AvivAr
Parking Attendant
Posts: 17
|
posted October 14, 2007 06:01 AM
Strange Misfiring Fault - ZX12R - Two Coils Don't Fire
Hello there,
I'm helping a friend out with a ZX12R (2004) project. The problem is that after I put the engine together for him (on the bench), he had the Kawasaki dealer here fit the engine....
At first the engine would not rev above 5-7k rpm, they said it was the harness and replaced it (with a brand new one) ...
What a shock, now the engine revs till 5k then as it goes above that, two coils stop working..
The dealer says that there's and fault in the ECU that claims to be an internal ECU fault and that from what they can tell (which is not much) the fault is with the ECU..
From my experience with diagnostics (10 years+) the dealer is, to put it lightly, not on the mark here.
Has anyone got an idea, or experienced something similar? Is there something that could cause this symptom - Limp Home Mode?
Any help would be well appreciated as my friend (the owner) is borderline suicidal at this point. (If I told you guys the whole series of events, it would take about 20 pages)
With thanks,
Aviv
The engine's a 1375cc + ported head with enlarged valves + cams (all from Muzzy)
  
|
Seth ZX12r UK
Expert Class
Posts: 238
|
posted October 14, 2007 07:06 AM
if its the ecu you should be able to put in dealer mode and read the Fault code from the F1 light.
|
ridgeracer
Pro
Posts: 1309
|
posted October 14, 2007 09:22 AM
The ECU does have a 'Limp Home Mode' if certain sensors fail but it consists of using some basic math formulas to substitute for MAP values dependent on the sensors. The engine won't run as smooth but there is no 5k limit that drops 2 coils mode that I've seen in the ecu software but then I haven't seen the 2004 32bit code either.
Actually it would probably be more helpful if you gave us the full 20 page story.
Just off the top of my head....
1) Which two coils are not firing? Is it always the same two?
2) Have you swapped the coil sticks? Does the problem follow the stick or stay at the cylinder?
3) Does it have the stock timing wheel on the end of the crank?
4) Obviously the engine was taken totally apart to put in the big bore kit...any chance the crank position sensor was damaged? If the spacing was wrong or the sensor was damaged and it was putting out a weak signal it could cause intermittent coil firing. It would also be sensitive to rpm and changing the harness could also make a difference in its failure mode.
Have they checked the crank and cam sensor resistance and signal output levels?
5) If the bike is cold, as in sat overnight cold, and you start it up and first thing try to rev it up over 5k in neutral what would happen? What about when the bike and engine are at normal operating temperature?
As for the dealer I would tell him to put up or shut up. In other words if he says it is the ecu tell him to order one but if it doesn't fix the problem your not paying for it he is.
____________
|
Avivar
Parking Attendant
Posts: 17
|
posted October 19, 2007 08:43 AM
Thanks a ton for the advice;
I'm going to help my friend out on this one; I called the dealer, he's going to put the engine back in so that I can do the checks on it.
I'm waiting for him to get his rear in gear. In the mean time;
He told me that the ECU fault is for coils one and four (he couldn't recall the exact wording).
1. It's allways coils 1 & 4 that shut off.
2. The coil sticks were switched with 3 & 2 and the fault remained with 1 & 4.
3. It's not the standard timing wheel. (Muzzy Ignition Advancer - it ren before the big bore kit was fitted, but you have a point; I dialed in the new cams and it could be completely off!)
I'll check the cam / crank sensor resistance and air gaps. Do you happen to have this data? (cam / crank sensor air gaps and resistances).
Symptom / fault temperature differentiation still need to be checked.
Many thanks,
Aviv
|
johnnycheese
Pro
Posts: 1008
|
posted October 19, 2007 03:13 PM
does it have a shiftlight on the bike?
____________
Builder and tuner of some of the fastest N.A. and P/A Hayabusas and ZX12 /14 in Texas
|
Avivar
Parking Attendant
Posts: 17
|
posted October 20, 2007 09:53 AM
No. It's got an air shifter, but it's disconnected.
|
tcchin
Zone Head
Posts: 867
|
posted October 20, 2007 10:07 AM
If the plugs are faulty, they will not allow the coils to discharge and will give the same error as a bad coil. Maybe try swapping plug positions to see if the error follows?
|
lonniemac
Expert Class
Posts: 225
|
posted October 22, 2007 08:59 PM
which kill box on the air shifter and is it just unpluged from the conector or is it cut loose at the coils and check the wiring on it. some can be activated a lot of ways with ground or with power. the the sensors u mention are used for each coil not to hit only 2. also r u using the horn button for shift? it goes to ground when pushed.
|
avivar
Parking Attendant
Posts: 17
|
posted October 25, 2007 12:23 AM
I went to see the bike myself;
1. All the 'add-ons' are disconnected, the engine management is connected as original.
2. The symptom is that as you accelerate the engine, all is well till about 5k, then you can see the spark die out ( on cylinders 1 & 4) till about 7,500 rpm when it returns as if nothing happened. When you release the throttle, the spark becomes very strong, but also dies away to almost nothing (tiny almost unnoticable spark) between 7.5 and 5 k rpm as the engine decellerates. We ran the engine with coil one fitted to an external plug.
3. The fault codes are for a wiring fault on coils 1 and 4 (forty something).
4. I checked the voltage on the coil when the engine is running; the positive remains strong all the way through the rev range, with do differentiation when the fault appears.
5. When checking the primary signal on a scope, the amplitude drops down to about 5v when the fault appears, but out of the bad range it's nice and strong.
6. Peak voltage was also checked on the crank and cam sensors and results were as per the reference values in the factory manual.
We've managed to find another bike who's owner is willing to donate his bike for parts for a few hours, so we're going to try the ECU (we've found a way around the immobilizer) and timing wheel.
Any ideas????
|
shiggsy
Expert Class
Posts: 128
|
posted October 25, 2007 02:15 AM
quote:
I The fault codes are for a wiring fault on coils 1 and 4 (forty something).
The coil fault codes are in the fifties, injectors are in the forties.
You probably have but you didn't say if you had tried swapping the plugs ?
Did you measure the voltage signals at the ECU or the coil, just in case you have bad wires downstream.
|
Avivar
Parking Attendant
Posts: 17
|
posted October 25, 2007 09:26 AM
Sorry, the fault codes were 51 and 54.
I measured the voltages at the coil(s) and the plugs were swapped and replaced.
|
ridgeracer
Pro
Posts: 1309
|
posted October 25, 2007 11:44 AM
quote:
...4. I checked the voltage on the coil when the engine is running; the positive remains strong all the way through the rev range, with do differentiation when the fault appears.
5. When checking the primary signal on a scope, the amplitude drops down to about 5v when the fault appears, but out of the bad range it's nice and strong....Any ideas????
The coils have a common +12V on one side and the ECU shunts the other side to ground with a mosfet driver transistor. So if I'm looking at the ECU side of the coil on the scope I would expect to see a waveform that alternated between 12V (ECU driver off ) or near 0V (ECU driver on)
I'm not sure how to read your comment number 5. During the fault the signal alternates between 12 and 5 or it alternates between 5 and 0 volts. I guess it doesn't really matter, its bad either way. According to comment 4 the 12V bus supplying the coils is not dropping to 5 and you said up thread you have swapped the coil sticks themselves. It really looks like a bad ECU. The ECU can only switch the coil drivers on or off. For it to be 5 volts something is not right somewhere.
However for it to fail only between 5k to 7.5k, repeatedly is a really strange failure mode for a mosfet, not impossible, but uncommon. And then to coincidentally have two drivers simultaneously fail in the same way seems against the odds. Unless of course this ECU actually internally has the coils paired up and has only two drivers, one for each pair. Given the 1-2-4-3 firing order it would make sense that the pairs would be 1-4 and 2-3.
I really think swapping the ECU is the next step to take.
One quick question though. Do you know if the bike fails the same way in gears other than 1st?
____________
|
Avivar
Parking Attendant
Posts: 17
|
posted October 26, 2007 03:12 AM
I tend to agree with what you're saying, and the fault appears in all the gears including N. Just to clarify what I mentioned in paragraph 5; the signal isn't shunting to ground, it's remaining at about 5v+. Everything seems to be leading to the conclusion that the problem is in the ECUs ignition hardware. We're going to try and get that other owner to donate his ECU for a test, but the dealer is making things complicated (there are only 6 B3s in the country, included).
I know this is out of the ordinary, but would you happen to know where a second hand ECU could be purchased in the US?
Once again, many, many thanks!
|
shiggsy
Expert Class
Posts: 128
|
posted October 26, 2007 04:18 AM
You could try emailing these guys, I think they have several types in stock.
http://cgi.ebay.com/ebaymotors/KAWASAKI-ZX12-CDI-ALL-YEARS_W0QQcmdZViewItemQQcategoryZ43976QQihZ009QQitemZ190129994821QQtcZphoto
Just quote the number on the ECU to them (B3 is 21175-0038) I think. You may also be able to get one from a different country but the number may not end in 0038 so it has to be positively identified as being for a B3.
|
Avivar
Parking Attendant
Posts: 17
|
posted October 26, 2007 05:34 AM
Thanks again for everything. I'll keep you informed as to how things go this week FYI.
Have a great weekend!
|
Avivar
Parking Attendant
Posts: 17
|
posted October 29, 2007 10:40 AM
The problem is solved!!!
After a few hours on the bike, I found the problem (no thanks to the dealer, though he has a very good technician)
1. The ECU was not at fault. We fitted the ECU from another bike, and the fault remained.
2. We bypassed (cutting the existing ones) the signal wires for coils 1 & 4.
3. The problem was simply an open / loose connector clip in coil 1. Black / signal.
Even though when the coil resistance was checked all was OK. Evidently, when the vibrations reached a certain level (4-5k RPM) the contact was lost.
All's well that ends well!!
Thanks for all your help!!
|
SHIGGSY
Expert Class
Posts: 128
|
posted October 29, 2007 12:44 PM
quote:
Even though when the coil resistance was checked all was OK.
The only positive result of a resistance check is a failure, (bit late saying that I know).
Cool, least it was a cheap fix.
|
AvivAr
Parking Attendant
Posts: 17
|
posted October 30, 2007 02:45 AM
It's not over till the fat lady sings.... After fixng the connectors, the fault reappeared similarily at slightly higher RPM. Only when we put the Muzzy timing wheel at 3-5 deg RETARD did the fault disappear completely.
Is it normal that we get a Sub Throttle Actuator fault when the Power Commander is connected. It seems logical, but I'ld like to be sure (Stage 3 map)?
Thanks
|
Avivar
Parking Attendant
Posts: 17
|
posted November 07, 2007 02:02 PM
OK, well the fat lady choked....:-)
When I wrote 3-5 Deg retard, I meant 3-5 notches.
When the rotor is at less than 15 deg retard we get our famed coil fault!?!?!?
Now can anyone explain that one????
To add insult to injury, the engine is overheating. There's no blow-by into the coolant, the fans are OK, you can see fluid flow via the rad cap, the temp just keeps rising.
Has anyone seen that ignition / fueling can cause drastic overheating in these engines?
We've asked for a PC map for the engine we built from Muzzy and are hoping for the best.
Desperately in Need of Help! Anyone seen or experienced anything like this??
FYI here's a list of what's fitted to the bike;
ZX-12R B3 (2004 Euro Spec);
1. Muzzy Adjustable Ignition Advance Rotor
2. Muzzy Race Cams with Springs and Titanium Retainers (Dialed-in as per instructions)
3. Muzzy 1375cc Big Block Kit with N2O pistons (on standard crankshaft, thus lower capacity)
4. Muzzy Air Shifter
5. Muzzy 75 shot dry N2O kit (will be running at 0, 25 and 50 shot options only)
6. Muzzy Pro-Pipe Ultra Titanium (race system) Exhaust
7. Muzzy Velocity Stacks
8. Power Commander (model 160)
9. Ported and polished cylinder head with Ferrea Valve Kit [oversized intake valves (+ 1mm) std. size exhaust, 3 angle valve seats] (std. chamber volume)
10. 18x41 Final Drive
11. Race Gas (105 Octane)
|
ridgeracer
Pro
Posts: 1309
|
posted November 08, 2007 11:00 PM
What color was the bike again?
(have no ideas)
____________
|
Avivar
Parking Attendant
Posts: 17
|
posted November 09, 2007 02:45 AM
It's black, but we're considering marine blue and throwing it into the Med....
|
ridgeracer
Pro
Posts: 1309
|
posted November 09, 2007 07:36 AM
quote:
OK, well the fat lady choked....:-)
When I wrote 3-5 Deg retard, I meant 3-5 notches.
When the rotor is at less than 15 deg retard we get our famed coil fault!?!?!?
Now can anyone explain that one????
Ok I slept on it and maybe I've got one idea for what it is worth which is probably not much.
To help me hack the ecu I built a black box that provides all the electrical signals a bike does to fake out the ecu so I could run it on the bench. The hardest part was providing the cam and crank pulses so the ecu would think it was hooked up to a spinning engine.
One thing I couldn't find in the manual was the exact relationship between the cam signal, which pulses once every two engine revs, and the crank pulses. The software shows that the crank pulses determine when to fire the injectors and coils and the cam pulse synchronizes it all telling the engine where TDC of cylinder 1 is. But it gets complicated: Does the rising edge of the cam pulse occur exactly with the rising pulse of that crank pulse? Slightly before? Slightly after? By how much?
To make a long story short I found that if the 'phase' was not correct the ECU would still run correctly at the lower RPMs but not the higher ones. What I found is the pulses don't have to be sync'd exactly, but that there was a window or range it which it would work but that window would shrink as RPM increased.
By changing the timing wheel settings you are in effect changing the relation phase of the cam pulse to the crank pulse. You say it won't run unless you run it 15 degrees retarded. Maybe you have to advance it that much just to get it in the window? This is only a possibility because you've had the cams replaced.
Two suggestions both a pain but you had noting better to do this weekend right?
Put the stock timing wheel on
Put the engine on TDC cylinder 1.
Mark the timing wheel tab closest to the crank sensor with a felt pen
Check the timing nub on the cam. If I remember correctly you can do this with out pulling the valve covers. Just remove the cam sensor from the side of the head and look down the hole.
Move the crank until the you see the cam nub centered in the cam sensor hole.
Record the position of the crank wheel tab you marked with the pen earlier
Repeat this procedure on another bike that runs. Compare results.
-OR-
You mentioned up thread you used a scope on the bike. If you have a dual trace scope measure the cam pulse on one and crank pulse on the other at the same time and measure the relationship of the cam pulse edge to the nearest crank pulse.
Repeat on a known good bike and compare. (you should use the stock timing wheel on both bikes)
That is a lot of work o go thru on a hunch and your newer 32bit ecu is probably different than the 16bit ones I bench tested but its not an impossibility. Also engine timing settings can effect engine temperature.
Its a long shot but it is an idea 
____________
|
Avivar
Parking Attendant
Posts: 17
|
posted November 09, 2007 08:17 AM
I believe that it's more than a hunch...
I just got the missing page from Muzzy's fitting instructions....
The state that the pickup has to be relocated by approx. 15 deg.....
I am a true believer in RTFM, but you need the whole FM in order to read it...
I'm sure that this is where the fault lies.
What I'm wondering now, is could igniton timing cause that much overheating while driving [sorry, riding] slowly?
I car engines, under low load, I've never seen that. The thermodynamics are different on bikes, I know. Less metal mass, faster heat exchange, etc.
Have you heard or seen something like that?
Thanks,
Aviv
|
|
|
HOME
NEW TOPIC