Home      Trouble Shooting

Engine Starts but Does Not Run Properly


It is often easier to determine why a car refuses to start than it is to figure out why the car is not running properly.  In order to insure one's car is running properly, one has to have a scan tool and a fuel pressure gauge at a minimum.  I am often amazed when someone asks me to look at a log and help them tune their car for better performance and I see things which are not right, yet, they are trying to race the car, or add mods to improve the performance.  If the car is not running properly to begin with, adding mods is generally slightly less productive than pouring money down a hole.  Therefore, one needs to understand the Basics and Scan tool parameters in order to know if the car is solid before wasting money on performance parts that will not do any good because the foundation is cracked.  If unsure about something, ask questions and learn.

A tool like PowerLogger is invaluable of speeding up the troubleshooting process and it makes it far easier for others to help you as all the data is there in one spot to be looked at.  A boost gauge with a vacuum scale on it as well is very useful as well.  If you have PowerLogger, you can monitor vacuum/boost, fuel pressure (if you have a pressure transducer installed), and other things not normally seen on a scan tool.

Make sure you are using a scan tool designed for Buicks and not one of the OTCs that uses 0.500 for stochiometric A/F rather than 0.441 which our tools use.  IF you are, then you have to correct all the readings to make them relevant to the numbers we throw around.   441/500 x the OTC reading.

In addition to the above, a wide band oxygen sensor may be useful....I am not quite as adamant about its use for the average street car as I once was.  In fact, sometimes I think they can be distracting when one begins to put all ones faith in one basket and finds the data confusing.

Now, troubleshooting problems can be very tedious and I am not sure exactly how to format the procedure in a logical way as the subject is a very broad one.  I am going to try to group things into a few general categories and see how that works out.  If I could classify 75% of problems into a few categories, I might be doing pretty good.

Always check for trouble codes when you start looking for a potential problem. 

Look for hints in the various parameters for something that is reading differently than it usually does...Is the vacuum reading the same as always, are the blm's the same, O2s, etc.  Knowing what your scan tool normally reports is a good help in getting started.

If you understand the basics of how the ecm supplies fuel under varying conditions, then it will make it easier to apply some logic when looking for a problem.

Now here are some basic things to check when trying to sort things out.  I have seen problems resulting from all of the below potential problems.

Did this happen over a period of time and gradually get worse, or did it suddenly occur and the symptoms have been consistent from that point? Had you worked on the car before it started? Did you recently buy the car or have you had it a long time?  Try to recall if you worked on anything shortly before the problem arose.

If you can see the problem at idle, this can speed up troubleshooting.

First I would look at the scan tool  and write down the readings for O2, maf, blm, clt temp, and voltage when the car is idling properly plus I would write down the vacuum reading and the fuel pressure.  Power Logger makes this very easy.  The neat thing about PL is that if the problem suddenly begins to occur while you are monitoring, you can hit record and it will go backward in time by grabbing the data in buffer and then go forward.  The means you don't have to continually record but can start when the problem shows its face.

Then when the car is idling badly when the problem occurs, write down all the same information again and see if anything stands out as being different.

Next I would start with some basic things such as verifying that the ground cable to the turbo support bracket is firmly bolted down and that the small ground cable off the negative battery post is attached to the inner fender sheet metal as it should be. Check to see if the coil/module is bolted down firmly against the mounting bracket and the bracket is bolted firmly to the intake/plenum.

Be sure to check the grounds on the back of the passenger head to make sure none are broken or loose.  These have to be well connected or you will have problems.  Many of us relocate them to the firewall with a kit, or pull them up and put them on an intake stud.

Make sure the positive battery cable from the battery to the starter is firmly attached to the starter bolt and that neither the cable or the fusible link bundle that attaches to the starter post is touching, or can touch the headers or anything else that could short out the positive feed.

Does the Scan Master show any stored trouble codes and does the SES light come on, or flicker quickly when this problem happens?

Put your volt meter on AC and read the voltage across the battery when the car is idling. It should be less than 0.5 v AC.

I would pull all the spark plugs and see if they all look like they are firing properly and none of them appear to be wet or seem to have not been firing. I would replace them with a new set of properly gapped plugs no matter how they looked.

With the car running, unplug the cam sensor and see if it keeps running right,  still incurs the problem, or runs considerably different. Don't forget to re-plug it after you turn it off or it will not start again.

Measure the resistance between the spark plug terminals on each coil pack (engine not running, of course) and see if the resistance falls between 11,000-13,000 ohms on each pack.

Check the base of the egr valve and make sure the clamp bolt is tight and there are zero leaks around the base and gasket. When the car is cold, pull the cover off the egr so you can get to the egr. Put your fingers under the egr and lift the diaphragm up and see if pops back down closed like it should.

Make sure the pcv valve is moving freely in its housing and is not "sticky".

Check the hose connections between the maf and the turbo and make sure all are properly installed and tight so there can be no air leaks. If by any chance you are using the original style hose from maf to to turbo, be absolutely sure there are no holes in it.

Check the hose connections between the intercooler and the throttle body to make sure they are all properly in place and properly sealed.

Check to be sure the crank sensor is not loose and is properly centered over all three of the blades on the balancer.

Your scan tool (not your nose) says the car is running rich-O2s are high, blms are low

Have you checked the o2 sensor cross counts at cruise and light throttle to make sure the sensor is properly active?   A bad sensor will screw up everything.  It is cheap, buy a Denso and put a new one in, if you have any doubt.  Remember that leaded gas, a too rich mixture, or old age can, and will, throw one out of calibration.

Is it rich from idle all the way to the shift point, or is it only rich at idle, or at high rpm only?

If it is rich primarily at idle, are the maf readings correct at idle?  Note that the oem factory mafs are dying quickly these days and the rebuilt ones that are offered for sale are often not calibrated correctly.  Typical failures are high readings at idle which makes the car run rich, or failure to reach close to 255 at wide open throttle which causes the car to run lean. 

 Are you taking into account that most modern chips run fairly rich in open loop at idle and this may create a situation where the idle is richer than normal in the name of a good stable idle.  If seemingly excessive, you can probably offset this a bit with fuel pressure, chip programming, and/or Translator adjustments if so equipped.

If it is rich all the way thru the operating range, have you verified the fuel pressure with a real pressure gauge and not one of those rail mounted Cracker Jack pos's that are lucky to be +/- 10% when new and get worse day by day as the vibration shakes them apart? 

 If you have a Translator, have you checked the dip switches for the correct setting for the chip you are using and are the dials set properly for the particular maf you are using with the Translator? 

 Have you used the programmability of the chip to adjust the fueling? 

Have you checked to see if the coolant temps reported by your scan tool are correct?  If they are very low in comparison to what they should be, have you checked the coolant temperature sensor? 

 Do you have a restricted exhaust-clogged catalytic converter or collapsed muffler?  Is the air filter clean? 

 Is the car in Limp Home mode (should have a Service Engine Light)? 

If a new car (to you), are you sure the injectors match the chip in the ecm?  Could it be that you have very old injectors that may need cleaning (which brings up the question-is it worth cleaning old injectors are is it time for upgrade/replacement)?

Is the chip programmed for 25 psi of boost and you are only running 18 psi to be safe?  If you do not reduce the fuel in the chip, the car will be very rich at top end under boost.

Scan tool says the car appears to be running lean-O2s are low, blms are high

Again...Make sure you are using a scan tool designed for Buicks and not one of the OTCs that uses 0.500 for stoch rather than 0.441 which our tools use.  IF you are, then you have to correct all the readings to make them relevant to the numbers we throw around.   441/500 x the OTC reading.

Check the O2 sensor and make sure it is actively cross counting as it should when at cruise.  A bad sensor will screw up everything.  It is cheap, buy a Denso and put a new one in, if you have any doubt.  Remember that leaded gas, a too rich mixture, or old age can, and will, throw one out of calibration.

Make sure there are no leaks between the Maf sensor and the throttle body. Air that is pulled in after the maf sensor will not be metered by the maf and will not be accounted for by the ECM other than the blm corrections being made when the O2 sensor sees that the exhaust is lean (low O2's) and the blm's go up showing the ecm is trying to correct the a/f.  These leaks can be in the connections to the maf, turbo, intercooler, or throttle body....or in the hose between the maf and turbo if the factory style hose is still being used-examine it closely.

Also check for leaks at the throttle body shaft seals, the joint between the throttle body and plenum, the plenum to the intake, or in any of the multitude of rubber vacuum hoses and/or steel tubes (they do crack at times).

At times, particularly if the intake has been removed, there may be a leak around the intake gasket...this may be under the intake and can not be checked from the outside...the intake has to come off.  Sometimes this happens because the heads and/or block was decked during a rebuild or head refurbishment.  This can prevent the intake from setting down far enuf to get good gasket crush on the intake gasket and a leak will show up after a few applications of "boost".

A crack in the exhaust manifolds before the turbo can also cause an apparent lean condition.  Air enters thru the crack and passes the O2 sensor which then reports a lean condition.  The ECM adds fuel to richen up the mixture...in this case the car can actually end up being rich because the O2 was tricked into believing the A/F was lean coming out of the exhaust port when it really is really not lean.

Again, try to use some logic in your trouble shooting.  Is the car lean all the way from idle through wide open throttle at the top end of third gear?  If it is, check your base fuel pressure.  Don't take your fuel pressure gauge reading for the absolute truth.  Add a couple of pounds of fuel pressure and see if that helps anything.  Then check the fuel pressure increase with boost and see if it maintains the required one pound of increase in fuel pressure with each one pound of increase in boost.  This requires a fuel pressure gauge that can be seen thru the windshield when driving so that pos cheap gauge mounted on the fuel rail that looks cool but is not good for anything will not work.  It takes either a quality gauge mounted on a hose that will reach the windshield, or, an accurate fuel pressure transducer that feeds Power Logger.

If the pressure does not rise properly with boost, then you have a fuel delivery problem and will probably have a bigger problem when the engine detonates enough to destroy something expensive.  Now, we have to find out what is causing the fuel delivery problem.  Is the fuel pump "hot wired"?  Is the voltage to the pump staying at 13.5 v, or higher, all the way through the run?  Is the fuel filter plugged up with debris?  Is the hose that couples the pump to the sending unit clamped properly and not leaking under pressure?  Are any of the rubber hoses in the fuel lines  kinked, or have pin holes in them?  Is the fuel pressure regulator working properly?  Is the fuel pressure gauge reading properly?  Is the fuel pump old and tired?

If the car is only lean at top end, it could be either the chip programming, or a lack of fuel delivery.  Start with the basics and make sure the pump is supplying fuel and then check to see if adding fuel in the chip programming will bring things in line.

Remember that if the chip has default fuel parameters for 18 psi of boost and you are running 25 psi, then the car is going to be lean because you are supplying a lot more air than the chip was anticipating.  You will have to add fuel in the chip.

Misses, or Backfires under Boost

If the car runs okay in light throttle conditions but starts to miss when the boost comes on, this is often a bad coil pack, and, sometimes, a bad module.  The best way to test the coil and module is by using a Caspers coil and module tester as this stresses both and the spark will break down as the rpm on the tester is increased.  The Casper's tester does not work on aftermarket modules, however.  The car will run when started, but, if the tester is used, there will be no spark at all...then you know you have an aftermarket module.    If you don't have the tester, then you can use a meter and check the resistance across each coil pack (between the two plug wire terminals on a given coil pack) when it is hot from running.  The resistance should fall between 11-13,000 ohms on a good Type I coil pack.  Type I is the original coil used on our cars...one piece unit with three individual coils molded into it.  This is only a test of the coil packs and not the module and is probably about 85% reliable.  Sometimes a coil will pass this resistance test but still fail under boost.  The best test outside the Casper's tester is to swap a known good coil and module onto the car.  I have never seen one that failed the resistance test to be anything but bad. 

If the scan tool is reporting timing retard, detonation may be causing the problem, but, typically, this will not be as severe as a bad coil or module and the symptoms will not be as obvious.

At times, a loose balancer, or misadjusted crank sensor can cause an obvious miss at higher rpm.  It is always a good idea to check the crank sensor for proper clearance on all three blades of the damper and to check the damper for wobble or movement when the engine is revved up. 

A bad cam sensor can also cause a problem.  Sometimes this becomes obvious after the car has been driven awhile.  If the cam sensor has been out of the car and reinstalled, be sure it was properly installed and not on the edge of the window, or 180 degrees out.  Sometimes, unplugging the cam sensor after the car is running will cause the car to run better.  This is normally a sign of some problem with the sensor.

See the note in the next section about the cam sensor-

Misses, or Runs Poorly at Idle as Well as at Higher Speeds

The first thing to determine is if the miss is specific to one or more cylinders.  I believe the easiest way to determine this is to crank the engine up and let it idle.

After it has warmed up, begin disconnecting one injector connector at a time.  Note if the idle gets worse, as it should, or stays the same.  If there is no significant change, it would appear that the cylinder that particular injector feeds is not firing properly.  Write it down...the cylinder number.

Reconnect the connector and and move to the next cylinder and disconnect that injector and note the effect.  Continue until all six injectors have been disconnected.  Now, if there is a single cylinder that had no significant change when unplugged, the problem is pretty well isolated.

If there were more than one cylinder showing a problem, then the first step is to determine if it (they) shares a common coil pack which would indicate that it may be a problem with that particular coil, or module.  Again, the Caspers Coil Tester makes it easy to show a problem with the coil and/or module.

If the problems do not correlate with a given coil pack (see the section on coils and module for common cylinders for a given coil), then one must determine if the problem is a bad plug wire, bad plug (it happens),  a bad injector, a bad injector harness, or, an ecm problem.

To check a plug, or a wire, I normally remove the plug, plug it back into the wire, and lay it on the valve cover where I can see the electrode end.  Then I crank the engine for a second while watching for a blue spark from the center electrode to the ground.  If it sparks continually as it should, I call the plug and wire good, and move on to the injector.  If it does not, I turn the engine off and stick a different plug into the boot and try it again.  If that plugs sparks, I consider the plug bad.  If it does not, then I look at the wire.  Note that I am assuming that the opposing plug on this particular coil is firing properly because that cylinder dropped off appropriately when I pulled the injector connector.  If you have a spark tester, you can avoid removing a plug at this point.  Never crank the engine without a plug connected and grounded as this can blow the coil or module according to G.M.

Okay, if the plug was firing properly, my assumption is that injector, harness, or ecm is bad-after I check the compression on that cylinder.  If the compression is good and the spark is good, the cylinder has to fire if the injector sprayed gas in to the cylinder.

Now, at this point we have to find out if the injector harness is delivering power to the injector and is being grounded by the ecm.  The best way to do this is to connect a Noid lite to the connector of the injector in question.  Crank the engine and see if the lite is blinking rapidly.  If it is, then we are back to the injector.  Noid lites are cheap and useful!  (will tell you that the cam sensor is sending a pulse as well)

If you want to verify that, you can swap a couple of injectors and see if the problem follows the injector, or stays on that cylinder.

If the problem seemed to be in the injector harness, get your meter out and see if there is power coming in on one lead (check the diagram for the wire color to that cylinder).  If there is no power, then you can figure there is a break in the wire on that lead somewhere.  If there is power, it gets to be more complicated.  The problem could be a broken wire going to the ecm, or it could be a bad driver in the ecm.  The easiest way is to swap ecms and see if the problem goes away, or is still there.

There have been some problems at part throttle and low boost reported that seemed to mimic compressor surge, but, were not.

The problem was traced back to the cam sensor-  Over the years, the gear on the cam sensor wears and this allows more vertical movement of the shaft.  This can cause two potential problems.  One, the reluctor wheel can rise up and and not get a clear signal from the magnet.  Second, it can rotate the position of the reluctor window in relation to the magnet resulting in a cam sensor that is on the edge of the window and confusing the fuel timing.

There are two ways to deal with the problem and I advocate them both.  First, as suggested in other sections, after setting the cam sensor, rotate it another 1/8" CCW.  Second, get some Mr. Gasket distributor shims #2820 and shim the vertical movement in the shaft so that it is closer to 0.040".  We used to suggest much less than that, but most of the new sensors I have checked have been between 0.40-0.050" so that should be good enuf.