Friday, January 31, 2020

2014 Nissan Altima S 2.5 Stalls At Idle

One of my good customers called with a description of stalling after a start at idle, battery light and a lot of corrosion at the battery terminals. I thought the stalling problem might be too low an idle rpm and suggested the one foot for braking and one foot for keeping it just off idle to raise the rpms. That worked well enough to get the car to the shop. When I drove the car inside there was no stalling or low rpm. In fact it ran very well but there was a check engine light. A light means a code. A code might be a clue to what had been going on.


A P0101 MAF sensor code was stored and actually is a pretty common fault code on these cars. So common in fact there is a technical service bulletin that suggests an updated PCV valve and a possible ecm flash update, depending on ecm part number.  NHTSA ID: 10119177
TSB ID: NTB16-030a 
That being said, nothing about the known P0101 problem would explain the stalling complaint. Also, the manner of driving with one foot on brake and the other on the accelerator could have set the code. In the few minutes I was with key on checking for codes the car battery voltage had lowered enough that the car powered down to save battery voltage. This brought me to what Danny (customer) had said about the battery condition. I decided to let the P0101 aside for now. I had another idea about the original complaint and its cause. I should explain that a car has a learned memory or an updated one that stores slight adjustments over time to compensate for wear or tear that has taken place since it has been getting older. If you reset the memory it goes back to before the learned behaviors and then has to relearn the adapts. One very common thing to happen over time is a buildup of carbon at the air intake throttle body and throttle plate. The buildup was slow and the idle strategy adapted to it to keep the car idling at a proper rpm. Clearing the adapts on a car that has a dirty throttle (carbon) will cause the car to stall at idle until the relearn adapts OR until you clean the carbon. If you clean the throttle the idle will not have to be relearned but will already match the reset strategy. I suggest cleaning the throttle after a reset. 
The battery here had not been disconnected and the battery had not failed to crank the car. You might wonder then why am I thinking of a problem with throttle carbon buildup and memory resets. A weak battery can drop voltage low enough during a crank to reset the learned adapts. Seriously. I've seen it more on a Dodge than other models so they may be more sensitive to it. It isn't just reset either. Some codes can set on a start that were a result of battery voltage dropping too much during an engine crank. The rule of thumb here, for me anyway, is that if the voltage drops lower than 10 volts during cranking its a problem. 
The plan was now to see if this battery was dropping too low on a crank and if the throttle plate was carbon dirty. Those two things and I'd have evidence this was the problem. 


The battery voltage starts at the left, the drop is down to 7.84 during the crank and then once the engine starts the voltage is output of the alternator. So yes the voltage is going low enough to reset the adapts and I do need a battery. But I need to see carbon buildup on the throttle to explain the stalling. 


Especially sensitive to any carbon is the edge of that throttle plate. Cleaning the throttle and replacing the battery should be the fix. At least it can't hurt. 


Cranking voltage with a new battery.


The P0101. I cleared the code and after 3 trips it had not reappeared even as a pending code. I decided not to pursue further. 

Thanks for reading!

Kenny@ggauto.repair


Thursday, January 30, 2020

'92 Astro Van 4.3Z (repost from 2008)

Diagnostic Tech: '92 Astro Van, 4.3Z

I had a '92 Astro Van come in on the hook for a no-start. The engine was a 4.3 with throttle body fuel injection. The owner said the van had been stalling and sometimes failing to start for some time now. He had replaced the fuel pump and filter because he though it seemed to be running out of fuel. No testing had been done and the same problem remained after the fuel pump replacement. When I turned the key on I could hear the fuel pump run. Turning the key to start engaged the starter, the engine spun nicely and .... it even started right up! It isn't uncommon for a no-start to be towed in and then start once it gets here. I'm sure there is an applicable Murphy's law for that.

What I did notice though was an intermittent engine cut-out. A sure fire sign of a secondary ignition problem. Even though the van owner was still sure he had a fuel problem, I decided to take a look at the ignition system. On the Astro Van the ignition system is easy to access once the engine cover is removed. I immediately saw what was probably causing the cut-out symptom I had noticed. There was a burn mark where the secondary spark from the coil wire had been jumping to a vacuum hose instead of following the intended path to the distributor cap. That meant the resistance in the secondary, downstream of the coil wire was so great that the spark found it easier to jump to the vacuum hose. A mental list of plugs, wires, cap and rotor was forming in my mind.



As I began the work of replacing the secondary ignition parts I began to be amazed that this engine was running at all. The coil wire tower in the cap was completely burned away.


The rotor was burned through and the distributor was severely corroded.



After a cleaning with electrical contact cleaner and penetrating oil the distributor was found to be in good condition. I replaced the spark plugs, spark plug wires, distributor cap and rotor. The van ran very well after that with no more stalling or cutting out. At 264,000 miles I wouldn't be surprised if it did need that fuel pump as well.

Kenneth Hayes
kenny@ggauto.repair

Originally posted 2008 for carfix online.

Tuesday, January 14, 2020

1995 Ford Checks: OBDI EECIV

1995 Ford F250 XL w/5.8 engine uses the EECIV engine control system. If you are going to use the EECIV system for self tests you need to know a bit about the system. The first thing to do is have the engine at operating temperature, better known as warmed up. Before doing so, remove the radiator cap and check the radiator is full. This particular truck was low of coolant, about a half gallon low. Pressure testing showed some slight coolant seepage at the radiator and the intake manifold. For now we will fill the radiator, start the engine and let the coolant temperature reach normal operating temperature in preparation for the self tests.



Ford self-tests on this system start with KOEO, key on engine off tests. Trigger the test and the truck runs a series of pre-programmed tests and outputs the results. Other than your beginning the sequence and listening to the noises as the tests proceed you are just an observer. Think about it though. KOEO codes/results are LIVE. Any results are for the current time and the current tests. 
Let's run through a KOEO test. The diagnostic connector is at the fender, under the hood and I do have a compatible scan tool. 



The engine is warm, the coolant level is full, the oil level is full and the battery is charged. Those are the prerequisites for tests.


 What are fast codes? If you run KOEO in fast codes you won't see the code output. The scan tool will capture the codes and place them in the "review codes" section for your later perusal. 




Slow codes is more like counting light flashes. You can count the pulses as the code is output. Here we have 1 pulse, pause, 1 pulse, pause, 1 pulse or 1 1 1. 111 System pass. The initial single pulse before the slow code output is the fast code output signal.



Cool huh? Now that you understand KOEO testing, I'll just mention there is more to it. There are actually two parts to KOEO tests. The first part, the part that outputs live codes is called KOEO On Demand. On Demand is the live codes for faults present at the time of the test. The second part of the KOEO test is called Keep Alive and is basically memory codes stored from previous faults. Keep Alive codes are output at the end of the On Demand codes. If you were counting pulses (slow codes) you would see a single separator pulse followed by any stored Keep Alive codes.



Make note of your Keep Alive codes because they are fragile. Sometimes just ending the test will remove the code so remember it had a code in memory and in this case it was 123. Now because Ford doesn't trust us to know the details of the self tests that were performed and because we got a system pass in the on demand codes we make a reminder that we have a 123 in keep alive and move on to KOER code tests because that is the proper sequence as designed. If there were on demand codes you would diagnose and repair those first then start over with the KOEO tests.
KOER tests also require the warmed engine, charged battery and full fluids. Most scan tools will request you to run the engine 3 minutes at 2000 rpm just prior to the test. You'll need to turn off accessories like heater/air to prevent false codes. If your scan tool is worth using it will walk you through the tests because KOEO tests are more interactive than the previous. At the prompts, run the 2000 rpm event, shut off the engine, start the engine, there will be a series of beeps to identify the engine cylinder number, 4 beeps for this 8 cyl., after the cyl id press the brake and release, turn steering wheel half turn, press the cruise set switch and the overdrive switch (as applicable). Then relax for a bit. The rpms will drop very low and a few seconds later you will be asked to "goose the throttle" which is ford tech speak for punch the accelerator to the floor and release very quickly. The tests will complete and you will get results. In this case I got another 111 system pass.
So now I have passes in both KOEO On Demand and KOER tests. The truck complaint was a hunting idle rpm and a check engine light. On the way to the shop though it ran fine and the light was out. So far that is the same here even during the checks. We have the 123 code in Keep Alive and by the book that is where to go but I suspect this problem is intermittent and the code may not be helpful. 123 would be the throttle sensor signal voltage too high which apparently was not the case in KOEO On Demand or during KOER since it ran fine.
Next step then, see what the tps signal voltage is. Not all old Fords have a datastream available but this one does.


Let me quickly touch on datastream. The scan tool has tapped into some of the data available from the engine computer. The data exchange doesn't look that nicely sorted and displayed between the two computers (ecm and scan). It would look like this screen capture of the exchange.


Within that datastream is what the engine computer thinks is going on and the information is being presented to you by the scan tool in a format you can understand. Before datastreaming you had to measure signals at the engine computer connector. You have to bear in mind when looking at scan data that it isn't necessarily the actual sensor signal present at the connector. For that you'd need to verify with a meter at the connector. What the scan data strength is sometimes though is that you are looking at what the computer sees going on, right or wrong. That insight can be helpful because you won't measure it, you have to see it in the scan data.
It can be hard to reach the computer connector, use a meter and properly identify the pins. Here is a Ford *breakout box* to help in that task. Connect the box between the ecm and ecm connector then use the box to access pin circuits.


Now, you won't see scan data without a scan tool but you can measure pin voltages at the ecm with a meter. You can also run Ford self tests without a scan tool. You can use a voltmeter. Remember all those pulses during the scan tool code tests? Works the same way with a voltmeter but you'll need to know what the pulses mean when the needle sweeps. Hook the positive lead of your voltmeter to the positive battery terminal.  Hook your negative lead to the STO terminal of the diagnostic connector. Use a jumper wire to connect the STI terminal to the Signal Return terminal. 


When you turn the key on the KOEO test will start and the pulses will become needle sweeps. The first needle jump will be the fast codes, followed by (in this case) sweep, sweep, sweep. 1 1 1 or the 111 pass. Two 111 code sweeps followed by the separator sweep and the KOEO Keep Alive codes. You can even run the KOER tests. As I said though you'll need to know what is expected of you during the tests. Still, very neat. Here's the initial KOEO sweep, fast codes followed by 111. 



If you saw my post on quora about reading GM codes with a paper clip well you can also do that with a Ford however it is more complicated because it triggers the Ford self tests which as you have seen are more than just code flashes. You need a wire paper clip. No voltmeter or jumper or scan tool. Bend the paper clip into a U shape and use it to connect the STI and Signal Return.


Now when you start your KOEO self tests you'll be watching the check engine light flashes instead of pulses or needle sweeps.


Hope you find that interesting or at least informative! Back to the Ford truck problem though. I haven't seen the truck do anything wrong and the throttle position sensor output is fine. I'm not out of options quite yet.


Ford has more tests available so there is a bit of wiggle room.


Sorry.
Wiggle test state lets you shake, pull and tug on wire harness and connectors while the system monitors voltages for change. Nothing found though. A few more test drives and still nothing. I may have to let this one go with no fix at this time. But didn't we have fun?!

Thanks for reading!

kenny@ggauto.repair