Battery cable removel

Tiny
JCLEMO
  • MEMBER
  • 2000 KIA SPORTAGE
  • 2.0L
  • 4 CYL
  • 2WD
  • 149 MILES
I want to elaborate a little further about my prior question about the Kia being hard to start once the battery cables been removed. So to explain a little more after removing the cables from the battery when replace the cables back on the battery the vehicle is hard to start like it's not getting fuel.
But after about 45 seconds it will start. I let it run for a few minutes and shut it off. Wait one minute and starts right up. Now if I remove the positive cable and wait for about an hour and then replace the cable back on the battery it's hard to start again. I've double checked all timing marks and everything is properly set all marks are in place. Could this have anything to do with the ECM resetting itself?
Thank you. JC
Wednesday, July 3rd, 2019 AT 12:04 PM

5 Replies

Tiny
CARADIODOC
  • MECHANIC
  • 33,752 POSTS
You're wiping out all the learned data in the Engine Computer. A lot of computers need to see a very specific set of conditions to relearn when they need to be in control of idle speed. Until that is relearned, idle speed can be too low for the engine to start, or it may tend to stall at stop signs.

To verify if this is what's happening, hold the accelerator pedal down about 1/4" when cranking the engine. If it starts right up that way, you'll need to visit the dealer to learn what it takes to do that relearn. The only one I have memorized is with Chrysler products, you need to drive at highway speed with the engine warmed up, then coast for at least seven seconds without touching the pedals. Once that is done, idle speed will be back to normal.

Disconnecting the battery cables can lead to real lot of expensive misery on some newer models. We hear horror stories all the time about cars that won't come out of "park", if the engine starts, it won't accelerate above idle, and a tow to the dealer is needed to unlock multiple computers. There's nothing to "reset" by taking a cable off. That thinking goes back to the early '90s when GM had a lot of Engine Computer trouble, and disconnecting a battery cable could get them working again.

Your computer starts off with a set of standard factory-programmed fuel metering values, then, as you drive, those values are constantly updated and fine-tuned for the best performance and best fuel mileage. That learning is all lost when the battery is disconnected, then you're back to the less-efficient starting values. When professionals replace a battery, they use a "memory saver" device that keeps the memory circuits powered up so that learning isn't lost. That was never an issue many years ago, but today those devices prevent a lot of frustration.
Was this
answer
helpful?
Yes
No
+1
Wednesday, July 3rd, 2019 AT 4:26 PM
Tiny
JCLEMO
  • MEMBER
  • 161 POSTS
Wow, that makes all the sense in the world. That explains why I'm smelling raw gas at first start. I was taking the positive cable off the battery for safety.
While replacing belt water pump and tensioner and idler bearing. I'm not a professional mechanic by any means
But I finally realized the only time it was hard to start was after that cable had been removed. I go out in the morning it starts right up. But I was going crazy trying to adjust the timing and the timing marks were all on the marks. So when I test drive it Morrow if it does anything weird
I need to see the Kia dealer from what your saying? If it. Drives okay I'm lucky.
Well mine is an old one it's a 2000 Sportage. 2.0 and runs excellent. I figured the belt needed replacing. But someone had replaced some time ago I could tell. But the tensioner needed replacing so I did it all. Thanks for the valuable information. This site is awesome l love it. I need to make a donation but I'm not sure how to go about it. But I will inquire. Thanks again, now I know.
I was right it's the electronics going nuts. I won't remove the cables again.
JC
Was this
answer
helpful?
Yes
No
Wednesday, July 3rd, 2019 AT 5:40 PM
Tiny
CARADIODOC
  • MECHANIC
  • 33,752 POSTS
The problem isn't in removing the battery cables. The problem is in knowing what to do afterward. For a lot of models, the computer relearns idle speed information by some other strategy, and the driver doesn't even notice that's taking place. As I mentioned, I only have Chrysler's procedure memorized because I have a lot of them. I don't know why they don't publish it in the service manuals. I first learned this at one of their classes I attended. The computer needs to know your foot is off the accelerator pedal, then it takes a reading from the throttle position sensor and puts that in memory. From then on, any time it sees that same voltage, it knows it has to be in control of idle speed. Otherwise it thinks you're controlling engine speed yourself, and it doesn't interfere.

To know your foot is off the accelerator pedal, the computer needs to see high manifold vacuum for seven seconds, with a perfectly stable TPS reading. If the reading bounces around, it knows your foot is on the pedal, so it ignores the relearn. You can get high manifold vacuum in the garage by snapping the throttle open, then releasing it quickly, but that will just last one or two seconds, not the required seven seconds. That's why you have to be at highway speed so you can coast and get that high vacuum for seven seconds. We were even told a lot of battery shops and dealership mechanics are instructed to perform a short test-drive when necessary to do this relearn, for customer satisfaction. I used to just inform people of the need to do that when they leave the shop, but every now and then you get someone who doesn't understand how to drive with one foot on the gas and one on the brake until this is done. It's just easier to do it myself. I kill the battery in my minivan about once a month from running a power inverter. I carry a jump-pack, but each time I have low idle speed until I hit a long off-ramp on my way home. I can see engine speed pick up after that seven seconds as I get to the end of the ramp. That's when the relearn has been completed and the computer is bringing idle speed up to where it should be.

The other thing you might like to be aware of is fuel trim numbers. Every Engine Computer starts off with a factory-programmed set of "look-up" tables. For every possible combination of engine speed, engine load, coolant temperature, intake air temperature, throttle position, throttle direction of change and rate of change, and multiple operating conditions, there is a magic number in those look-up tables that tells the computer exactly how many milliseconds to pulse the injectors open to achieve the proper fuel / air mixture for best performance and fuel mileage, and lowest emissions. The actual mixture is affected by the temperature of the gas in the tank, since that affects how quickly it evaporates and is collected in the charcoal canister, and it's affected by how much unburned gas has snuck past the piston rings and diluted the engine oil. Fumes from both places are drawn off to be burned. That makes the mixture too rich. The computer accounts for that when it sees the switching between "rich" and "lean" about twice per second from the front oxygen sensors. Based on those readings, it fine-tunes the injectors' on-time to readjust the mixture.

That's the long version of how the computer keeps tabs on the fuel metering data. It starts out with those factory default values, then, while you're driving, when it sees the need to inject slightly more or less gas, it modifies the numbers in the look-up tables. Those are called the "short-term fuel-trim", (STFT) numbers. The higher positive they are, the more fuel it's adding right now over the factory values.

Eventually the computer will see it's always making the same corrections for the same set of conditions. At that point, it moves the short-term numbers into the "long-term fuel-trim" (LTFT) numbers. Those are the starting numbers it uses now instead of what was programmed in on the assembly line. There should now be less need to change the short-term numbers while you're driving. All of these numbers can be viewed on a scanner, either as a set of a real lot of data on a table, or simply as the current STFT and LTFT numbers. When you see the short-term numbers are high, the computer is making adjustments right now. When you see the LTFT numbers are high, the computer has made big changes over what was expected to be "normal" by the engineers. If you do a lot of short-trip driving where the engine never gets fully warmed up, there is going to be more unburned gas in the oil. That results in a richer-than-expected mixture when that is pulled out through the PCV system to be burned. That won't happen effectively until you make a nice long trip on the highway with a hot engine and hot oil. You can expect to see high negative fuel trim numbers because the computer is reducing the amount of gas it's injecting to offset what's coming in from the oil or the charcoal canister.

Fuel pressure is also a big player in the fuel / air mixture. Some cars have fuel pressure sensors, but on most models the computer simply knows what the pressure is supposed to be. Spring-loaded mechanical fuel pressure regulators are usually very reliable and accurate, but a weak spring can let gas go through it and back to the tank too easily. That results in low fuel pressure. If the computer doesn't monitor pressure, it will see the resulting lean condition from the oxygen sensors, then make appropriate adjustments in how long it holds the injectors open during each pulse.

Fuel pressure and manifold vacuum are the two forces acting on a molecule of gas as it's ready to leave the injector. Vacuum goes real high during coasting, and that would result in a rich coasting condition and excessive emissions because that vacuum is tugging harder on the gas molecules. To offset that high vacuum, fuel pressure is lowered so the difference between those two forces remains constant. Fuel pressure is most commonly lowered by connecting a vacuum hose to the regulator. Vacuum helps to pull the relief valve open so the fuel can go back to the tank easier, leaving the pressure lower. With higher vacuum and lower pressure during coasting, the net difference is the same, so there's no rich condition.

In this story, a leak in the vacuum hose will look to the regulator like the engine is under hard acceleration. With lower vacuum, the pressure relief valve is harder to push open, so fuel pressure rises. In reality, the engine is just cruising with minimal load, but with the higher fuel pressure, more fuel will flow from the injectors and you'll have an excessively-rich condition. The oxygen sensors can't measure gas, but they will detect the lack of oxygen during the "lean" pulses of exhaust gas. Once again, the computer will adjust for that, but there's a limit to how much it can adjust the short and long-term numbers.

The range of adjustment is designed to handle normal variables and the typical range of those variables. It's when something major happens to affect the fuel / air mixture beyond what the computer can adjust for that performance or emissions problems occur. Emissions problems are detected by the computer and usually include a flashing Check Engine light meaning too much unburned gas is going into the exhaust system where it will burn in the catalytic converter and overheat it. You're supposed to stop the engine as soon as safely possible to avoid expensive damage. Excessively lean conditions also contribute to increased emissions due to incomplete combustion.

The point of this sad story is when you disconnect the battery, the fuel trim numbers are lost. With a properly-running engine, those tables start being rebuilt as soon as you start the engine, and you will rarely notice anything unusual in engine performance.

As a point of interest, Chrysler invented the computer-controlled automatic transmission for '89 models, and they do the same thing with constantly updating shift schedules and clutch overlap to offset the normal wear of the clutch plates. That results in nice crisp like-new shifts throughout the life of the transmission That learning is also lost when the battery is disconnected. It can take up to two miles or a dozen up-shift cycles for that data to be relearned. Until then, the transmission could shift harsh like a dragster, or it could be soft and mushy. By 2000, most other manufacturers had copied the design and now have their own versions. If you notice unusual shifting for the first few miles, don't panic. That will clear up as the data is relearned and updated.
Was this
answer
helpful?
Yes
No
+1
Friday, July 5th, 2019 AT 12:25 AM
Tiny
JCLEMO
  • MEMBER
  • 161 POSTS
Thank you for the update. That's serious knowledge and good to know your average weekend mechanic would
Never suspect anything closely related to this problem I know I would have never known to do this driving procedure but I'll do it today. I have a scanner an actron but for some reason it doesn't recognize the 2000 kia Sportage. The engine light is on and has been since I bought it six months ago. I've never had any problems untill
I decided to change the belt. The light is steady and does not flash.
Thanks again. JC
Was this
answer
helpful?
Yes
No
Friday, July 5th, 2019 AT 6:29 AM
Tiny
CARADIODOC
  • MECHANIC
  • 33,752 POSTS
Some other manufacturers have different strategies for relearning various parameters, so if normal driving over a day or two doesn't work, visit the dealer and ask someone there. You might have to ask to speak with their engine performance specialist as this isn't always well-known, even among dealership mechanics. That's because it doesn't cause enough problems to generate customer complaints.

I don't like to hear the Check Engine light is staying on, even when you know what the fault codes are. The problem is if another different problem develops, you'll never know it because the light is already on. Some of those problems, when ignored unintentionally, can be very minor, but turn into expensive repairs.

The other problem is there is always a long list of conditions that must be met for the computer to set a diagnostic fault code. One of those conditions is usually that certain other codes can't already be set. One example is the computer knows when the engine has been off for at least six hours, the intake air temperature sensor and the coolant temperature sensor had better be reading the same temperature. If they aren't, the computer has ways of figuring out which one is wrong. If you were to unplug the intake air temperature sensor while the ignition switch is on, the signal voltage will get "pulled up" to 5.0 volts, which is a defective condition, to force the defect to be detected. At that point, anything the computer uses that sensor reading for comparison, it knows can't be trusted, so it suspends those other tests. If the coolant temperature sensor simply goes out-of-range, the computer won't know it, and that won't set the appropriate fault code but it could cause engine running problems. With no fault code to direct you to the circuit that needs to be diagnosed, you'll never figure out why the engine isn't running right. At that point the only solution is to diagnose and repair the first, known problem, erase that fault code, then the other tests will resume on the test-drive or when the customer picks up the vehicle. That's when the new problem will be detected, a new fault code will be set, and the Check Engine light comes right back on again.

Be aware too that of the roughly 2,000 fault codes that can be set, only about half of them refer to things that could adversely affect emissions. Those are the codes that turn on the Check Engine light. A lot of those defects do not cause noticeable engine performance problems. That's why we often hear the light has been on but the engine runs fine. The politicians don't care how well the engine runs. They're only concerned with emissions.
Was this
answer
helpful?
Yes
No
+1
Saturday, July 6th, 2019 AT 12:38 AM

Please login or register to post a reply.

Sponsored links