Help! No Spark. Very Up A Creek.

[baraka]

[restarting this topic in the right forum]

So. I just finished a 2500 mile trip on my '02 VFR800A, 48,000 miles on the clock. On the last leg, the bike died as I clutched in and dumped the throttle to slow down after a long highway haul. No worries, it started right back up and I kept rolling.

Several days later, I noticed that it was not starting well cold. For the first time ever, I had to open the throttle up a bit to get it to start. It got progressively worse: it wouldn't start without the throttle being cracked and held open, and held at 3-4,000 rpms until it warmed up. Then it would run all just fine, except once or twice it would just die when slowing down and going to idle.

I tore it apart today thinking it was the starter circuit/wax unit. The starter circuit did need to be freed up--it was stuck at the warm setting and wouldn't open up when cold. But even manually actuating the cold starter circuit didn't help it start. It would run really rough if held at 4,000 rpms, until it warmed up.

So I pulled the plugs, which were blackened. I believe it has been running rich. I cleaned them up, put them back in, and couldn't get it to do anything. Pulled each plug out one by one and checked and was unable to get any spark from three of them. I was able to get one or two very weak orange sparks from #4, out of a good thirty to forty seconds of turning the engine over. During the whole spark plug test, I had my battery jumpercabeled to my buddy's car (A Honda Fit, so it's all in the family) and his car was running, so as to rule out the no spark because of weak battery issue.

Needless to say, I'm shocked (terrible electrical pun, I know).

I am definitely ordering some new plugs tonight. I have a ride to work tomorrow morning, which is good, because the bike is my only mode of transportation. :)

The engine sounded like this the last time I had it running:

http://www.youtube.com/watch?v=dDi0XfqrJfA

I was holding a steady throttle.

My current guesstimation is this: that surging was caused by the ignition cutting in and out. And now the ignition has completely cut out. I cleaned up all four plugs pretty good before putting them back in, and they look decently serviceable, unless iridium plugs are much more picky than the old ones I used on my old bikes. Nevertheless I am going to pick up some new plugs ASAP.

So. I doubt that the new plugs are going to fix the problem. Is there a chance it's a fault somewhere in the wiring harness, and if so, where should I look? Or should I just go ahead and replace the pulse generator or the ECU?

[MBrane]

You need to check your wiring. All of it. Don't just look at it check it with a known good meter for actual resistance. Hopefully you find a problem, and that problem hasn't toasted your CDI. Be prepared for the possibility that it has.

[Guest]

Not familiar with your specific wiring, but there will be a sensor/sending unit on the crank that will send a signal to the ECU . This will in turn send a pulse to the individual coils ( one per cylinder). Bad grounds /faulty connections would be an fairly easy fix. Do you have a factory wiring diagram? I know you cleaned the plugs but new might be the answer.

GL

Jeff

Of course, it could be the ECU, too.
Think positively.
GL
Jeff

[baraka]

Thinking positively. It's not currently flashing any FI codes except MAP and Intake Air Temp sensor, and both of those are unplugged right now.

I have no idea where to start on the wiring harness, and a very rudimentary knowledge of voltmeters. Any suggestions would be greatly appreciated. Otherwise, I guess I'm going to have to enter the long dark of Moria...err...start pulling panels and guessing where wires go.

[YoshiHNS]

Did you check for spark again when you replaced the plugs? Might be as simple as bad plugs. Had it happen on the ST, just as you described.

[dalesvfr]

Fuel pump went out on my vehicle once. Triple A got me to a garage and since there was no sound from the pump they commenced to drop the tank and put a new one in. When they tried to start the truck there was no sound from the new pump. They disconnected the ground at the frame, made it nice and shiny and the pump came to life. My theory is that the old pump was fine but had a bad ground. Just before it went out I was astonished to see that I was getting 28 mpg from a Chevy truck with a 5.7 liter, so I think it was working intermittently until it finally lost ground fully. In other words, have you checked all your grounds?

[baraka]

VFRNess and four new spark plugs ordered. While I wait: time to check the grounds. Preliminary inspection seems to indicate that the 30A fuse looks fine. Pulled the plastics this afternoon, will look at the rest later. Will pick up a new battery as well, juuuuuuust in case.

[MBrane]

One of the problems with CDI ignitions is they don't like no-load conditions i.e. if there is no path to ground for the spark energy it will feed back into the electronics, and destroy them. Seen this issue more than once especially on bikes that have been cranked over with the plug wires disconnected. Bad idea.

Checking the wiring for resistance is a simple matter of checking the voltage drop or resistance over a circuit. Any length of wire has a certain amount of resistance or V drop. What you need to determine is if the R or V-drop is more than a comparable length of known-good wire. Connections are always suspect, but wire can go bad underneath the insulation. I've seen lengths of wire that were completely green with corrosion once the plastic insulation was stripped off. At that point the wire basically becomes a really long thin resistor which generates heat, and adds to the resistance which heats up the connectors which adds to the resistance.....

[baraka]

MBrane: So, if the ground was bad for the ignition circuit, it could have simultaneously fried all four coils? Or the ECU? Or the pulse generator?

[MBrane]

MBrane: So, if the ground was bad for the ignition circuit, it could have simultaneously fried all four coils? Or the ECU? Or the pulse generator?

Probably not, but a bad plug wire or badly fouled plug could do it if left that way for any length of time.

The reason for checking the wiring is it's the cheapest, and easiest fix. If the wiring checks out then move on to the various components starting with the cheapest one first. CDIs are not cheap, and are difficult to test.

[baraka]

Is the CDI the coil itself? or the ECU? I am having a hard time finding the CDI.

So I freshly charged the battery and replaced the sparkies. Cleaned up the two ground points by the fuel tank mounting hinge. Cleaned the terminals on the battery and the starter solednoid, just for kicks. Checked the notorious 30 amp fuse, and it is only slightly discolored at the point where the fuse leads connect with the harness.

All the leads had a fine white powder--battery, ground, fuses--between layers in the connection. I used a wire brush to shiny these points up.

Put it all back together, but left the #3 plug/coil hanging out. Bike turns over strongly and confidently with bright lights, but zero spark AND zero gas coming out of the spark plug hole in the cylinder. Air, yes, gas, no. Checked with my hand, nose, and eventually a lighter (my dad is a redneck at heart--his recommendation).

Could this indicate a faulty ECU? Or a faulty pulse generator? Or more wiring hiccups? How does one test for a faulty ECU?

[spud786]

Does the fuel pump prime(the winding up noise). thats the only thing I can think of.

[MBrane]

Is the CDI the coil itself? or the ECU? I am having a hard time finding the CDI.

All the leads had a fine white powder--battery, ground, fuses--between layers in the connection.

Ah, sorry should have been more clear. On fuel injected bikes like the VFR the CDI is usually part of the ECU itself. The coils are what supply the big voltage to the plugs, and the pulse generator is what tells the CDI where the engine is in it's rotation.

Really hard to test a ECU without tools that cost more than a new ECU. Best bet is to troubleshoot down to that part through POE, and find a known good used replacement.

That white stuff is corrosion. If left alone it will eventually destroy all your wiring. Nasty stuff.

[baraka]

Copy that.

The fuel pump does whirr for five seconds as normal whenever the ignition is activated. That's what confuses me about the lack of fuel to the cylinder.

When the bank angle relay or the engine cutoff relay drop, would that cut power to the fuel pump? Should the presence of the fuel pump working, but the injectors not injecting, indicate a faulty ECU, or could that be a result of:

a) bank angle sensor failure/misconnection

b) engine cutoff relay failure/misconnection

c) kill switch fault

d) something else I'm not thinking of.

My process of thinking is, if the fuel pump is working, and the fuel regulator is working (it isn't leaking when I pull the vacuum hose), but there's no spark and fuel in the cylinder, then whatever sends the signal to the fuel injectors and to the coils is faulty: either the cam pulse generator, the ignition pulse generator, or the ECU.

The reason I'm not looking at the sidestand switch or the kill switch primarily is because the engine will turn over right now. Could it be possible/probably that a fault in these switches will allow the engine to turn over without spark or fuel injection, but with the fuel pump working?

[baraka]

Guess what? Bad ground from the bank angle sensor due to the rewiring from the previous owner...

[MBrane]

Guess what? Bad ground from the bank angle sensor due to the rewiring from the previous owner...

Anytime I find "wiring mods" the entire harness is considered suspect. Especially if those mods involve crimp connectors, and electrical tape. I've seen some pretty ugly "fixes".

[baraka]

Well. To be honest, I'm not really sure. Around the thirteenth or fourteenth hour of fault hunting in a dark garage by yourself, you enter a sort of Heart of Darkness like experience of increasing bizarreness and desperation. The thoughts you are thinking seem to be intelligent and rational, but after sleep and re-exposure to other people...

I decided to check the ignition circuit since I had no spark. As I took the bike's front fairing apart, the problems got worse: no spark, then no fuel injection, then finally no fuel pump as well. I decided to check the engine cutoff relay. It was not relaying. Sweet! But before I shell out some money to buy a new relay, I tested the relay by connecting it straight to the battery as described in the manual, and voila...it worked fine. Crap. I decided to check the voltage in the circuit from the top down. Started at the bank angle sensor, then the engine cutoff, then the fuel cutoff, then the kill switch...and they were all showing 12.5-12.3 volts. I was confused. How could there be 12.5-12.3 volts at every connection in the system, but the relay not work?

Well, if I had paid attention when I was a child, I would have known the answer. To explain, there are two types of people in my family: electrical engineers (My dad, and my brother) and liberal arts majors (my mom, my other brother...and me.) So I bit the bullet and called Dad, and explained my process, and he laughed and pointed out that I was testing voltage from each point on the system to the ground lead of the battery, so I had verified that power was available all the way through the circuit, but the circuit was not complete because I hadn't checked the grounding of that circuit--by grounding to the battery terminal for ease of testing, I was bypassing the circuit's ground. Then he said something about how if I had paid attention during the eighty billion times he dragged out the voltmeter to work on the car or washing machine or furnace, then perhaps I would not have these problems, and maybe I should have studied something more useful in college than Political Science.

Actually, he was really nice about it.

So I went down and measured voltage across the leads of each relay and the bank angle sensor, and none of them had any voltage. So I ran a tester lead from the battery ground and touched it to the ground lead of the bank angle sensor, because that was the only ground lead I could find in the wiring diagram for the ignition circuit and WHAMMO! Lights came on in the instrument panel, the fuel pump whirred to life, the relays all clicked and I started singing the hallelujah chorus. I checked the voltage to the ignition coil, and it was there, so this should solve the spark problem (unless all four coils went bad simultaneously).

I tore apart the wiring harness and this is what I found:

Previous Owner, my God have mercy on his soul, apparently decided to rewire the R/R into what I can assume is a dedicated circuit. My best guess is that the ignition circuit was grounded and powered through the R/R ground. PO (mGhmohs) then installed a new R/R and a new wiring harness, but didn't know what to do with the old harness...so he cut the wires off the R/R and looped them to each other and electrical taped the whole thing up and said f*** it, let's ride. Oddly enough, the quality of the work is high: copper crimping, no corrosion whatsoever, purple patch tape over the crimp spots, electrical tape and heat shielding over it all, and routed carefully, with OxGuard on the connectors.

So here is 12 or 10 gauge ground wire, run in a loop to itself, with a connector in the middle of that loop, and it goes to nowhere. They are tied independently into the small green ground wires from the ignition circuit and headlight circuits. Also, there is another deviation from the wiring schematic: there is supposed to be a small green ground wire going through the gray 16 pin connector from the ignition circuit back to common ground on the bike, but it is not there. There are only 15 wires in that connector, when there are supposed to be 16. So there is no direct path to ground for this circuit.

Instead, the whole headlight, high beam, and ignition circuits are currently going to ground through a green/black wire that grounds out the instrument cluster. All that green ground wiring in the picture is joined into one itty bitty green wire through that yellow junction block. That one ground wire goes into the instrument panel, where it encounters 110 ohms of resistance before connecting to the instrument panel ground wire, which proceeds with 0.4 ohms of resistance through the 16 pin connector and on to the chassis ground.

My best guess is that there was a connection in the instrument panel between the ignition circuit ground, and the instrument panel ground, and when the PO realized that the ground was working after rewiring the R/R, he said, "F*** it, let's ride!" and our inadequate little Horatius on the bridge heroically held out for eight or nine thousand miles and now is so fried that he can't hold on any longer. The connection is there...you can get from any point in the ignition circuit to the ground, but you have to go through 110 ohms of resistance to get there. Dad did a little very rough guesstimation over the phone and decided that we are looking at something in the realm of 100 amps to make that circuit work against that resistance.

All the wires in the ground are internally continuous--no resistance. The only resistance is between the leads on the instrument cluster where the green wire goes in, and the green/black wire comes out.

The red/white wires are tied to themselves, with a smaller red/white tied in, which leads to a fuse, which leads to a green/yellow wire that goes to one of the headlight relays and doesn't appear to carry any current. I have no idea what it is for. There's also a white/black wire tied into one of the red/white ones, which goes to the connector, but there is nothing on the other side of that connector for it to connect to. It's a very lonely connector.

Since I can see no point in any of the large green/red and white wires in that loop, I'm going to chop them off, wire in a nice phat 10 gauge ground line to all those little green wires, and run that independently back to the chassis ground underneath the seat. Then we'll see if she fires up. I'll probably test the R/R and stator while I'm in there as well. Just to be sure. VFRness is due to come in the mail today, so I'll stick that in while I've got the whole thing torn apart. Hopefully the custom harness job isn't so hacked that the VFRness won't install...

[MBrane]

Obviously the PO (or whoever re-wired it for him) had no understanding of how electrons actually flow through circuits.

Anyone who deals with audio will tell you ground loops are a bad thing, but poor/no ground can cause all kinds of bizarre problems. Electrons are like water: they will find their own way to ground.

[kaldek]

The red/white wires are tied to themselves, with a smaller red/white tied in, which leads to a fuse, which leads to a green/yellow wire that goes to one of the headlight relays and doesn't appear to carry any current. I have no idea what it is for. There's also a white/black wire tied into one of the red/white ones, which goes to the connector, but there is nothing on the other side of that connector for it to connect to. It's a very lonely connector.

Since I can see no point in any of the large green/red and white wires in that loop, I'm going to chop them off, wire in a nice phat 10 gauge ground line to all those little green wires, and run that independently back to the chassis ground underneath the seat. Then we'll see if she fires up. I'll probably test the R/R and stator while I'm in there as well. Just to be sure. VFRness is due to come in the mail today, so I'll stick that in while I've got the whole thing torn apart. Hopefully the custom harness job isn't so hacked that the VFRness won't install...

For reference, here is what the wire colours indicate on the VFR800 harness:

• Red/White: Main power feed from Regulator to Engine Stop Relay
• Black/White: Common +12volt for EFI systems from Engine Stop Relay

The black/white wire leads to nearly everything in the EFI system. This includes the ECU itself, coils, injectors, EVAP solenoid, O2 Sensor heater circuits, PAIR solenoid, and the fuel pump (although it converts from Black/White to Brown as it exits the Fuel pump relay). Note that all the EFI systems controlled by the ECU are ground-switched; that is, the ECU controls the circuit by switching the ground on and off, and it's the ground wires for each component which are controlled by the ECU - except for the fuel pump which is relay-switched.

[baraka]

Kaldek, If these things are all controlled by the ECU controlling the ground, did I just FUBAR everything by wiring an independent ground into the ignition loop?

[baraka]

Well. I ran a new 10ga ground from the front wiring harness straight to the grounding point by the fuel tank hinge. Voltmeter says it's a huge success, and the ignition relay circuit lit up quite nicely. I now have initial power to the spark plugs, a nice whirring fuel pump, and...still no spark at the plugs. I can smell gas when I turn the engine over, but I can't get any ignition when I hold a lit lighter up to the open spark plug hole.

I've checked ground from the ECU (both green/pink wires and the green wire), and it's good (0.1-0.5 ohms). I checked continuity on the wires from ignition pulse and cam pulse generators, and they are good. I even grounded out a coil and plug and ran 12v to both leads and got good sparks, so it's not a coil or spark plug problem.

At Kaldek's suggestion, I unplugged the ignition pulse generator and turned it over for a bit, and it threw the FI light for the ignition pulse generator. Did the same thing for the cam pulse generator, and it threw the code for the cam pulse generator. In theory, this means that both are working, because signal from one but not the other is how it indicates that the other is not working.

The only thing left in the system, as far as I can tell, is the ECU. I picked up a used ECU from an '02 ABS from Poniesatmybagel, and plugged it in today, and nothing changed. The ECU's appear identical, but his is labelled "38770-MCW-L02-9697-100420" and mine is labelled "38770-MCW-L02-9697-101205". That's the only significant difference, and I have no idea if it's important or not. They are both made by Keihin.

I guess the next step is either shell out the big money for a new ECU, or find a kind soul with a running bike to swap ECU's with for a test. If anybody in the Atlanta area would like the help, I would be in their debt. I just found out that I have a funeral to attend in Michigan next weekend, and I don't have any running mode of transportation, so I'm getting sort of desperate for answers.

[dalesvfr]

Those are probably serial numbers is all...either they owe you a refund for a dud ECU or the problem is elsewhere.

Dumb question - could the engine be out of time? Stretched chain?

[baraka]

Dales--I sure hope not. I have to take the righthand crankcase cover thingy off to replace the ignition pulse sensor, so I'll check timing while I'm down there. But my main problem is that I don't have any spark to the plugs at all. So I'm not terribly concerned with the timing.

I did preliminary checks on the stator, and it has proper resistance between the coils, and does not have any shorts to ground. But I haven't laid eyes on it yet. Could a problem with the stator or the rectifier cause a no spark issue? People are suggesting that I'm being foolish in attempting to troubleshoot an ignition problem without addressing the stator/RR issues, but the RR is an updated one, and the stator passes the initial voltmeter test (resistance between coils, no shorts to ground).

[baraka]

Well. As the doctor would say, there's good news and there's bad news.

The good news is, I have found problem #2. A giant point of confusion here has been that the ignition pulse generator has passed all the external tests, but all symptoms point to it not doing its job. So I pulled the right crankcase cover of and, lo and behold, a bolt drops out into the oil pan...hmmmm...

Sure enough, further inspection reveals that the ignition pulse sensor is simply hanging by its wires. I have found my problem! Apparently the bolts holding it in place vibrated out and it fell out of place. It looks a little the worse for wear: the magnetic "nub" is reduced a little bit, probably from contact with the timing sprocket. The sprocket looks pretty good, and I have a replacement sensor and a new gasket for the crankcase cover. I simply drain the oil from the oil pan, recover the sensor mountain bolts, and I should be back in business.

Except there is only one bolt in the oil pan...and I distinctly remember only one bolt falling out when I took the cover off. There are supposed to be two bolts.

Either a) the PO only mounted the sensor with one bolt, or b) there is a bolt sitting in my oil pan.

Did I mention that one of my favorite mechanic games is "Find the Bolt?"

Welp. I'm going to try the whole "magnet on a stick" trick before I go and pull the oil pan. Maybe I'll get lucky...at least it looks like I have found a solid reason for the original problem.

Once I get this figured out, I'm gonna locktite those two bolts in.

[baraka]

Well. Update for the troops: I am back on the road after fixing three things:

1. New grounding line to the front wiring harness, to fix the poor ground in the ignition circuit.

2. New power line run to front high beams, and extraneous loop cut out of front wiring harness.

3. New ignition pulse (or crankshaft position sensor) and bolt. The old one's mounting bolts had vibrated out and were chilling out in the oil pan. :( But, :), they were not chewed up.

Much thanks to all who helped trouble shoot, and PoniesAteMyBagel and Tightwad for timely help with parts.