Kawasaki KLR Forum banner

1 - 15 of 15 Posts

·
Registered
Joined
·
9 Posts
Discussion Starter · #1 ·
Please help, I'm not an experienced forum user so hope this works OK. Sorry, very long post coming up, but please try and stick it out. I’ve also read older posts with similar/same problem:

KLR650 C, Gen 1, 1999 – Australian Model (no tacho) – Random, intermittent misfire – Red Box CDI issue?:

Was working perfectly until it was stolen when my garage was broken into. Bike was recovered 2 days later with ignition barrel gone and evidence of trying to hotwire it. As OEM ign switch is $$$, I temporarily installed a made up one (from Schneider ZB4 series key switch and contacts) on a custom-made bracket until I could buy an online NON OEM switch. Once done, tried to start her but no spark whatsoever. From this I assumed they had damaged the CDI trying to hotwire it, so ordered an aftermarket Red Box type (made in Argentina) from Ricks Motorsport Electrics in USA. She started-up first go… fantastic!

So, went for a ride and she was pulling like a train, except I encountered a random misfire – it seems like a single misfire rather than a stutter or hesitation – it occurs with no particular condition, throttle position, speed, engine load, rpm, lean angle. It can occur 5 times in a km (0.6miles for USA), or maybe just once, but you can bet your life it occurs in the roundabout just when you don’t want it 😊. It also occurs when just idling in neutral on the stand, maybe once or twice every 5 minutes.

I eventually bought an aftermarket ignition switch – not quite exactly the same and had a bit of re-positioning of connector wires, but it did the job OK, but still had the random misfire exactly the same.

Checked the following (not necessarily in order) – and maybe a few more I’ve forgotten about:
  • Exciter coils – resistance, insulation resistance to ground (disconnected) at 250VDC, voltage output – about 250VAC. All was OK.
  • Trigger Coil – resistance, insulation resistance to ground, voltage output – about 3V to 4V pulses.
  • Visual Inspection – removed the crankcase cover and took off the rotor to inspect the coils – all looked OK. Checked gap between rotor magnet and pick-up – looks OK but not measured. Rotor key OK and crank bearings good – no play and no poling of the rotor.
  • Wiring from Exc/Trig Coils – all good no breaks or chafing, continuity tested OK whilst bending wires looking for breaks.
  • Diode Pack – all OK, also directly grounded the safety switch output from the CDI to eliminate this.
  • Harness – bypassed this by directly wiring the exciter coil output (and trigger coil) straight to the CDI. Earth was doubled-up and connected directly frame.
  • Ignition Coil – tested OK. But bought an aftermarket one anyway. Result was: engine pulled stronger on original coil, so put it back. Primary side earth connected direct to frame and cylinder head as well.
  • Spark Plug – checked gap OK and no fouling, but tried new plug anyway, still same misfire.
  • Ignition Lead – tested OK but replaced with a new bit of HT cable anyway just for testing.
  • CDI Replaced (again) – Ricks Motor Sport Electrics kindly replaced the first CDI I bought; the replacement made no difference – same random misfire.
I can visibly see the ignition misfire with a spark tester in line with the plug whilst idling. It occurs more often than heard/felt because is occurring on the wasted spark cycle (exhaust stroke) as well.

I am an electrical engineer and have so far failed to identify the random misfire. I also have had a low-end digital storage scope on it, but need more than 2 channels, and don't have access to and can't afford to buy. Ideally a 4Ch Picoscope would be great for this. However, I’m sure the problem is NOT with any of the external electrical devices. To everyone reading this, although well intentioned and trying to be helpful is both welcome and greatly appreciated, please don't suggest other reasons such as petcock vacuum, carb problems, bad earths, valve clearances etc. etc. etc., it is an ELECTRICAL fault, which I believe “May” be inherent in the red box CDI's*

*Qualification:
I decided to try a secondhand OEM CDI from a wrecker (who swears it was functioning OK), but again had no spark whatsoever. After returning it for my money back and having a bit of an argument over it - I rode the misfiring KLR up to him with the red box in, with intention of swapping it out with his to prove his CDI had no spark... he didn't want to see this and told me Kawasaki have a "hidden" resistor inside the original OEM ignition switch, which is not shown on the Klymer manual drawings. The reason being - according to the wrecker - is it's an antitheft device, to make it difficult to hot-wire. I was quite skeptical about this at the time and thought he was b...sh.... me, so argued and got my money back. Now, after what I’ve read about resistor on other models, I'm thinking maybe there was some truth to this ?? Anyone? Is it time for me to eat humble pie 😊

If so, would this resistor provide a -ve DC offset bias for the exciter coil AC input? Without an electronic schematic of the CDI internals, it’s hard to know what they are doing in there.

I’m very OCD about my bikes and can’t really put up with a misfire, but don’t think much chance of selling it like it is. Although not worth much, she’s too good to just chuck away or send to wreckers… apart from which I think I’ve done my dash there 😊. Oscilloscope shots of exciter coil and ignition coil primary can be found here:

Oscilloscope Shots
 

·
Premium Member
2013 KLR 650/692, 2017 HD Electraglide Ultra
Joined
·
1,179 Posts
I’m an engineer too, and electrical intermittent problems are the most frustrating. Re the switch having an anti theft resistor inside: that’s in Gen2 bikes in the USA, but I would expect the same in Oz. Gen2’s around the world have a different ignition system than Gen1’s. So a Gen1 CDI should not be affected. The 1998 US model I picked up a week ago was missing its key, and the PO wired around the ignition switch to a regular toggle switch and it works fine. This would not work on a Gen2.

Your ignition traces look normal except that in the first picture, there are two traces that show the negative spike occurring after the positive buildup, rather than in the middle as the other signals demonstrate. I’m guessing that is related to the misfire.

At this point I would try swapping components with known good ones, if you can get some. Obviously, the CDI box, but also the coil. Coils can do strange intermittent things when they start going bad, although that’s usually at higher load and RPM.

I don’t know if you’ve tried this method yet: run the bike stationary and wiggle and pull on every wire you can reach to see if it causes misses. Breaks within wires under the insulation can occur where they are subject to vibration or sharp bends (someone here recently found one of those). You will have to prop up the tank or remove it and run fuel from a hanging bottle or something like that, in order to access the wiring loom under the tank. And while you’re in there, pull apart and grease all the connectors you can find.

Hopefully Damocles will jump on this thread. He knows more about these ignition systems than just about anyone else.

And here’s a very useful site for things KLR if you haven’t found it, Tom Schmitz’ “Souperdoo”, specifically his article on ignition switches:

 

·
Registered
Joined
·
9 Posts
Discussion Starter · #3 ·
PeteK, thanks for taking the time to read it. Yes, I work in the commercial marine industry and apart from general electrical/control/instrumentation design work, I have lots of experience with complex control systems on defence vessels, tugs/rig support vessels etc. such as power management, dynamic position and propulsion control systems to name a few, and the fact I can't find a misfire on a bloody simple CDI system is damaging my rep! :LOL:.

Re: [there are two traces that show the negative spike occurring after the positive buildup, rather than in the middle as the other signals demonstrate. I’m guessing that is related to the misfire. ], these are occurring every revolution and I think are somehow caused by the capacitor discharge. I have knocked up the below basic drawing of the key wiring to help evaluate the waveform in the oscilloscope shots. The CDI internal details are very simplified and purely theoretical as I have no schematic drawings of what goes on in there. The regular negative spikes every period are a bit of a mystery to me; if you can explain or have a theory, I'd love to know.
29420

Yes, I will check again the wires from exciter and trigger, just in case I've missed a break somewhere. I'm not really going to buy any more parts for swapping out though, as I also have a 1999 LR Defender TD5 which is draining the purse badly as well. Any other feedback welcome :)
 

·
Registered
klr 650 c
Joined
·
160 Posts
The bastards who stole the bike shorted out the cdi, I wonder if the exciter coils got damaged somehow in the proces, shorting a bit when hot, or wiring leading from them to cdi, two cdi-s changed with same symptoms, its odd. Run the engine to operating temp and check the exciter coils again
 

·
Registered
klr 650 c
Joined
·
160 Posts
Also can You mimic those exciter coils by disconecting their wires and adding external power source, needs to be hv ac source? Might be bad/weak solder joints on exciter coils?
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #6 ·
The bastards who stole the bike shorted out the cdi, I wonder if the exciter coils got damaged somehow in the proces, shorting a bit when hot, or wiring leading from them to cdi, two cdi-s changed with same symptoms, its odd. Run the engine to operating temp and check the exciter coils again
Thanks. Good thought and I have considered this before, but there is no real way to damage them from attempted hot wiring. Even if they connected full battery voltage +12V, it wouldn’t heat them up as would only be about 85mA for a 140 Ohm coil (1W), the coil winding wire is plenty big enough or that, and the excitor output is grounded to -ve anyway to stop engine through key switch or kill switch. Coils seem to check out fine.
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #7 ·
Also can You mimic those exciter coils by disconecting their wires and adding external power source, needs to be hv ac source? Might be bad/weak solder joints on exciter coils?
It’s not easy, but not impossible to simulate the excitor output waveform using a function generator and step-up transformer with matched impedance ( it’s only about 142Hz from excitor at idle), but tricky bit is also synchronising the simulated trigger signal (pickup coil) to it. I think first of all though, I’m going to see if I can get hold of a “good” 4-channel, digital storage oscilloscope (Picoscope or similar) that I can check / store all 3 signals - excitor, trigger and ignition coil primary output - that way I can capture it when a misfire occurs and see exactly which signal has missed. My money’s still on the CDI’(s) though.
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #8 ·
Just a bit of an addition to my comment above, if any electronics capable guys/gals are thinking of trying to simulate an excitor source with a function generator, it would need to be able to source a fairly high power level or run its output through a transistor amp to boost it as well as being stepped up in volts through a Txmr. I haven’t yet bothered to work out how much, as it would depend on input impedance of the CDI. For this you’d need to measure the AC rms current flowing into CDI and voltage (expected around 40-60Vrms). Typical FG can source 20V p-p into 50 Ohm load.
 

·
Premium Member
2013 KLR 650/692, 2017 HD Electraglide Ultra
Joined
·
1,179 Posts
Some more of my amateur comments: I expect that the CDI does not need the exciter coil waveform as an input. I’ve seen other CDI systems that used +12V as the input. It appears to me that Kawasaki used this system as a carryover from pre-CDI bikes so that they could be self-powered for kick-starting, even without a battery. As an experiment, you could try putting 12v to the power input instead of the exciter coils. As Hughie pointed out, the internal resistance should be high enough to prevent shorting out. It either will work or it won’t. If it works, see if you still get a misfire.

I suggested earlier to swap parts with someone who has the same bike. Since I re-read your original message, I see that you tried two CDI boxes (plus your Red Box) and still had the misfire. I think you can eliminate that as the source of the misfire. Also, from what I’ve read from other folks, the CDI either works, or it doesn’t work at all.

So what’s left? The exciter coils, trigger coil, and ignition coil. You tested the output of the trigger coil at 4v, which is greater than the minimum spec, so that should be good (did you hook up a meter or o-scope and watch the output from the trigger coil while it was running and misfiring?). For the exciter output, If you can substitute a generated waveform or +12v and it doesn’t misfire, then you will isolate the problem circuit. Test the ignition coil. I know you don’t want to spend more money on parts, but they aren’t expensive on eBay, and even if the problem isn’t the ignition coil, you’ll have a spare (I keep a spare coil for my Gen 1 and Gen2 KLRs). I’ve had bad coils do weird things. They tend to fail over time, intermittently, rather than all or nothing. Although not a high probability bet, if I had to put my money where my mouth is, I’d bet the ignition coil is going bad.

Where’s Damocles?
 

·
Premium Member
2013 KLR 650/692, 2017 HD Electraglide Ultra
Joined
·
1,179 Posts
As I reread your message, I see that neither your original CDI box, nor the replacement from the wrecker worked at all. So at this point, you have tried two red boxes and they both misfire.

And a couple more questions: does the Red Box CDI unit plug directly into the KLR harness? Does it use the exciter coils for its power or does it have a separate connection to the battery? And did you test your original CDI box for resistances as described in the manual?
 

·
Registered
klr 650 c
Joined
·
160 Posts
I have read that you wired directly the cdi bypasing all the switches and diode pack, i have a feeling that this is going to be something stupid simple but hell to diagnose, hope you will find why it is misfiring, myybe pickup coil got damaged from hotwiring? Probably a long shot, looks like the purpose of yellow/black wire from cdi to starter relay which is used on 1990 models and up is for spark retarding during startup, maybe You can try disconecting it if the earlier models can drive without it? Dont know is it safe as earlier cdi-s probably have different spark timing. Maybe that part of aftermarket cdi is giving you headaches?
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #12 ·
Some more of my amateur comments: I expect that the CDI does not need the exciter coil waveform as an input. I’ve seen other CDI systems that used +12V as the input. It appears to me that Kawasaki used this system as a carryover from pre-CDI bikes so that they could be self-powered for kick-starting, even without a battery. As an experiment, you could try putting 12v to the power input instead of the exciter coils. As Hughie pointed out, the internal resistance should be high enough to prevent shorting out. It either will work or it won’t. If it works, see if you still get a misfire.

I suggested earlier to swap parts with someone who has the same bike. Since I re-read your original message, I see that you tried two CDI boxes (plus your Red Box) and still had the misfire
[2 x red box CDI, both misfired, and one "used" OEM, the used OEM did not spark whatsoever] I think you can eliminate that as the source of the misfire. Also, from what I’ve read from other folks, the CDI either works, or it doesn’t work at all.

So what’s left? The exciter coils, trigger coil, and ignition coil. You tested the output of the trigger coil at 4v, which is greater than the minimum spec, so that should be good (did you hook up a meter or o-scope and watch the output from the trigger coil while it was running and misfiring?)
[yes I hooked up scope (not the same scope I used for checking exciter and ignition coil), but it was very low quality analog and it's extremely difficult to get stable triggering on a scope when the input channel is bouncing all over in frequency - due to gold old single pot bangers :) - one revolution is rarely same exact speed as next one. And I was never able to capture a signal during a misfire - they seem to get stage fright when I get the oscilloscope out :). Please understand as I've said before, this is a SINGLE misfire and sometimes have to wait 5min at idle before it happens (ONCE). When riding it's more often because more revolutions = more chance of misfire - maybe 1-5 times in a km]. For the exciter output, If you can substitute a generated waveform or +12v and it doesn’t misfire, then you will isolate the problem circuit [exciter input to Gen1 CDI needs 30-60VAC rms to charge the main firing capacitor with sufficient voltage - please have a look at the waveform, the peak voltage (+ve half cycle) is around 120~150Vp. I have no personal experience/knowledge of Gen 2 CDI, but I've read on this forum it has a 12VDC supply, but I would think, and @Damocles may correct me on this - inside the Gen 2 CDI, this 12VDC voltage is inverted** (by a switching PSU), or has some form of an electronic voltage doubler circuit (inductor/capacitor arrangement in oscillating circuit arrangement) to provide sufficient voltage to the ignition discharge cap. Test the ignition coil. [I know its very long for which I apologize, but please see my original post - ignition coil has been tested and also has been replaced with an aftermarket one, but I put the original back in as the bike pulled a bit stronger on original one and the aftermarket one did not cure the random misfire] I know you don’t want to spend more money on parts, but they aren’t expensive on eBay, and even if the problem isn’t the ignition coil, you’ll have a spare (I keep a spare coil for my Gen 1 and Gen2 KLRs). I’ve had bad coils do weird things. They tend to fail over time, intermittently, rather than all or nothing. Although not a high probability bet, if I had to put my money where my mouth is, I’d bet the ignition coil is going bad. [It's not soooo much about the cost (shushhh don't tell my wife), but in Australia, particularly WA - stands for Wait Awhile (or Windy Always) - there's not much available and takes forever to get here. Just to give you an idea, there's hardly a single KLR around mine's vintage for sale here.
By inverted I'm referring to changing from DC to AC and stepped-up in voltage.
Where’s Damocles?
PeteK, thanks for your suggestions. It's easier for me to give you my replies in [blue] above. I agree, most of the time CDI's, do fail completely, but I have read a number of earlier posts on this forum (which are now closed) of other people having the same problem as me, who also have the red-box CDI and claim to have changed-out everything, even rotor, but I know only too well it's very difficult to diagnose problems remotely as the problem owner is very subjective. However, I've been in this game a very long time and am extremely thorough. I appreciate everyones input, but I think I'm going to leave it at this for the time being. However, I WILL get to the bottom of it when time permits and when I've managed to beg, steal or borrow a 4 channel oscilloscope (DSO) and I promise to post my results on here for anyone who's interested. If I don't identify the problem as being exciter, trigger or ignition, I think I'll make my own bloody CDI for it, because I now know what all the signals are, I just need to know how much timing advance is needed between idle and max RPM and if it's linear or not through the range.

Stay tuned....
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #13 ·
I have read that you wired directly the cdi bypasing all the switches and diode pack, i have a feeling that this is going to be something stupid simple but hell to diagnose, hope you will find why it is misfiring, myybe pickup coil got damaged from hotwiring? Probably a long shot, looks like the purpose of yellow/black wire from cdi to starter relay which is used on 1990 models and up is for spark retarding during startup, maybe You can try disconecting it if the earlier models can drive without it? Dont know is it safe as earlier cdi-s probably have different spark timing. Maybe that part of aftermarket cdi is giving you headaches?
C model, thanks for reply. Yes pick-up (trigger) coil tests out OK and is giving a nice clean trigger spike of about 3-4V at idle (will increase with RPM though). But, please see my reply to PeteK regarding not being able to see or capture any "Possible" misfire of trigger signal. I need to test with better (higher spec) digital storage oscilloscope (DSO). I like your comment regarding the yel/blk wire though. I can't remember if I checked with new red-box CDI on with this disconnected... I think so (i've tried so many things now), but it's easy to test. Oh... actually I think I did because I remember it made no difference to engine starting. But will check again. I wouldn't think it would be cause of a random miss though.

As I said to PeteK: I WILL get to the bottom of it when time permits and when I've managed to beg, steal or borrow a 4 channel oscilloscope (DSO) and I promise to post my results on here for anyone who's interested. If I don't identify the problem as being exciter, trigger or ignition, I think I'll make my own bloody CDI for it, because I now know what all the signals are, I just need to know how much timing advance is needed between idle and max RPM and if it's linear or not through the range.

Stay tuned....
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #14 ·
As I reread your message, I see that neither your original CDI box, nor the replacement from the wrecker worked at all. So at this point, you have tried two red boxes and they both misfire.

And a couple more questions: does the Red Box CDI unit plug directly into the KLR harness? Does it use the exciter coils for its power or does it have a separate connection to the battery? And did you test your original CDI box for resistances as described in the manual?
PeteK, sorry I missed this one. Yes both red box CDI's have random misfire, although the overall engine torque appears to be greater than original. Yes red box is direct OEM replacement from Ricks Motorsport Electrics in USA. No it does not use a battery connection, it's exactly as per OEM connections.

Yes tested original (failed) CDI with both my Fluke 187 DMM and an old Hioki analog multi meter... I keep for sentimental value :). It did not correlate to the Klymer manual test values, and it was definitely dead. BTW, although it's much more basic than the KLR CDI, I have previously made a CDI to replace a failed DANSI one on my old Aermacchi Harley Davidson RR250 Grand Prix bike, which works OK - tested up and down the local street to the neighbors dismay, but not tested in anger yet though. The bike is same as this one but I haven't restored properly yet... work in progress :)
29452
 

·
Premium Member
2013 KLR 650/692, 2017 HD Electraglide Ultra
Joined
·
1,179 Posts
Hughie wrote: “However, I WILL get to the bottom of it when time permits and when I've managed to beg, steal or borrow a 4 channel oscilloscope (DSO)...”

HAH! I see now—it appears you just WANT to play with a 4-channel recording o-scope! 😁

Good luck, and I look forward to your findings, since I’m really curious now. And yes, it’s difficult to “consult” in diagnosing an intermittent problem from half a world away. But let me be the annoying bug in your ear one more time—find someone else with a Gen1 KLR and offer them a few beers for you to come to their house and swap parts for testing. And you may get a riding partner in the bargain.
 
1 - 15 of 15 Posts
Top