Author Topic: Fuel system system mystery (Read 21684 times)

cnoon · « **on:** September 06, 2015, 06:31:07 PM »

Hi everyone. I thought I'd start a first post with something that's got me stumped. It's a 1991 K75s and here's the background:
1. Riding on a warm day 5 months ago and the bike just quits. It starts for just short bursts and is getting spark so I suspect a fuel issue.
2. Months later, I finally get time to deal with it and test the fuel pump via the fuse box and it's dead.
3. Replace the fuel pump and filter (and bench test the prior one and confirm truly dead).
4. Start the bike and it idles for a few seconds and then dies. Won't restart.
5. Confirm that the new fuel pump is working via the fuse box and also confirm that it pumps fuel into the front of the FI rail (also to confirm I didn't put the pos/neg on backwards on pump).
6. Now here's where it gets weird. If I disconnect the 4-wire plug between the bike and the tank (with the yellow, white, brown, green&green/white). The bike will start and idle fine if on full choke. This means the fuel pump is not on and it's just relying on gravity feed(?). If I try to rev it at all, it will starve out. If it is idling in this fashion and I connect the plug, it immediately dies and will not start (until I disconnect the plug). If I try to start it with the plug connected, I can hear the fuel pump engaging on the start button and continuing for a second or two after start button release.
7. I've confirmed that the incoming green/white wire is delivering power when the starter is engaged but I'm not getting any power to the yellow or white wires when the key is on or when the starter is engaged.

So, any ideas on what to check next? I'm thinking of idling it unplugged and then running the fuel pump by jumping it at the connector (essentially bypassing the yellow and white connections). If that kills it, then there's something related to fuel pressure I guess. If it doesn't kill it and even allows it to be revved, then it's something electrical related to the yellow and white wires (fuel level indicators). I suppose I could always just snip those but that's a kluge.

- thanks, Chuck

Scott_ · « **Reply #1 on:** September 06, 2015, 07:23:10 PM »

That does sound interesting......

I suppose it's possible that something could be breaking down in the wiring pass thru where the wiring enters the tank. (ie the sending unit assy)
There have been reported issues(other owners) with the brown ground wire connection failing where it connects to the tank. Other riders have had fuel pump related issues with this, generally the pump doesn't run at all.
The repair was to replace the sending unit assy.

cnoon · « **Reply #2 on:** September 06, 2015, 08:41:38 PM »

I'm thinking that it could be a faulty sending unit. But does the sending unit have the ability to tell the bike to quit running. I thought it would just be a passive sensor.

johnny · « **Reply #3 on:** September 06, 2015, 09:14:53 PM »

greetings...

have you removed the motronic connector and cleaned up the connections...

what also stops working when you plug in that 4 pin... head light... dash lights... neutral light... clutch switch...

j o

Scott_ · « **Reply #4 on:** September 06, 2015, 10:15:50 PM »

I'd recommend you check the ground path continuity from the 4-wire connector(brown wire) to the lead connected to the fuel pump.
If your ground path is interrupted, or not clean and solid, all kinds of funky stuff can happen.

rbm · « **Reply #5 on:** September 07, 2015, 02:00:52 AM »

You won't get power at the white or yellow wires because they are sensor outputs. Furthermore, the yellow terminates in the aux. instruments connector, which is essentially open circuit.

The sudden nature of the failure indicates electrical. Check all grounds as Scott says. Did you investigate how the original pump died? Was it a mechanically frozen rotor or did the motor burn out? This might help diagnosis.

cnoon · « **Reply #6 on:** September 07, 2015, 05:42:41 AM »

I'll definitely remove the motronic and clean the connections although there's nothing else that seems to go offline when the connection is made. It is worth noting that before all of this went down, the bike had a slight short which required me to disconnect the battery or it would discharge in a couple of days. There's no evidence of that short now after changing out the fuel pump.

Not sure what exactly caused the fuel pump to fail. The strainer was still intact but I think the hose connected to it was missing or broken. The fuel pump vibration damper was in bad shape so it could be conceivable that the pump got gummed up. When bench testing the old pump, it sort of thrusts when you apply power as if it is mechanically jammed.

The brown ground wire path must be clear to the fuel pump since the pump will run either when the connector is connected or if I just run positive and negative into the tank side of the connector. Now, whether it's only partial or not clean, I don't know but can check.

The way it immediately dies when connected seems to imply something electrical, but that would imply there's some sort of signal being sent upstream and that's what's perplexing me. Thanks!

rbm · « **Reply #7 on:** September 07, 2015, 08:22:25 AM »

It could be a bad FI relay. That is the device that controls the pump. If the contacts were bad on the relay, power to the pump could be interrupted when the load is placed on the circuit. Power to the injectors is supplied through a separate set of contacts so it's conceivable that the injector power circuit works and the fuel pump circuit does not. Do you have a spare relay to test? If not, maybe taking the existing relay out of the bike and bench testing it is a possibility, before forking out money for a new one.

The Mighty Gryphon · « **Reply #8 on:** September 07, 2015, 10:04:51 AM »

So what you are telling us is that the engine will idle without the fuel pump running, and won't run at all when it is. I am having trouble believing that gravity will allow enough flow through the injectors to run the engine, even at idle. I know my bikes won't run that way. I suspect some residual pressure is supplying fuel to the injectors. The question is: how long will it run like this? I would guess not more than a few seconds.

To see if indeed the problem is in the fuel pump wiring in the bike I would try connecting the fuel pump wires in the tank connector to an external power supply so that the pump is running independent of the rest of the engine management electrics.

If the engine runs that would confirm an electrical problem. If not, I would suspect something in the fuel circuit like the fuel pressure regulator, or possible the venting of the tank is interfering with fuel delivery when the pump is running.

If the engine runs, I would connect a meter to the fuel pump wires in the bike's fuel tank connector to see if indeed there is 12v present when the engine is running. If not, the problem is probably in the fuel pump circuitry.

johnny · « **Reply #9 on:** September 07, 2015, 11:27:20 AM »

id hook up one of those drill pumps to the fuel line and see if it will go with the 4 pin disconnected...

j o

billday · « **Reply #10 on:** September 07, 2015, 04:03:33 PM »

Quote from: cnoon on September 06, 2015, 06:31:07 PM

6. Now here's where it gets weird. If I disconnect the 4-wire plug between the bike and the tank (with the yellow, white, brown, green&green/white). The bike will start and idle fine if on full choke. This means the fuel pump is not on and it's just relying on gravity feed(?).

Can't be possible! The injectors can't inject without fuel under pressure. Fuel can't be under pressure without power to the fuel pump. This just doesn't add up.

cnoon · « **Reply #11 on:** September 07, 2015, 07:12:09 PM »

I took a video for you un-believers:
https://www.youtube.com/edit?o=U&video_id=dsH6NOyYN18

Also, I later used a car booster battery to power up the fuel pump by directing power to the tank side (green and brown) of the plug. It runs the pump when invoked. I then started the bike with the plug disconnected (and it idled fine) and when I turned on the power to the pump, the motor promptly died. Hence, I'm now more convinced it's a fuel pressure (rather than electrical) issue and will check out the regulator, etc.

Hey, if any K75 owners wanted to unplug the tank and see if it idles (with the "choke" on full), that would be helpful information.

- thanks, Chuck

rbm · « **Reply #12 on:** September 07, 2015, 07:14:12 PM »

Putting the theory aside for the moment, empirical data and observation shows the theory to be wrong, if what's written is to be believed. If the FI relay is energizing the injectors and the ECU is actuating the injectors, fuel must be entering the cylinders but not under normal pressure. This could be just dribbling in or there could be residual fuel pressure in the system that just manages to get the bike started and running for a few seconds.

In your video, it certainly seems that the engine is running on residual fuel that is leaking into the cylinders. The fuel would not be under pressure since 32 hours have elapsed since you last started it. I'm guessing vacuum generated during intake cycle is sucking fuel into the cylinders from the injector rail. Once that fuel supply is exhausted, the engine will quit.

cnoon · « **Reply #13 on:** September 07, 2015, 07:31:52 PM »

Robert, I just read your message, walked outside, confirmed the tank plug was disconnected, set the stopwatch on my phone, and started the bike and the stopwatch. It ran for 1 full minute and then I turned it off with the key. If pressure is a necessary condition for fuel to enter through the injectors, then there must be some other source of pressure, but from where? If the gravity/vacuum is enough to pull fuel into the cylinders, then we've all been relying on a bad assumption. This is becoming quite interesting (although somewhat frustrating). - best, Chuck

Freelancer · « **Reply #14 on:** September 08, 2015, 12:24:05 AM »

Just a shot in the dark,

Could be the injectors. If they are shot/ stuck open to the point that fuel can gravity feed then the pump would flood out your bike.
Also, could be something in the control circuit that is locking the injectors in the open position.

I'd suggest looking at your spark plugs after the pump has been on. If they are wet then there is a big clue.

Good luck on finding/fixing the problem,
Freelancer

chris846 · « **Reply #15 on:** September 08, 2015, 04:20:34 PM »

Has a previous owner plumbed in a fuel pump somewhere on the bike, maybe to cope with the iffy item you replaced? ".......the hose connected to it was missing or broken" That could just about match all the symptoms.

Sorry if this is a stupid suggestion, but I don't know how familiar you are with all the bits and pieces on your bike. Apologies otherwise.

Because as others have said, if I pull the 4-pin from the tank my bike dies there and then. There's no 'minute' of running on....

The Mighty Gryphon · « **Reply #16 on:** September 08, 2015, 11:46:51 PM »

Quote from: Freelancer on September 08, 2015, 12:24:05 AM

Just a shot in the dark,

Could be the injectors. If they are shot/ stuck open to the point that fuel can gravity feed then the pump would flood out your bike.
Also, could be something in the control circuit that is locking the injectors in the open position.

I'd suggest looking at your spark plugs after the pump has been on. If they are wet then there is a big clue.

This sounds logical to me. Combined with intake vacuum to pull fuel through the injectors. Opening throttle leans the mixture to where it won't ignite. Might very well be enough fuel in the rail to run the engine at idle for several minutes.

Turning on the furl pump floods the engine through the open injectors. Whatever you do, I hope you keep us posted on what you find.

MicStanz · « **Reply #17 on:** September 09, 2015, 09:11:47 AM »

Putting the fuel pump and fuel pressure questions aside for the moment,
Unplug the wires to the >injectors< and see if it will run at all. if the injectors are whack, that will define whether it's mechanical (stuck open), or electronic/electrical (Motronic / FI control brain fart). Either way I find it pretty hard to believe all three injectors would do that at the same time unless they all sucked up some debris (that somehow would have had to get past the fuel filter). I don't know if that could mess up the FPR too. AFAIK, that's only a diaphragm and spring set-up. So I don't think there would be anything 'clog-able' in there, but ...Brings me back to the comment on the missing or broken hose and the gummy fuel pump isolator. (Maybe the PO had a clogged fuel filter and moused it to get home?)
If you could prove out the mechanical side (fuel pump [even hot-wired if necessary] will produce correct pressure [not under spec, not over spec...got a gauge?], hoses are all in the right place, and injectors arent jammed open, then it's probably safe to go after the electrical gremlins.
Just remember to follow the Professor Mayhem's First rule of troubleshooting (anything)... Never overlook the obvious.
Keep us posted...

cnoon · « **Reply #18 on:** September 10, 2015, 07:57:39 AM »

Thanks for the ideas, everyone. I plan to do a few more tests this weekend and will report my findings.

TrueAce · « **Reply #19 on:** September 10, 2015, 09:56:05 AM »

Since you've invited everyone to throw ideas on the wall ( maybe getting tiresome?), let's go back to your initial post "replaced fuel pump": did you use a stock or after-market, and could you test the pressure, should be about 65 lbs. to the pressure regulator, about 35 lb. to the rail. Seem you have a functioning pump, functioning injectors, but under a load greater than idle, the fuel supply is insufficient for the engine to rev.
Do let everyone know the solution to your puzzle!

cnoon · « **Reply #20 on:** September 10, 2015, 09:15:00 PM »

I checked out a few things this evening and here's the update.
- I started it up with the fuel pump unplugged at the 4-pin connector and it idled as it has been doing. I then connected it and it promptly died as before. I immediately pulled the #2 plug and it was very wet.
- I disconnected the forward gas line from the fuel injector rail (as pictured) and redirected it through a clear tubing and into a gas can. With the 4-pin disconnected, I started it and it idled as usual. I then plugged in the 4-pin and it kept idling but there was gas now flowing through the clear tubing. Keep in mind that with the fuel line disconnected from the rail, there was absolutely no positive pressure within the rail but the bike was still idling.

Based on the above, I'm beginning to think that all three injectors are wide open and the combination of gravity and cylinder vacuum are pulling any residual fuel from the rail and maintaining the idle. That would also cause immediate flooding when the system is all connected up.

So, what could cause all three injectors to be in the wide open position or perhaps something is telling the control unit to open them up (perhaps an air flow sensor)? Thanks!

TrueAce · « **Reply #21 on:** September 10, 2015, 09:41:28 PM »

Was gas flow through the clear tube under fuel pump level pressure? How about rate of flow through the hose to the rail?

cnoon · « **Reply #22 on:** September 10, 2015, 09:44:55 PM »

The fuel flow was brisk but I don't have a gauge (yet) to test pressure. I put in a brand new Bosch pump just last week.

The Mighty Gryphon · « **Reply #23 on:** September 10, 2015, 10:42:26 PM »

I think the no load flow of the fuel pump should be on the order of 1 liter in two minutes.

Next test is to unplug the injector wires and see if the engine still idles. If not, the problem is electrical, not stuck injectors. If it runs, the injectors need to be looked at.

cnoon · « **Reply #24 on:** September 12, 2015, 06:47:56 AM »

The replacement fuel pump was a Bosch exact, by the way, not an alternative requiring adaption. Given that fuel flowed in the right direction when the plug is connected suggests I have the pos/neg and fuel lines hooked up correctly. I'm going to check out the air sensor this weekend. I'm wondering if it is sending a signal that results in the injectors opening richer. I also just noticed there appears to a crack in the S-shaped vent hose towards the rear of the engine that seems to connect the air collector with the crankcase but not sure if that's enough to be the culprit. More experimenting this weekend. This has gotten personal.

News:

Author Topic: Fuel system system mystery (Read 21684 times)