ECU and fuel pump for 8-10s before the engine fires.

I originally told myself I wouldn't buy an EFI dirt bike unless it could be started with a dead or damaged battery, but just like a pretty women (people), you over look the blatant flaw and give go for it anyway.

Great to hear things are sorted out! This has been quite a learning session for me. I'm sure I'll own another EFI bike eventually, but until then... Ahhh the elegant simplicity of the dead-nuts reliable dinosaur carburetor.

I did build a battery from cells. It's clear that testing under load that one of the cells has largely given up, the pack sags to ~9V. The chemistry is better than SLA regarding voltage drop, so with healthy cells there should not be a problem. Here are the specs for the cells I used:

http://www.batteryspace.com/prod-specs/4203.pdf

Given the relative cost and country of origin I should have suspected a cell gone bad earlier.

My bike came with a Yuasa YTX5L-BS (from shop owners 'race bike', not new) as they didn't have any YTZ7S charged. So 'normal' starting speed for my bike was based on the smaller YTX5L-BS. I think I've been running 'marginal' for awhile now. The LiFePO4 may have very well started life with a marginal cell, or I inadvertently damaged one while soldering. Because of the mini how-to I wrote, the company gave me a gift cert for $50, so I can build another for essentially just the cost of shipping, $8. I've already ordered the parts.

Keep in mind, this problem I've been chasing was not consistent until just this last couple of weekends when I got serious about testing and finding the culprit.

Clearly the (my) 08 ECU is very sensitive to low voltage, even voltage that is plenty to start the bike 'normally'. This is a bit of a warning system though, if the bike runs badly after e-start, I'll use the kick/bump, if good then I'll suspect the battery is not healthy. Even a 'bad' battery should have enough float voltage to power the ECU for normal start, I would worry the fuel pump might not produce enough pressure though. The point, the bike should kick start with a very very sad battery, just don't try to e-start. The capacitor is a great idea, or carry a spare, very small 1.5a/hr LifePo4 pack that weighs ~6oz. It may not be enough to use the e-start, but it will power the ECU and fuel pump.

In my profession, I frequently perform root cause analysis of computer networking issues, dealing with far more variables than a single-cylinder bike. I enjoy the process and do not relish paying someone else to 'fix' things. Not only because of the money, but I lose out on the experience and knowledge gained.

Thanks again to all that provided suggestions, hopefully collectively we have saved someone else troubleshooting time in the future.

I'm not convinced the FI system was all that involved here...the ECU starting up with borderline low voltage caused it to misinterpret the crank position and because of this be all mixed up causing the other systems to send/receive bad info. Even a carb bike won't run properly if the ECU thinks the ignition should be fired during the wrong point in the stroke, never mind what is happening with fuel.

That said, there sure are a bunch of things that must go just right for the complete FI system to work. Too many switches, connections, wires in my opinion. It could be simplified and thus made more reliable. One simple change would be to reverse the kill switch, currently it must make contact for the bike to run, opening the circuit kills the ignition and fuel pump. It's vastly easier, practically guaranteed, to maintain an open circuit than a closed one.

And good God, technical writers at Husqvarna, please correct the errors in the manual!

I think I've beat this one enough, I'm going riding now.