Preamble

The full text of the original problem description has been retained (below, on this page), for anyone unfamiliar with this  - a rather nasty, hard-to-diagnose problem I posted questions regarding to HeD and other places last fall, in an attempt to resolve.  Basically, it was a very intermittent, but quite crippling (riding-wise) stuttering problem with my EFI Evo engine.  Dang, it's been frustrating!

But now - partly in thanks to all who offered helpful advice and partly to provide some information that might be helpful to others if they encounter the same situation - I can finally report back what the problem stemmed from and the best way to attack intermittent problems of this sort.

Having an EFI bike with a problem is sometimes a matter of working in the dark, to some extent, unfortunately - one reason for perhaps sticking with carburated.  The backyard mechanic does not have the resources to trouble-shoot certain problems.  The best you can do is eliminate the things that can be verified without the assistance of expensive and restricted (to HD dealers only) equipment.  Shoot, HD won't even let you buy their breakout box tool! 

But, by eliminating all the possible causes you can on your own, you can at least narrow the range of things you eventually pay the dealer's mechanic for (should your own efforts not succeed).

As this case study will demonstrate, the thing you definitely do not want to do is to start replacing parts willy-nilly - in the hopes that you stumble on the cause of your problem.  You can very easily (and quickly) throw a lot of good money down the drain that way.  If an EFI engine is malfunctioning, there can be many, many possible causes - not all of which seem very obvious.  Well, maybe I should modify that.  If indicators make a part seem very suspect and that part is inexpensive, maybe give it a shot, but don't do that with costly items.

The Solution

The solution, in my case, hinged on the use of a diagnostic tool that the HD dealer had (Heritage Harley-Davidson of Edmonton).  I'm not sure what it's proper name is, but basically it is a unit that you can attach to the bike that records data from virtually all aspects of the bike's electronic appendages.  I am also told that it is an expensive little gadget (in the neighbourhood of $14K Cdn).  It is equipped with a control stick that has a button on it (which you tape onto one of the handlebars).  Using it, the mechanic can ride the bike until the problem occurs and, at that instant, press the button to begin recording data. 

When the button is pressed, the unit records all sensor information for a period of 4 minutes prior to the button-press (the unit has a buffer) and for 4 minutes after the button was pressed.  Afterwards, that information can be displayed graphically on a computer for careful diagnosis.

Here's the important part of the graph produced while the mechanic rode my bike:

Notice that big 'step' in the Engine Temperature graph?  That definitely shouldn't happen, right?  Certainly it's expected that, as the engine warms up, that the graph will slowly trend upwards (at a very slight slope), but engine temperature doesn't change in discontinuous steps like that.  There's a sudden, 115 degree Farenheit shift showing here.  Something's gotta be wrong with that blasted engine temperature sensor! 

One look at that graph and I was buying myself a new sensor (no need to pay the dealer to put it in).  Removed the tank.  Replace the sensor.  Put 'er back together and, in the space of an hour, my sentence of frustration was commuted.  The bike ran absolutely perfectly after that!

Lessons Learned From This

That engine temperature sensor was not a cheap part.  It cost me $180 Cdn.  That's not the kind of expenditure you want to make if you are not sure.  Quite frankly, I was very strongly leaning towards fuel injectors being the cause of the problem (largely due to when the problem began, as explained in the section below).  And, without the evidence the graph provided, that is surely what I would have began by replacing first.  And I'd be poorer by $185 per injector with the problem still not resolved.  That shooting-in-the-dark method is a bad way to go.

And, not initially aware there was a sensor-logging unit available for HD bikes, I was extremely reluctant to bring the bike to the dealer.  I had visions of them spending days and days; trying to get to the bottom of this; especially as it was so intermittent.  Who knows?  They could spend a thousand bucks of shop time on it, hand the bike back to me and say, "We couldn't find anything wrong with it."  And then, "Oh yeah, before we forget... here's your bill."  Stuck between a rock and a hard place, man.

I did sort of suspect the engine temperature sensor, but had no solid proof that it was bad.  Regarding it, I did the best I could on my own:  I removed the fuel tank and measured the resistence across the pins of the sensor.  They didn't shown anything unusual, so I was very reluctant to throw $180 buck at it, on a gamble.

What's more, without the sensor graph showing the discontinuity, it would have been very difficult to pick up on this problem any other way.  Both the pre-step 117 degree reading and the post-step 231 degree readings are within the acceptable range of values for the sensor, so there's no bright red flag there to catch one's attention.  Sure, there are ohm readings for various engine temperatures that could be compared (if you have an accurate idea of what the real engine temp should be, compared to the readings), but even that could be missed quite easily. 

What's more bedevilling about a problem like this is the fact that the problem was so darned intermittent.  It's just a matter of luck, getting it to happen while the bike is inside the shop.  Unless you can do it, you'll still be left without a clue.

So, in conclusion, don't throw parts at problems like this.  You really have to get that data-logging unit strapped to your bike to figure out problems like this in the most efficient way.  If your dealer doesn't have such a device, better find one that does.

Background (Original Problem Description)

The bike is a 1998 Sequential Port Fuel-injected Road Glide with about 70K miles on it. Basically stock, aside from having a Stage I kit. The rear base gasket is beginning to sweat, but no major work done on the engine yet. The problem described below has been occurring for a month now; initially not getting better or worse, but recently getting more frequent, it seems.

I should note that the "Intake Air Temperature Sensor" has been going bad on the bike since about March; causing the Electronic Control Module to throw a trouble code indicating that. It didn't seem to be interfering with the running of the bike, so I've been ignoring it (maybe fix it in the off-season). Other than that, no trouble codes are being thrown (I've checked for them many times since the problem described below began). But, perhaps due to the way the storing of repetitive codes gets spaced, that doesn't mean anything. The codes for the Air Intake Sensor are causing the engine light to be lit about half the time, whenever I ride.

I should note that there appears to be no correlation between the engine trouble light - associated with the Intake Air Temp sensor - coming on and the onset of stuttering.

Brief Description of the Problem

It started a few minutes after refueling at a small-town, independent gas station. The engine began stuttering and almost dying every time I tried to apply throttle. Running at a steady RPM, the engine seems to run quite normally; it's just whenever I applied throttle (to accelerate or even when just shifting) that the engine would sputter.

And I mean stutter badly; throwing-me-forward-in-the-seat badly - as if I were letting go of the throttle for a second or two. And then, after belching out some black smoke, it would slowly build up revs to obey the throttle command. Almost un-rideable and quite dicey to take corners with now, as the power-delivery sometimes comes so late.

Initially, the problem never seemed to occur right after starting the bike but, now, it sometimes does. It may be getting worse and more frequent, in fact. It seems to require a few minutes of running before it starts acting up, but sometimes almost right away lately. I should note that, here in Edmonton, the outside temperatures have been getting significantly colder lately. Also, when the stuttering does happen, it keeps on happening for a minute or two, then it runs fine in stretches.

It acts just like a motor being started after being flooded; first dying and then attempting to burn out the excess fuel - black smoke coming out of the pipe, as the engine finally manages to rise in RPM, in response to the throttle's command. The fouled front spark plugs seem to indicate too much fuel (the deposits on the plugs are dry/fluffy, not wet/shiny); the front plug being quite a bit more fouled than the rear.

Another odd thing about it is that the problem is intermittent. For almost every ride, it acts up about 40-50% of the time - maybe even more, lately. The rest of the time the engine runs normally and has good power.

Although I somewhat suspect a bad coil or bad spark plug wire, I've been having difficulty getting to the wire, where it plugs into the coil, to remove it and test it. Any suggestions for an easy way?

But it might not be ignition failure that's causing the fouled plug condition. It seems that the rear plug is getting a bit fouled as well. It's possible it might be the injectors, but I've run a test (below) that seems to rule that out.

Could also be a leaky induction module (the thing that replaces a carburator on fuelies) seal. When I sprayed a garden hose around the induction module while the engine was running (engine warmed up), the idle speed changed dramatically (even momentarily stuttering) in response to the first blast of water. Subsequent sprays only changed the idle a tiny bit (say, 100 rpm), if anything.

Photos of Problem's Symptoms

Front Spark Plug
(photo somewhat washed-out - in reality, it is completely black in colour). Seems to be running quite rich on the front cylinder (although only at times when stuttering, perhaps).

Rear Spark Plug
(this one looks like it's burning a bit rich as well, but not as bad as the front plug)

Just FYI, this is how I have the fuel pressure gauge rigged up for mobile use (had to buy a special fitting, used an old 0-150 psi oil gauge)

This is what the fuel filter looks like (in case you've never seen it). I replaced it, as the old one with 70K miles couldn't be blown through at all (the new one can be blown through easily). Didn't resolve the stuttering.

Things Tried So Far, Leading To No Improvement

• Put in different set of (old) spark plugs

• No Code 12 indicating barometric sensor failure, so unplugged sensor, generating Code 12, to rule our wiring harness being faulty.

• Ran injector-cleaner through 4 tanks of gas and ran menthyl hydrate through as well

• Replaced gas completely

• Replaced fuel filter, as shown above

• Visually inspected regulator, noting no obvious problem

• Checked fuel pressure, which stays constant at near 50 psi, whether idling or revving engine

• Serviced electrical connections (cleaning with elect. contact cleaner and applying dielectric grease), including spark plugs, fuel pump connectors, relays and barrel connector for crankshaft-position sensor

• Replaced with new fuses for both ECM and fuel pump

• Checked battery condition (12.96 volts across terminals)

• Checked resistance in fuel pump (and other) relays - 0.08 Ohms for all

• Removed ECM and mounted it in front of a hair dryer for 20 minutes to dry - in case any moisture existed inside

• Conducted test to determine that injectors are not leaking

• Sprayed water around the induction module (fuel-injection equivalent of a carburator) while engine was running (warmed-up condition), to see if idle speed changes. Initial spray caused the engine to stumble quite badly, briefly and, additional spraying seemed to affect engine idle somewhat (may lowering it by 100 RPM). Hard to say, but that might indicate a leak in the induction module seals. Either that, or the coolness of the water caused the change?

Preamble

The full text of the original problem description has been retained (below, on this page), for anyone unfamiliar with this  - a rather nasty, hard-to-diagnose problem I posted questions regarding to HeD and other places last fall, in an attempt to resolve.  Basically, it was a very intermittent, but quite crippling (riding-wise) stuttering problem with my EFI Evo engine.  Dang, it's been frustrating!

But now - partly in thanks to all who offered helpful advice and partly to provide some information that might be helpful to others if they encounter the same situation - I can finally report back what the problem stemmed from and the best way to attack intermittent problems of this sort.

Having an EFI bike with a problem is sometimes a matter of working in the dark, to some extent, unfortunately - one reason for perhaps sticking with carburated.  The backyard mechanic does not have the resources to trouble-shoot certain problems.  The best you can do is eliminate the things that can be verified without the assistance of expensive and restricted (to HD dealers only) equipment.  Shoot, HD won't even let you buy their breakout box tool! 

But, by eliminating all the possible causes you can on your own, you can at least narrow the range of things you eventually pay the dealer's mechanic for (should your own efforts not succeed).

As this case study will demonstrate, the thing you definitely do not want to do is to start replacing parts willy-nilly - in the hopes that you stumble on the cause of your problem.  You can very easily (and quickly) throw a lot of good money down the drain that way.  If an EFI engine is malfunctioning, there can be many, many possible causes - not all of which seem very obvious.  Well, maybe I should modify that.  If indicators make a part seem very suspect and that part is inexpensive, maybe give it a shot, but don't do that with costly items.

The Solution

The solution, in my case, hinged on the use of a diagnostic tool that the HD dealer had (Heritage Harley-Davidson of Edmonton).  I'm not sure what it's proper name is, but basically it is a unit that you can attach to the bike that records data from virtually all aspects of the bike's electronic appendages.  I am also told that it is an expensive little gadget (in the neighbourhood of $14K Cdn).  It is equipped with a control stick that has a button on it (which you tape onto one of the handlebars).  Using it, the mechanic can ride the bike until the problem occurs and, at that instant, press the button to begin recording data. 

When the button is pressed, the unit records all sensor information for a period of 4 minutes prior to the button-press (the unit has a buffer) and for 4 minutes after the button was pressed.  Afterwards, that information can be displayed graphically on a computer for careful diagnosis.

Here's the important part of the graph produced while the mechanic rode my bike:

Notice that big 'step' in the Engine Temperature graph?  That definitely shouldn't happen, right?  Certainly it's expected that, as the engine warms up, that the graph will slowly trend upwards (at a very slight slope), but engine temperature doesn't change in discontinuous steps like that.  There's a sudden, 115 degree Farenheit shift showing here.  Something's gotta be wrong with that blasted engine temperature sensor! 

One look at that graph and I was buying myself a new sensor (no need to pay the dealer to put it in).  Removed the tank.  Replace the sensor.  Put 'er back together and, in the space of an hour, my sentence of frustration was commuted.  The bike ran absolutely perfectly after that!

Lessons Learned From This

That engine temperature sensor was not a cheap part.  It cost me $180 Cdn.  That's not the kind of expenditure you want to make if you are not sure.  Quite frankly, I was very strongly leaning towards fuel injectors being the cause of the problem (largely due to when the problem began, as explained in the section below).  And, without the evidence the graph provided, that is surely what I would have began by replacing first.  And I'd be poorer by $185 per injector with the problem still not resolved.  That shooting-in-the-dark method is a bad way to go.

And, not initially aware there was a sensor-logging unit available for HD bikes, I was extremely reluctant to bring the bike to the dealer.  I had visions of them spending days and days; trying to get to the bottom of this; especially as it was so intermittent.  Who knows?  They could spend a thousand bucks of shop time on it, hand the bike back to me and say, "We couldn't find anything wrong with it."  And then, "Oh yeah, before we forget... here's your bill."  Stuck between a rock and a hard place, man.

I did sort of suspect the engine temperature sensor, but had no solid proof that it was bad.  Regarding it, I did the best I could on my own:  I removed the fuel tank and measured the resistence across the pins of the sensor.  They didn't shown anything unusual, so I was very reluctant to throw $180 buck at it, on a gamble.

What's more, without the sensor graph showing the discontinuity, it would have been very difficult to pick up on this problem any other way.  Both the pre-step 117 degree reading and the post-step 231 degree readings are within the acceptable range of values for the sensor, so there's no bright red flag there to catch one's attention.  Sure, there are ohm readings for various engine temperatures that could be compared (if you have an accurate idea of what the real engine temp should be, compared to the readings), but even that could be missed quite easily. 

What's more bedevilling about a problem like this is the fact that the problem was so darned intermittent.  It's just a matter of luck, getting it to happen while the bike is inside the shop.  Unless you can do it, you'll still be left without a clue.

So, in conclusion, don't throw parts at problems like this.  You really have to get that data-logging unit strapped to your bike to figure out problems like this in the most efficient way.  If your dealer doesn't have such a device, better find one that does.

Background (Original Problem Description)

The bike is a 1998 Sequential Port Fuel-injected Road Glide with about 70K miles on it. Basically stock, aside from having a Stage I kit. The rear base gasket is beginning to sweat, but no major work done on the engine yet. The problem described below has been occurring for a month now; initially not getting better or worse, but recently getting more frequent, it seems.

I should note that the "Intake Air Temperature Sensor" has been going bad on the bike since about March; causing the Electronic Control Module to throw a trouble code indicating that. It didn't seem to be interfering with the running of the bike, so I've been ignoring it (maybe fix it in the off-season). Other than that, no trouble codes are being thrown (I've checked for them many times since the problem described below began). But, perhaps due to the way the storing of repetitive codes gets spaced, that doesn't mean anything. The codes for the Air Intake Sensor are causing the engine light to be lit about half the time, whenever I ride.

I should note that there appears to be no correlation between the engine trouble light - associated with the Intake Air Temp sensor - coming on and the onset of stuttering.

Brief Description of the Problem

It started a few minutes after refueling at a small-town, independent gas station. The engine began stuttering and almost dying every time I tried to apply throttle. Running at a steady RPM, the engine seems to run quite normally; it's just whenever I applied throttle (to accelerate or even when just shifting) that the engine would sputter.

And I mean stutter badly; throwing-me-forward-in-the-seat badly - as if I were letting go of the throttle for a second or two. And then, after belching out some black smoke, it would slowly build up revs to obey the throttle command. Almost un-rideable and quite dicey to take corners with now, as the power-delivery sometimes comes so late.

Initially, the problem never seemed to occur right after starting the bike but, now, it sometimes does. It may be getting worse and more frequent, in fact. It seems to require a few minutes of running before it starts acting up, but sometimes almost right away lately. I should note that, here in Edmonton, the outside temperatures have been getting significantly colder lately. Also, when the stuttering does happen, it keeps on happening for a minute or two, then it runs fine in stretches.

It acts just like a motor being started after being flooded; first dying and then attempting to burn out the excess fuel - black smoke coming out of the pipe, as the engine finally manages to rise in RPM, in response to the throttle's command. The fouled front spark plugs seem to indicate too much fuel (the deposits on the plugs are dry/fluffy, not wet/shiny); the front plug being quite a bit more fouled than the rear.

Another odd thing about it is that the problem is intermittent. For almost every ride, it acts up about 40-50% of the time - maybe even more, lately. The rest of the time the engine runs normally and has good power.

Although I somewhat suspect a bad coil or bad spark plug wire, I've been having difficulty getting to the wire, where it plugs into the coil, to remove it and test it. Any suggestions for an easy way?

But it might not be ignition failure that's causing the fouled plug condition. It seems that the rear plug is getting a bit fouled as well. It's possible it might be the injectors, but I've run a test (below) that seems to rule that out.

Could also be a leaky induction module (the thing that replaces a carburator on fuelies) seal. When I sprayed a garden hose around the induction module while the engine was running (engine warmed up), the idle speed changed dramatically (even momentarily stuttering) in response to the first blast of water. Subsequent sprays only changed the idle a tiny bit (say, 100 rpm), if anything.

Photos of Problem's Symptoms

Front Spark Plug
(photo somewhat washed-out - in reality, it is completely black in colour). Seems to be running quite rich on the front cylinder (although only at times when stuttering, perhaps).

Rear Spark Plug
(this one looks like it's burning a bit rich as well, but not as bad as the front plug)

Just FYI, this is how I have the fuel pressure gauge rigged up for mobile use (had to buy a special fitting, used an old 0-150 psi oil gauge)

This is what the fuel filter looks like (in case you've never seen it). I replaced it, as the old one with 70K miles couldn't be blown through at all (the new one can be blown through easily). Didn't resolve the stuttering.

Things Tried So Far, Leading To No Improvement

• Put in different set of (old) spark plugs

• No Code 12 indicating barometric sensor failure, so unplugged sensor, generating Code 12, to rule our wiring harness being faulty.

• Ran injector-cleaner through 4 tanks of gas and ran menthyl hydrate through as well

• Replaced gas completely

• Replaced fuel filter, as shown above

• Visually inspected regulator, noting no obvious problem

• Checked fuel pressure, which stays constant at near 50 psi, whether idling or revving engine

• Serviced electrical connections (cleaning with elect. contact cleaner and applying dielectric grease), including spark plugs, fuel pump connectors, relays and barrel connector for crankshaft-position sensor

• Replaced with new fuses for both ECM and fuel pump

• Checked battery condition (12.96 volts across terminals)

• Checked resistance in fuel pump (and other) relays - 0.08 Ohms for all

• Removed ECM and mounted it in front of a hair dryer for 20 minutes to dry - in case any moisture existed inside

• Conducted test to determine that injectors are not leaking

• Sprayed water around the induction module (fuel-injection equivalent of a carburator) while engine was running (warmed-up condition), to see if idle speed changes. Initial spray caused the engine to stumble quite badly, briefly and, additional spraying seemed to affect engine idle somewhat (may lowering it by 100 RPM). Hard to say, but that might indicate a leak in the induction module seals. Either that, or the coolness of the water caused the change?