KTM 640 - engine randomly dies

Power has to originate from the ignition switch.

We need to find the correct wiring diagram, then you take it to the copier and enlarge it. I think a dealer or KTM USA technical will help you.

http://www.ktm.com/us/contact.html

Once you get the wiring diagram enlarge it.

The plug socket size is 18mm.

Do you have any jumper wires?
Better make some around 3 feet long with alligator clips on either end.

Buy a good roll of 3M Scotch super 33+.
Buy some dielectric or silicone grease.

http://www.amazon.com/3M-Scotch-Electrical-75-Inch-66-Feet/dp/B00004WCCL

And a small pair of sharp scissors to cut the tape(don't tear it!).

The scissors go in your tool kit. A small bag or old sock holds the tale, jumpers and multimeter. Grease goes to the tool kit

You are looking like a motorcycle DC electrician.

And smile, you are among the elite.

Reread everything Gunner wrote.

Unscrew the plug cap and cut off 3/8" off the wire end, re-screw the cap.

Check and grease every electrical connection especially on the wires leaving the ignition switch.

Check, plug and unplug every connection on the rest of the motorcycle and grease with dialectic grease.

Bring the motorcycle up the steps into the dining room. It's warmer in there...oops...you have to run it, darn!!!

Don't give up, we have all been there.

bill

http://www.advrider.com/forums/showthread.php?t=169958. Look at post #7

Here is a link. It is about the ignition switch. This is a common problem caused by vibration. I have had 5 640As, and two of them had this happen! Look at post #7

Have the coil off the bike, here are the readings:

setting
200 ::: blue-white - ground ::: ~.16 - 2.4 (bounces around)
20k ::: blue-white - ignition ::: ~ 12.0

If I'm doing the first test correctly, the coil appears to be out of spec.

Wow - ok, I just pulled the bottom half of the switch. The top half is set in with what look like screws that aren't designed to come out after installation.

Going to try and take the cover off the bottom half of the switch next.

EDIT
Got the bottom half of the switch off. No broken or loose wires.

Mine wore out up above in the contact area, really dirty to the point of no contact.

A simple cleaning and all was well again.
bill

Is it possible to pull this down further?

Slightly off topic, here's a similar thread where some inmates suggested a bad coil.. the other 640 owner had same symptoms and resistance reading on his/her coil.

http://www.advrider.com/forums/showthread.php?t=816258

It's possible, though I don't know why the resistance would jump around. Try to scratch the contact points of the multimeter leads so you get a better reading. Might be jumping around due to corrosion where you're touching your leads.

Assuming it's clearly an ignition problem (no spark from plug lead if you crank) ...

My 640 had a sort of similar irksome intermittent fault -- ran fine, then wouldn't start. After fiddling, or just being left along, started just fine.

Turned out the pulser coil was packing up. relatively easy to check, follow the wires from the flywheel side up to the connector, and check the resistance. Off the top of my head it should be around .3 ohms. Can confirm, but if it's going it'll prob be very high.

I also had an intermittent fail caused by a broken solder joint in the ignition switch (although you seem to have ruled that out...)

A quick bush-mechanic test of the coil is to put 12V across the primary off and on (blue wires), with the plug lead near (cm or so) a good earth ... you should get a big spark when you break the connection. Keep fingers clear of painful shock.

Quick update this AM - I tested the coil with a different multimeter (a digital unit from Radio Shack), and this time I get:

blue-white - ground ::: ~.6 (steady)
blue-white - ignition ::: ~ 13.0

The primary is still a smidge out of spec if this reading is exact and trusted. Then again, my meter may be rounding up from 5.75 to 6. Not sure that small a difference would be causing the bike to cut off. Gonna do the "bush-mechanic test" D0dja recommended to see if the coil is capable of producing a healthy spark from a 12v DC source.


About the switch.
Is it safe to put dielectric grease in the switch (the side with the contacts)? Do I just dab a little on each contact or swab it all around?

EDIT:
According to the Wikipedia section for dielectric grease, I should *not* put this on the actual contacts

This explains SO MUCH.

Spades,
Did you break into the contact portion of the switch, that's where mine failed. It was full of dirt/dark gunk. I cleaned all the contacts, then a very light coat of dialectric, reassembled.

Been working fine since.

Thanks for sharing the Wiki info, I'll be more careful with it. For me the issue with connections is corrosion and it has to be prevented. I have never had an oily connection fail and often just put a drop of oil on mine and give them a wiggle. I don't say much about that technique because it is unusual.

I have had to repair the ignition switch on every motorcycle I ever owned.

Earlier I wrote that power originates from the ignition switch, I should have called it switched power.
FWIW.

bill

I pulled some tape off the main harness to trace the blue/white off the coil. It did a U turn and went back to what looks like the CDI (on the right side, behind the side panel). Since my switch was apart and inside the house, I couldn't do anything more than check resistance on that piece of wire - it checked out fine 0.0 ohms.

EDIT:
I recall reading somewhere that switched power on the KTM is indicated by a yellow/blue wire, and "always on" power is carried in yellow/red wire.

Have you made any progress on this? I just went out and checked my bike, a 2002 model ADV, and I get a constant 45 or so volts at the coil (with blue/white wire disconnected from coil) with spikes of 150 volts or so as the motor cranks. (you might not see the 150 volt spikes as I have a DVA meter which is designed to read momentary voltage spikes, multimeters can't react fast enough to register a spike)

With the blue/white wire connected to the coil and getting the MM lead into the connection to get a reading I get 0 volts with spikes of 200 volts while cranking.

YMMV..


...also means the bush mechanic 12volt test mentioned above probably won't provide spark.

Update - got the switch back together last week with a light coat of di-electric grease on the internal connector plates (after first lightly cleaning them with a metal scotch brite pad).


Findings tonight:

Before turning the bike over, I connected my multimeter red terminal to the blue/white female connector, and the mm black terminal to the battery NEG post. I read 11.x volts, whether the key was on or not. (?)

I reconnected the blue/white connector at the coil, then started the bike up - it ran for maybe 10 seconds before dying. Confirmed no spark by using a timing induction light.

Disconnected blue/white at coil, and repeated first test above. This time it read ~50 mv or so. I also noticed a distinct whine coming from the area of the CDI (on the left side near the battery). This whine was not audible when I performed the first test.

Is this whining noise normal? Could it be related to the drop in voltage at the blue/white wire?

Cross referencing a similar thread:

KTM LC4 no spark? HELP?

If you've got 12volts at the coil then our bikes have different ignition systems.


The whining noise sounds like "flyback" which Luke describes in the thread you linked as "very faint whining sound like a tube TV"

I see you responded to that thread, but the OP of the thread hasn't been back to ADV since 12-23-2010, so you might not get any reply.


Seems like Luke's reply is a good for some areas to check.

http://www.advrider.com/forums/showpost.php?p=14764451&postcount=8

I remember seeing a red or orange looking wire going into the CDI. Tried to trace it back just by observation and couldn't find the other end. It was cold in the shop and fresh-baked chocolate cookies were calling.

I have a wiring diagram but I noticed some inconsistencies with my harness and what's on the diagram. I don't know if it's allowable to post that diagram - read the concerns from the mods about posting service manuals.

Cross referencing another interesting thread, similar problem as mine.

From 2006: LC4 CDI issues?

In reply #9, bigtators wrote,

Another cross-reference, a post by CountPackMan also from '06:

You have to be careful when testing wires around the CDI, it is not unheard to fry the thing and they are quite expensive to replace...

A few pages back I mentioned for you to visually inspect the stator and pickup coil, have you done that? I have seen more than a few times where the stator/ pickup running surfaces were scored up and gave intermittent electrical problems similar to what you have been describing...

It is very easy to remove the stator/cover { 4 bolts} and if your careful you may get away with reusing the gasket... You'll have to remove the counter sprocket cover as well...

Another clue by CountPacMan: