From time to time, I will be posting a request for a diagnostic discussion on behalf of others. These posts will generally revolve around a broken vehicle that’s causing sleep loss for some anonymous soul. There are lost learning opportunities that result from the intimidation that comes with the thought of sticking our necks out and asking for help. This is the reason that I have chosen to do this.
I’m going to call this series, Dear Diag.Net
Please, be patient with me. I will do my best to be the liaison between the individual requesting this conversation and the members that chose to take part. By design, these posts will be more about learning than fixing the vehicle. We will try to put closure to as many of these discussions as possible but there will be times that doing so will not be possible.
My first post will be about a crank but no start on a 2012 Nissan Versa with a 1.6L HR16DE engine. We will call our anonymous friend Gandalf.
Full disclosure; Gandalf states that he may not get the opportunity to have any additional time with this vehicle. Either way, we thought that this was a good conversation piece and wanted to share.
The vehicle was purchased by a used car dealer as a crank but no start. The dealer installed a used engine and a PCM attempting to resolve this concern. Gandalf was asked to program the replacement PCM. Once they realized that the PCM didn’t make a difference, they asked Gandalf to perform some testing.
This is the shake down.
During the first effort, Gandalf performed the usual checks monitoring scan data. Seeing nothing out of the norm Gandalf started unplugging key sensors to make sure that there wasn’t anything odd that he didn’t catch while viewing scan data. He also verified that spark was present several times throughout the entire ordeal. Gandalf then acquired a cam/crank correlation but had a hard time getting a known good capture from the same model year. Next, he acquired an injector voltage capture and found that the injector on time didn’t seem sufficient at about 1.2-1.5ms. This engine uses two port injectors on each cylinder, so he did some investigating, attempting to gain some idea of how this dual fuel injector system functioned. Unfortunately, he didn’t find any useful information. He did find that the fuel injectors were firing sequentially during the crank but no start condition but both injectors for each cylinder were being commanded on at the same time. This didn’t agree with something that he and I found in service information. When I was looking at this with him I found a document suggesting that the injectors should be batched fired during engine cranking. This 2012 is not batch firing during these start attempts. Could this be a problem?
Gandalf had to move on to his other appointments, but he advised the dealer manager that he would be back for further testing.
During Gandalf’s preparation for his return visit, he could never find an exact match on the cam/crank correlation. He did however, find a known good Cam/Crank correlation from a 2015 1.6L Versa. The capture looked similar enough that he felt that it was most likely a good reference. That said, the decision was difficult to make because the cam sensor appeared to be inverted like Nissan’s with crank, but no starts often do. After some serious analyzing, he decided that the timing was most likely okay based off the 2015 Versa capture. When Gandalf returned, he found that the dealer manager had his team install another used engine because they were suspecting that the valve timing was off. They also replaced the cam sensor due to the seemingly inverted signal. Neither service made a difference. Gandalf decided to acquire the injector events again. When looking at the injector waveforms from his first trip, there appeared to be a pintle hump but he never saw any indication of a pintle hump during the second trip. He also found that the vehicle would run with an alternative fuel source. Every injector current or voltage trace he acquired now seemed to have a 1.5ms or less on time with no sign of a pintle hump. He removed the spark plugs and didn’t see any evidence that fuel was making its way into the cylinder.
To recap, Gandalf never had any fault codes that would give him any diagnostic direction,
He never saw any data that looked suspicious other than the seemingly low 1.2-1.5ms injector Pulse width. He felt that a 1.2-1.5ms pulse width seemed low even with two injectors per cylinder during a crank but no start condition. He suspects that the injectors are not on long enough to open and deliver fuel to the cylinders. He didn’t see any signs of pintle hump, the spark plugs looked dry and the vehicle ran on an alternative fuel source.
Monitoring and unplugging sensors that are key in the base pulse width calculation didn’t make any difference. Scan data matched the 1.2-1.5ms pulse width shown on the injector scope captures.
The PCM powers and grounds were verified.
What do you think? Gandalf suspects that there is either some compatibility concern or some kind of PCM strategy is allowing the PCM to command the injectors but not long enough to actually deliver fuel into the cylinders.
Well, that's the story. I look forward to some discussion on this one. Please jump in and share your thoughts.
Just a stab in the dark, but has Gandalf had the fuel tested? Diesel, Methanol, Ethanol, Water all could make the vehicle hard to start. Will it start when adding propane or other flammable?
it says " He also found that the vehicle would run with an alternative fuel source".
I was thinking the same thing. Either it's bad/not gas, or it's all plugged up somehow.
Good thought Michael. I just checked with him and he did say that the fuel was questionable because the vehicle has been sitting for months due to the crank but no start condition. He didn't perform any formal test but he verified that the fuel was flammable. He recommended replacing the fuel but ultimately, he's pretty sure fuel isn't making its way into the combustion chamber.
The injector wave forms show no pintle hump and the spark plugs showed no evidence of fuel.
Jaime, thanks for asking. I'm sorry forgetting to include that detail. I just spoke to Gandalf. He did check fuel pressure and it was 60 psi. He did state that he didn't check specifications and assumed that 60psi was adequate.
Just spitballing here Brin, is that maf dirty? Could be under reporting intake and as a result cutting down injector open times. Any data with key on that may be out of specs? And I know this sounds trivial, and not trying to be rude, but is he absolutely sure that there's good clean power source? These cars are real finnicky with their source voltage. (Totally missed the unplugging key sensors, disregard maf)
Mario, thanks for jumping in. He didn't mention the condition of the battery but he did mention using either a jump box or a iota maintainer during diagnostics.
Yes, specifically there were no G/S displayed koeo and he thinks he remembers 1.5 G/S or so during cranking.
He did unplug sensors one at a time, including the maf, while attempting to start the vehicle with no luck.
Well at this point I'd say the biggest determining factor would be wether or not his known good correlation waveform is to be trusted. Time and time again I see that correlation varies from year to year. I'd hunt down that same year model known good correlation waveform. Looking at the known good and his capture side by side, it almost seems as if his wiring was questionable, plus the obvious discrepancy between the "teeth." Id also confirm that his positive, ground, and signals are in the correct order at the sensor connector. Aftermarket sensor? Nonetheless a proper known good is in order. I still have the "runs with alternative fuel" in the back of my mind though. This should be interesting. So I'm sticking with a key input issue, or questionable source voltage and ground to the pcm(load tested?).
I agree, the captures look close enough to be the same but I would feel much better if we could verify. The good news is that there's another 1.6l Versa on his schedule tomorrow. He's hoping that he can get that job and acquire that wave form.
Try swapping the Intake cam and exhaust cam position sensor connectors over. A fellow tech at my work did some repair to the vehicle, can’t remember what it was. But he plugged the sensors in backwards. The vehicle would not start, and there were no codes. Attached is a know good cam/crank for this vehicle. I see the intake cam signal that was captured by Gandalf, has an issue on the first pulse being wider than the rest.
Great point. I spoke with Gandalf and he said that he did acquire his cam captures while probed near the sensor. He said that he usually verifies wire color to wiring diagrams and he's pretty confident that he did so in this case. I can certainly see how this could cause some headaches if the connectors are very close together.
I'm sure that your comment has him thinking but his initial response is that he was pretty sure he verified wire colors on the connectors were matching the appropriate sensors.
Thanks for that capture. I'm on a phone now but he and I will take a better look when we can.
I subscribe to a different philosophy than Gandolf did when you wrote "he did acquire his cam captures while probed near the sensor". Many times I've seen a perfect signal while testing at a sensor yet not at the PCM. I suggest the signal acquisition occur at the module whenever possible.
I spoke with Gandalf about this. He said that he acquired the CKP at the ECM but the Cam signals from the sensors. He stated that he realized that he should have acquired the cam sensor traces from the ECM but the combination of how difficult it was to access the module, how easy it was to access the sensors and the fact that the vehicle wasn't coding for Cam sensor faults motivated him to make the decision to acquire the signal the way he did. In hindsight, he does agree and hopes to get back to the vehicle to check.
Since the injectors aren't batch firing, I wonder if the PCM doesn't realize that the engine is cranking. Is there a PID for that?
I'm not sure. I'll ask. I do know that he paid attention to rpm while cranking and it seemed accurate.
Since the vehicle runs off an alternative fuel source in my opinion that rules out timing especially if it runs half way decent. First I would like to see the wps on the fuel rail and use an injector signal for sync and see if Gandalf is able to see the injectors pulse. If so then my next check would be to disconnect the fuel system and use one of the fuel system cleaning tools that hooks to the rail and pressure can be dialed in. That would eliminate the fuel possibility without having to drain the tank. I don't think you have a fuel problem but that would take it off the table completely.
In my opinion the only thing left at this point is injection strategy. The fact that the cam sensor has been replaced and the engine has been replaced twice as well as the pcm means something has been continually causing the strategy to be screwed up. I have seen cam sensor patterns upside down caused by the incorrect cam sensor so I'd want to make 100% sure the cam and crank sensors are the correct ones. I'd make sure the maf is clean although I don't think it's a problem I want to take it off the table. I would make sure the IAT and ECT are correct after a cold soak to make sure they aren't changing the strategy. I'd also like to see when the injectors are firing compared to TDC. Next I'd want to make sure cranking speed was the same compared to a know good engine. I don't think this will be a problem either but again I want to take it off the table.
Wow, I just hit the post button on my message and then saw yours. Looks like we had some of the same ideas.
Since the engine starts with an alternative fuel source, that eliminates a lot of things such as engine mechanical and the ignition system. I would definitely focus on the fuel delivery system.
Is has to be either BAD fuel or a fault with the injection system. It would be nice if good fuel could be pumped into the fuel rail, possibly with a injector cleaning machine. I have done that in the past so I could drive a vehicle with a bad fuel pump into the shop.
The injector current ramps also suggest a problem. Usually pulse width cranking will be very high in the 15-20ms range. Could the system be in a fuel cut mode? Immobilizer issue perhaps?
With used cars the possibilities can be endless, especially after multiple engine changes and god knows what else.
That was my biggest take away here. Gandalf has three or four injector voltage and current files. All of them indicate that the injector is not opening. I'm not sure that I've seen all injectors being turned on without opening on any vehicle that I've worked on. We wonder if this is due to the low pulse width. I agree, PW during cranking should be much longer in my experience. Should this vehicle be batch firing 8 injectors 2 times per 4 strokes like service information seems to suggest? If so, that would give us about 6ms of pulse width per cylinder during a four stroke cycle on a 1.6L engine.
The second image you posted in the original post, the one of the injector voltage, shows the pintle hump. The pintle HAS to be moving in order for that to appear (ie - influence the voltage). The images you speak of now, may be indicating no pintle movement, which may have developed since the engine swap (harness not connected properly?) and if corrected, may resolve the original complaint as well.
Yes, in the original post I mentioned that he did see a pintle hump in his 1st capture on his 1st trip. On his 2nd trip, he acquired several injector captures and none of them were showing pintle humps.
I agree, you know the drill. At this point, he could be dealing with more than one concern and some of them are most likely self induced from the dealers technicians.
Has this vehicle been repaired yet? I'm curious as to what is happening.
John, I can't guarantee the outcome in this case. I mentioned in the original post that I couldn't guarantee that Gandalf was going to have the opportunity to put closure on this one. That said, I think he's going to do his best. The dealer made it clear that they were not going to authorize any additional testing. If Gandalf goes back out, it'll most likely be a voluntary move on his part. He's no different then us in this regard so I know he's wanting the opportunity to see it through but it might not be up to him.
I'm sure I'll here from him throughout the next day or two and I'll be sure to keep you updated.
Honestly, my next step is old school. You said that he checked fuel pressure. With the gauge already hooked up, I would have powered up at least one injector with my power probe, key off, just to see if they would cause a pressure drop. If no pressure drop occurs, I would then be looking at pulling the fuel rail. If pressure drop occurs, then I would continue focusing on the short pulse width, or fuel quality.
I would also look at vacuum cranking. To me, 1.5 gm/s seems a little on the low side for the maf. What does it read running on alternate fuel?
I love the manual operation of the injectors idea. So simple but effective.
I don't have an issue with the g/s when cranking. Traditionally at 500 rpm you could expect 1 g/s per liter. That's why a rough rule of thumb is that you should be a little over 1 g/s per liter at a hot idle. I believe this engine was cranking at between 2-300 rpm's.
I'll ask if he paid attention to g/s when it ran. I'm not sure.
The manual operation of the injectors is something I have done in the past, when I suspect an injector flow issue with one injector. I will pump the fuel pressure back up between each, and check the pressure drop on the injectors while powering them up for 1 second each. I have caught clogged injector screens on multiple occasions that way.
Are we sure the used car lot installed the correct year engine? Correct pcm? Any way to verify this? For clarification, when you said "runs on alternative fuel" I'm assuming spraying in the intake? It wasn't attempted with the rail pressurized, taking the vehicles fuel system out of the question?
Hey Brin great write up. I noticed this in reading the story. " When Gandalf returned, he found that the dealer manager had his team install another used engine because they were suspecting that the valve timing was off. They also replaced the cam sensor due to the seemingly inverted signal. Neither service made a difference. Gandalf decided to acquire the injector events again. When looking at the injector waveforms from his first trip, there appeared to be a pintle hump but he never saw any indication of a pintle hump during the second trip. He also found that the vehicle would run with an alternative fuel source. Every injector current or voltage trace he acquired now seemed to have a 1.5ms or less on time with no sign of a pintle hump. He removed the spark plugs and didn’t see any evidence that fuel was making its way into the cylinder." Originally it was a no start with pintle hump and a possible bad CMP sensor. When he returned it was a no start with a "different engine" and a new CMP sensor. It's possible that the injectors in the second engine may be stuck closed not allowing fuel flow. Still doesn't explain the short injector pulse though. Kind of hard to diagnose a problem when the components keep changing. The only constant through this ordeal is the wiring on the vehicle. 3 engines all with short injector pulse. Maybe there is some sort of anti theft device on the vehicle or maybe a remote starter with some sort of start disable strategy. What about wide open throttle fuel cut? TP APP