firstly wanted to say great job with what you've done so far. very logical approach to repairing this.

its possible that when your old pump went it created a ground fault...the possibilities are endless there. one question: when the pump was replaced did you have the battery disconnected? you may have inadvertently caused a shot which may have damaged the ECU. also possible wiring has been damaged.

to that end,take a look at the wiring under the dash where the ECU is....you might get lucky and have a damaged wire or two, but i'm thinking the ECU may be done/

 

The fuel pump will not continue to run after it initially pressures the system if the computer does not get a signal that the engine is running. So take a look at your CKS (crankshaft position sensor) and CPS (camshaft position sensor). These are easy to access on the 1.6L engine, so check the wiring condition, perhaps the cleanliness of the probe and the connector. If you know someone with similar engine it would be an easy swap to test this, but the CKS is not expensive. Off hand, not sure of CPS cost.

 

I think I did leave the battery connected while I did the swap. Before the swap, the car would just crank and never catch, and that's when I noticed that the pump wasn't coming on. I also tried to run an OBD scan before I did the swap to a new pump, and it wasn't reading any codes. I didn't put a lot of thought into it at the time, but the OBD was getting power, just not reading any codes. I assumed that it was because the car didn't throw a code, but it could just as easily have been the same failure to pick up a signal that I'm having now.

Is there an "oldschool" method to check that the ECU is functional? I seem to remember a method that involved the length of time the CEL is illuminated when the key is turned, but that might have been specific to the Terceleporter I used to have.

One of the things that I checked after doing the work and observing the stalling characteristics was both the CKS and CPS. I had to replace the crank sensor a little while ago when the engine would stall at temperature, or else fail to start again if it had been parked for a short while and still close to running temperature. It would shut off the engine without a hiccup, but not throw a code. The OBD reader did work at this time. The old sensor was definitely faulty, and the new one was good when I double checked it yesterday. I did move the induct that sits beside the battery to get better access to the CPK, so I checked around it to make sure that I didn't knock anything else loose. I will re-re-re-check the sensors though.

 

It finally connected to the OBD after a long, cold wait. It lit up to tell me that there were no trouble codes, but the yellow light indicates that the results aren't completely awesome.

The M(Misfire monitor), F(Fuel system monitor), and CC(Comprehensive Component Monitor) light up solid, indicating that they have finished the self-check.

The C(cat), O(O2 sens.), OH(O2 sens. heater) and EVAP monitor are flashing, which, according to the manual, means that they haven't yet finished their diagnostic self-test.

I left the scanner plugged in for a while with C, O, OH and EVAP flashing, and it beeped and indicated that it was reading again, but it came back with the same monitors flashing. The scanner cycles this same sequence over and over again until you're too cold to sit in your car.

Correct me if I'm wrong, but it seems promising(in terms of being able to narrow it down) that the outstanding monitor checks are all to do with exhaust and EVAP? On the other hand, these might be the kind of sensors that require x number of drive cycles since codes last cleared or battery disconnected, and since the car will not stay running, they'll continue to flash..?

 

i have a similar thing on my code scanner, 'emissions status' or some crap. that light is pretty much useless. you can ignore it. without a DTC, its meaningless.

you are correct to say that there are certain parameters in the ECU that require a few drive cycles to get them fully tested properly. if you've disconnected the battery, yes it will require a few drive cycles. if not then it should not be an issue.

as to if the ECU is properly functioning or not...its hard to say at this point. you got the scanner to check, but it was kind of abnormal....if i were in your shoes, i would take my car to a stealership and ask them to check the computer is fully functional and nothing more. one of the few times i'd pay them a visit.

 

Hi guys, the first thing that came into my mind when reading this one was the fuel pressure regulator. Does your engine have a pressure regulator on the fuel rail or is it a returnless fuel system? If it's a returnless system, did you replace the whole pump assembly with the fuel level sender etc or just the electric pump? The reason I ask is that in a returnless fuel system the pressure regulator is inside the pump assembly. If you replaced the whole tank unit it may have the wrong (or no) fuel pressure regulator.

Also, is the scan tool communicating properly with the engine PCM?
Confirm by reading engine RPM in live data. If live data works, check the B1S1 voltage while it's running to confirm if it is going very rich when it cuts out. (that's your suspicion..right?) While your at it take a look at your coolant temp and make sure it's showing a plausible value.

Regards.

Scottie.

 

I removed the entire in-tank apparatus, replaced just the dead electric fuel pump with a new Bosch, and then placed the apparatus back into the tank after inspecting it.

The scanner doesn't have live data functionality, and there are no codes when it dies out.

I'm actually beginning to suspect that it is a camshaft position sensor, as David C suggested. In an effort to keep the engine running a little longer than 3-5 seconds and hopefully throw a code, I tried pressing the accelerator as soon as it fires. If I hold it down at about 25% throttle, the RPMs begin to fall after 3 seconds, all the way down until it sputters and dies. It is still delivering spark, but no fuel.

When my crank sensor went bad, it didn't send any codes either. I'm dealing with complete blizzard conditions right now, so there isn't a lot I can do, but I will hopefully narrow it down to the cam sensor, or the connection, wiring, etc. I took the sensor out and cleaned it up, but it had no effect. The thing I remember noticing the night before this stalling issue is that the car took a bit more cranking before it caught. I automatically let go of the key when I normally would, and the engine hadn't caught yet. The fuel pump failure is looking more and more coincidental.

I appreciate all of the suggestions, and I feel more confident that I will be able to solve this issue soon. I am going to take a recording of the ignition and stalling behaviour, and if I get it fixed, hopefully it'll help someone else in the future.

 

If your getting spark it's highly unlikely to be a cam/crank sensor issue. When you say you have no fuel what exactly do you mean? In your original post you seemed to suggest it was over fueling. It seems strange that your car was driving fine before the pump died. You replaced the pump and your now having a suspected fuel delivery issue that wasn't there before you replaced the fuel pump.

 

yeah more info....its hard to say that it would be the regulator as that is integrated with the pump in this design. its possible the new one is a dud, but the only way to test that is with a fuel pressure test gauge on the fuel rail. it would explain an erratic fueling problem though.

the fact that the pump died suddenly doesn't tell us anything spark. pumps can die suddenly regardless of if its the usual or not.

i'd be checking the install over thoroughly before blaming this on the new pump, but its a good possibility. i've had defective parts myself a few times.

and i hear ya on the blizzard....oh man. shoved a good foot of snow this morning and the crap is still coming down!

 

2004 Hyundai Accent 1.6 Stalling - YouTube

There's a video of the stalling behaviour. Video begins just after the 2 second fuel pump priming. I tried to boost the volume so the behaviour would be audible. My only inputs are clutch and ignition. This is exactly what the car does consistently.

I had a suspicion that the engine was running rich because of the smoky exhaust and the slight backfiring it does if I try to keep it alive with the throttle.

Regards

 

I had a suspicion that the engine was running rich because of the smoky exhaust and the slight backfiring it does if I try to keep it alive with the throttle.

this is the kind of information we need if you want us to help you. i didn't see anywhere in your previous posts that this was mentioned.

given that, yes i would agree its likely running rich and may have a fueling issue. once you've checked everything over, maybe see if you can get the new pump replaced under warranty.

 

Good evening gents, I think this shows the importance of being able to use all you senses when diagnosing a fault. It was a great idea putting your clip on youtube.

Listening to the engine in the video clip, it sounds like it is fuel starvation that's the problem rather than too much fuel. What I think I would do is confirm that the fuel pump is being supplied with power after the initial crank. My suspicion is that the pump is operating to prime the fuel rail but isn't actually running when the engine starts. The engine will run for a couple of seconds on the fuel that is in the rail (from the initial prime) but dies once that fuel has been injected.

My suggestion for now before doing anything else is to connect a volt meter (or even better, a 12V test lamp) to the fuel pump test connection in the engine bay fusebox. Check that there is 12V being supplied to the pump during and after engine crank.

Scott.

 

plugs should reveal weather this is running rich or lean. checking the electrics is always a good idea, but if we have a pressure regulation problem or a problem of the pump cannot supply sufficient pressure a test gauge is the only way to show that.

 

For the plugs to tell us anything about the mixture would the engine not have to have been running for a while? His engine has probably only ran for a minute or so since the fault has appeared, so do you think the plugs could tell us anything meaningful in that short space of time?

I'm guessing he hasn't got a fuel pressure gauge or he would have already done a pressure test, so I suggest doing the electrical test first because he is likely to have the tools required for that to hand.

But let's say he follows your advice and buys a pressure gauge. He does a pressure test and it shows that the pump isn't supplying fuel pressure to the rail. If that was the case, what would you advise him to test next?

 

Simple. It's a new Bosch pump with a warranty. Beyond a simple leak search and wiring/electrical test - which we already covered in previous posts- it's time to use that warranty.

 

Beyond a simple leak search and wiring/electrical test - which we already covered in previous posts- it's time to use that warranty.

OK, for now let's pretend that he bought a pressure gauge, done the test and he found that the pump wasn't supplying the rail when the engine was running. Your now happy to pull the pump and return it for warranty at this stage?

 

Please excuse my crappy drawing, but I think it shows clearly enough what terminals in the fusebox to bridge together. Remove the fuel pump relay and bridge the terminals on the relay socket in the fusebox. That will power up the fuel pump and may let the engine run. Worth a try anyway.

 

I tested out jumping the fuel pump relay and it ran the pump constantly (the original pump that I replaced would not run at all like this). The battery in the car has drained to the point that it won't turn over and catch for a few seconds like it had been before, so I'm going to put a charge on it before testing, so as to avoid inaccurate results.

After some earlier testing, it doesn't appear that the pump is receiving the proper voltage for priming on a consistent basis. Not sure if this was a matter of battery drain, but I tried swapping identical relays into the fuel pump slot and it had no effect on consistency--if it was a matter of a bad relay, the swap should have had an effect, right? Since the bridge successfully brings the fuel pump on constantly on battery power, the PCU could be sending no/incorrect power to the coil of the fuel pump relay? I wonder if this inconsistency may provide a clue?

I'm going to try to borrow a fuel pressure gauge tomorrow and see what that reveals.

Thanks for all the help AUTOSPARK, zero_gravity, et al.

Regards,
John

 

but I tried swapping identical relays into the fuel pump slot and it had no effect on consistency

Total shot in the dark here.... Try swapping the fuel pump relay with one that doesn't matter like the horn.

 

John, I think your right, the PCU isn't sending a control signal to the pump relay.
But it's the PCU that controls the relay during the initial prime, so it is capable of sending the signal. That tells us that it's likely to be a fault on the input side of the PCU.

Someone mentioned the cam and crank sensors near the start of the thread (David C) and I kinda dismissed those at the time because the description you gave was of over fueling and you said you had spark. Personally, I've never came across a Hyundai engine that can control the ignition coils with no crank sensor signal. Anyway, we now know that over fueling isn't what's happening and I'm now thinking that David was right on the money with his suggestion. Perhaps when you bypass the pump relay once the battery is recharged, the engine will run long enough to log a trouble code.

I'm guessing that the PCU isn't seeing a crankshaft sensor signal. For safety reasons, the fuel pump won't run in the absence of a crankshaft signal. If the car is in an accident and a fuel line bursts, when the engine dies due to lack of fuel the PCU will switch off the pump because there will no longer be a crankshaft signal input.

Let us know how it goes with the relay bypass and if you manage to get a code out of it.

Scott.

 

This reminds me of something that happened with my dads Renault Safrane. We had a short hill to get out of the drive and the poor thing couldnt even get up it - it sounded like a bag of hammers in a cement mixer, idle was fine, but any attempt to rev the engine did exactly what you describe - with a noise that would make even a grandmother wince...
The problem? a broken oxygen sensor - which I assume would either be something to do with the catalyst circuit or the intake sensors. From memory it was one of the sensors on the intake side.
When the sensor was replaced it was entirely back to normal - but it wasn't something that even the engineers expected from what the car was doing... a case of too many sensors spoils the broth..
The fact you mention the OBD - you might have a problem with wiring, which would mean the wiring in other connectors from the ECU might be suspect (= nightmare) - but the Bluetooth OBD dongles are notorious for being useless to a greater or lesser extent..

The problem you have with this are there are so many possible lines of enquiry and it may be just luck that you find, or dont find, the particular thing thats causing the problems.

 

The problem you have with this are there are so many possible lines of enquiry and it may be just luck that you find, or dont find, the particular thing thats causing the problems.

Hello Rosie. Your right, that's one of the challenges of diagnosing these types of faults. There are so many different sensors and many of them interact in some way making difficult to determine if what your seeing is the actual fault or just a symptom of an issue somewhere else in the chain.

I think we have a pretty good idea of what's happening with this one now though. Just a pitty we are at post 25 or 26 in this thread and someone told us what was wrong way back in post 3. I think we should have listened to David_C:laugh:

Scott.

 

Total shot in the dark here.... Try swapping the fuel pump relay with one that doesn't matter like the horn. There's nothing so far to say that the ecu isn't trying to keep the pump running and the relay coil just won't stay on for some reason.

5 minute test and its free. Not much to lose here.

 

Friends don't let friends buy aftermarket crankshaft position sensors.

I took a day off to charge the battery, test the fuel pressure, send off some law school applications, and write a paper.

With a full charge on the battery, the issue eventually changed from a start-stall, to a cranking no start. I decided to take the ECU/Crank sensor fuel cutoff theory and get intimate with the crank sensor. Resistance was within spec, so I took it out to clean and inspect. I noticed that the flange on it was slightly offset compared to the original--enough to have the actual pickup sit about .4mm further away from the wheel inside the engine, so I decided to re-install the crank sensor without the o-ring and test that theory. The filthy ******* started up right away, stronger than it ever has for me, idled, and runs forever.

I wanted to thank you all for the help, suggestions, etc. I was getting extremely frustrated with this problem, and I'm very, very relieved to have it sorted.

I don't feel comfortable running without the o-ring, so I'm going to pick up a crank sensor directly from Hyundai. I doubt I will ever use an aftermarket electronic or powertrain component again. The variety of symptoms that occur when a sensor is slightly out of spec amazes me. Then again, I come from the simplicity of a small engine racing background(karting).

Thanks again!
John

 

Hi John.

Thanks for posting back in with the fix. I'm glad to hear you finally got it going.

It was your video clip that put me onto the what was happening. Sometimes your eyes, ears and nose are the best diagnostic aids to have. And I've now learned that you can have spark with no crank signal to the computer. That's a new one to me, so I've have to file that away for next time. It seems an odd design though. Why would the computer create a spark but not run the fuel pump to give the spark something to ignite?

Anyway, all the best. Speak to you next time it breaks down

Scott.

 

Inside the dash bord... There is another relay,its located in the driver side in the botom ,,under the stering wheel,,in the left side
i think the relay is bad
--let me look for the video in youtube

 

THANK GOD IT WORK ...LET ME PUT THIS VIDEO MAY HELP OTHERS
WHEN YOUR HYUNDAI ACCENT WONT START

 

following this thread

 

following this thread

 

great posts guys. Thought Id sit this one out. Some don't know this but to add one last thing. It is important you TEST the amp draw to spec on the new pump before running it. A fuel pump that draws more than the spec for that pump can damaged your computer. Always test this part before you hook it up. You can use a DVOM in amp mode or use a coil type meter or even a scope with an amp probe to read the amp draw. I have seen first hand a person who replaced 2 computers and found out it was the amp draw on the fuel pump causing the computers to fry


Good job!

 

I have seen first hand a person who replaced 2 computers and found out it was the amp draw on the fuel pump causing the computers to fry

Thanks for the heads-up on that. I've never heard that before...every days a school day in the motor trade.

How does the current draw of the pump fry the engine computer?

Scottie.

 

Think this is the video where it mentioned this


https://www.youtube.com/watch?v=ohS2FSIc1iU