Idle Way Too High.

[BiKenG]

Just replaced a failed O2 sensor and now the bike runs without any errors. But it idles at about 3,000 rpm.

When first started from cold, it seems to idle at about 2,500, which is too high even for a cold start fast idle, but as it warms up, the idle rises to at least 3K and then as I let the clutch out (in gear), the revs go up another 3-400. I assume the latter is normal for a VFR and not a problem if it was idling at the correct 1150. But when it's already racing at over 3K, the last thing you need is it getting even higher as you let out the clutch.

Anyway, something is obviously up here, but there are no errors being reported. The TCP sensor (controlled by the throttle cables) seems to be shutting completely as when the engine is not running, I can hear it snapping shut. But for some reason the engine is idling at a stupidly high speed and the ECU is not reporting any problem. In fact in all other respects the bike seems to be running perfectly.

Anyone any suggestions?

[DaveVXR8]

Have you tried disconnecting the battery for 20 minutes and plugging it in again?

[Switchblade]

Did you do any else to the bike ? change air filter ?

[BiKenG]

Have you tried disconnecting the battery for 20 minutes and plugging it in again?

The battery was disconnected briefly (a couple of minutes at most) since I first ran the bike and found the high idle speed. I'll try it again for a longer period. Is this a recognised procedure? I've not seen it in the manual anywhere.

Did you do any else to the bike ? change air filter ?

The air filter has I'm told been changed, but even if fitted badly, I cannot see how it would affect idle speed, let alone as dramatically as this. My plan is to check all that today.

[BiKenG]

I should point out that this bike was 'drowned' in a ditch and although it now looks immaculate, there of course has been some 'drying out' work done on it - not by me. When I got it there were several fault codes stored, but the only one that was real was the O2 sensor. Now that's been replaced, it generates no fault codes. So this is a process of tracking down the last little problem. I realised there would be a few issues and so I'm not surprised. Just a bit puzzled at the moment.

This is what I'm thinking about the problem.

THE thing about which to be most concerned when designing a TBW/RBW (Throttle/Ride By Wire) system is the possibility of the throttle being held open more than the rider is asking/expecting. That's one of the reasons why there's a lot of duplication within the system, so there's a failsafe, a backup if something should malfunction. With this in mind, it seems to me that if the engine is revving higher than the ECU thinks it should, it would generate an error. So if the throttle is apparently closed, but the engine is still running at 3K rpm, the ECU would notice that and report at error. Ergo, the ECU must be thinking the rider is asking for that rpm.

If the throttle actuator motor is in some way faulty so it doesn't shut the butterflies completely, it must see that as being an error.

If the TPS (Throttle Position Sensor) is reading wrong, so it's reporting butterflies shut, when they're actually still slightly open, again the ECU MUST realise this is an error.

If thee's an air leak, so more fuel is added to compensate, once again the engine is running far faster than the ECU is expecting, so there must be an error.

If the ECU thinks the twistgrip is closed and hence the engine should be idling, it will try to control the engine speed at its programmed idle speed of 1150 and whatever the actual fault, it must report an error if engine speed remains at 3K rpm.

The only thing that makes sense is if the TCP (Throttle Control Position) sensor that is actuated by the throttle cables is reporting the throttle open more than the actual position of the twistgrip. In this case the ECU cannot know about the discrepancy, only the rider will notice that. All the ECU thinks is that the rider is holding the throttle slightly open. It CANNOT know that the twistgrip is actually closed.

This is supported by the fact that even the cold start idle is too fast (about 2,500) and it then rises slightly as the engine warms up. This is exactly what would happen if you held the throttle slightly open. Idle speed would be higher than the automatic cold start fast idle speed which would then be irrelevant and it would of course speed up a bit as the engine warmed to full operating temperature - all the while NOT throwing an error because as far as the ECU is concerned, the rider is asking for that.

However, there are TWO TCP sensors. If only one was faulty, the ECU would throw an error and it seems unlikely to me that both TCP sensors would misread in the same way - which is exactly why all vehicles that use TBW systems run duplicate sensors.

So the only fault that seems to fit the symptoms cannot be occurring. Which means some part of the above analysis is flawed. If anyone has any suggestions about this, they'd be much appreciated.

[BiKenG]

OK, here's another thought. I must be wrong about the ECU noticing the engine speed is too high. If the ECU actually takes no notice of engine speed and simply does what it thinks is right, then the engine speed could be sky high and the ECU would take no notice. So, e.g:-

THE TCP sensor says the throttle should be closed, so the ECU shuts the butterflies and adjusts fuelling to suit the other sensor inputs, like manifold pressure etc, but NOT actually checking the result (i.e. rpm). In this case, if there's an air leak, the ECU would simply respond by increasing fuelling to suit which would result in a higher engine speed than expected, but since the ECU is not checking this, it would report no error.

So it looks to me like the ECU is not as smart as I first thought and is NOT checking engine speed and hence is not aware that even though it's supposed to be idling, in fact it's racing at 3K rpm and no error will be generated.

This then would be the flaw in my earlier analysis. The ECU does NOT check the result of its 'commands'. ECU says 'IDLE', but the engine races and the ECU doesn't even notice.

Does anyone have any more in depth knowledge of how the ECU operates?

[Switchblade]

Is there anything in the service manual about re-calibrating the throttle after ECU or TPS replacement ?

[BiKenG]

Is there anything in the service manual about re-calibrating the throttle after ECU or TPS replacement ?

Basically, no.

In the Troubleshooting section, it mentions idle too high as a symptom with the following diagnostic procedure:-

1. Check idle speed
2. check throttle operation and freeplay
3. check the TBW system

None of which is much help because:-

1. The manual section that relates to this just says it's fixed and cannot be adjusted.
2. The cables are not too tight and I can hear the TCP sensor clicking as you close the twistgrip, so it IS shutting completely.
3. Yes, good idea. But all the checks are based on the DTC error (fault code) that has been displayed. But there are no errors.

Looks like I'm going to have to pull off the throttle bodes and just have a look around, but it would be better if I knew what to look for.

[Tamworth]

If it started happening after you replaced an 02 sensor then that is where your problem lies

[BiKenG]

If it started happening after you replaced an 02 sensor then that is where your problem lies

Sorry if I wasn't clear. When I received the bike it was idling too high and there was an O2 sensor error. I've replaced the sensor, cleared the DTC and ridden the bike long enough to see that NO codes are now being generated, but it is still idling too high. So the O2 sensor replacement and idle speed issues are not related.

I think it's time to take the throttle bodies off and have a good look.

[SimonAudi]

Before you take that apart - maybe it's better to pull all the electrical connectors between the throttle, throttle body and ecu apart and check they are clean and dry and corrosion free - corrosion can cause a slight short circuit that may be the cause

Then spray everything with an electrical contact cleaner and reassemble everything

I once had a corroded rear break light bulb that stopped my Gilera 50 running - cleaned up the connector and the bike would start and run

Strange things happen with potential shorts

[DaveVXR8]

Water in the ECU if it went for a swim?

[BiKenG]

Water in the ECU if it went for a swim?

Well yes it was downed, but the ECU is sealed and all the relevant connectors are also sealed so wouldn't have got wet. All those I have had apart are dry and pristine inside. It's not an old bike and never ridden in the rain (which is somewhat ironic)

If the ECU had suffered from water ingress, I think there'd be more obvious problems.

[InlineSk8r]

I'll venture a guess and say the throttle cable got water in it and possibly turned to rust. My bicycle's back brake cabled got stuck this weekend. Problem was water in cable rusted so it didn't move freely.

Good luck on finding the problem

[BiKenG]

I'll venture a guess and say the throttle cable got water in it and possibly turned to rust. My bicycle's back brake cabled got stuck this weekend. Problem was water in cable rusted so it didn't move freely.

Good luck on finding the problem

No, not throttle cable. Everything works perfectly smoothly and in any case, as the twistgrip is shut, I can hear the clack of the TCP sensor hitting the bottom stop so it is definitely closing all the way.

[Switchblade]

I guess the O2 sensor was for your bike . Wide band ?

[BiKenG]

I replaced the O2 sensor with the std Honda part. The bike is completely standard.

Having now got the top of the air box off, I can look at the throttle butterflies while the engine is running. With the ignition, or the kill switch off, the butterflies do not move with the twistgrip (hardly surprising). Switch to RUN mode and the butterflies flick open briefly then back to the idle position (more on that later). If I touch the butterflies or shaft, I can feel a vibration, a twitching which I think must be the effect of the ECU constantly controlling the servo motor and monitoring their position. With the engine off, the butterflies only open to about a quarter and no more. It's possible this limit applies also when the engine is running but just idling (in neutral or clutch in) as some sort of safety measure. Just guessing there, but would make a certain amount of sense.

Start the engine and it idles at 3K, with the butterflies not quite closed. Turn the engine off and I can see the butterflies snap completely shut. They close further with a noticeable movement. This means that when my bike is idling, they are not shut with engine speed being controlled solely by a fixed air bleed. The engine speed IS being controlled by the butterflies, but is that my actual problem? Maybe they should be completely shut at idle as I originally surmised and the fact they are slightly open is causing the fast idle. Under no load, it would require very little opening of the butterflies to raise the engine speed to 3K rpm, so the amount I can see them open could be causing it.

The problem now is, I do not know whether this is correct. I can find no information about the air flow through the throttle bodies and what SHOULD actually be occurring. Typically this is all 'black box' stuff - secret. Honda's solution if there's anything apparently up with any of the throttle body components - replace the entire assembly. Well, at £2,500 that's not gonna happen and I think this sort of profiteering should be made illegal, but that's a whole other issue.

So, does anyone have any advice about how the idle speed might be controlled? Should the butterflies be completely shut with just an air bleed controlling the airflow? Or is the idle speed simply controlled by the butterflies as the engine watches the rpm to keep it at the correct idle speed? In which case, why is mine idling at 3K, without any error being generated?

[Switchblade]

Just for trouble shooting put the old O2 sensor back in and see if it returns to normal and measure your voltage at the old O2 sensor and compare to the new one .

[BiKenG]

Thanks for thinking about my problem, but the O2 sensor is NOT the problem. The bike was idling fast with the old sensor (which was open circuit between heater wires, i.e. buggered) and is still doing it with the replacement sensor (whose heater circuit measures correctly). The difference is that there is now NO O2 sensor error being shown when testing the bike.

In fact it's hard to see it being any other sensor as it would register a fault, but the ECU seems happy that everything is correct.

Since there are no errors being generated, the ECU must think the throttle should be slightly open. Which indicates a mismatch somewhere between what the twistgrip (via the TCP/IP sensor) is actually requesting, and what the ECU 'thinks' is being asked for.

If only Honda published detailed information about all this, troubleshooting would be so much easier.

Sent from my iPad using Tapatalk

[Mohawk]

Ken, Check the TB for a means to manually adjust the idle &/or the TPS/throttle plate servo position. Every EFi I've ever seen has the butterflies fully closed at idle ! Pretty certain that is your issue.

The control servo will be a stepper motor hence the vibrations you feel, but they don't normally vibrate unless they can't achieve a steady state & are working against a load. I assume that FbW throttles have a fail safe spring like normal throttles, so that might be the load.

Sounds to me like the TB sensors are out of adjustment. The butterflies should be closed on a dead engine, they should run full cycle when ignition is turned on & before it will attempt to start. That pre-start sequence is to ensure the sensors & ECU can see full movement & return to calibrated location.

[Switchblade]

Thanks for thinking about my problem, but the O2 sensor is NOT the problem. The bike was idling fast with the old sensor (which was open circuit between heater wires, i.e. buggered) and is still doing it with the replacement sensor (whose heater circuit measures correctly). The difference is that there is now NO O2 sensor error being shown when testing the bike.

In fact it's hard to see it being any other sensor as it would register a fault, but the ECU seems happy that everything is correct.

Since there are no errors being generated, the ECU must think the throttle should be slightly open. Which indicates a mismatch somewhere between what the twistgrip (via the TCP/IP sensor) is actually requesting, and what the ECU 'thinks' is being asked for.

If only Honda published detailed information about all this, troubleshooting would be so much easier.

Sent from my iPad using Tapatalk

Ah miss that yeah it sound like your ECU is trashed !

[BiKenG]

Ken, Check the TB for a means to manually adjust the idle &/or the TPS/throttle plate servo position. Every EFi I've ever seen has the butterflies fully closed at idle ! Pretty certain that is your issue.

I agree. I would expect the butterflies to be closed and idle speed controlled by some other air bleed/bypass. There appears to be NO adjustment. Honda's technical info just says, it will be set correct by the ECU with no real troubleshooting when it's wrong.

Sounds to me like the TB sensors are out of adjustment. The butterflies should be closed on a dead engine, they should run full cycle when ignition is turned on & before it will attempt to start. That pre-start sequence is to ensure the sensors & ECU can see full movement & return to calibrated location.

It's almost as if the TCP sensor (driven by throttle cables) and TP sensor (actual butterflies position) are somehow out of sync. But no obvious procedure to re-sync them.

The control servo will be a stepper motor hence the vibrations you feel, but they don't normally vibrate unless they can't achieve a steady state & are working against a load. I assume that FbW throttles have a fail safe spring like normal throttles, so that might be the load.

Yes, there is a return spring, but when the engine's off, the spring doesn't seem capable of closing the throttle and I don't know if it should. I would think the drive from the servo is geared down, so in reverse, moving the butterflies would require spinning the motor quite fast. But I still would have thought it should happen easily. I cannot discount this as maybe being, or related to the problem.

What does seem to be the case is that the butterflies are for some reason slightly open while idling when they should be closed and the ECU thinks this is correct. As if the ECU thinks the twistgrip is not completely closed and you're asking for more engine speed.

[SimonAudi]

Hi - As your idle speed increases as you let the clutch out then I think there is only 2 common things

1. Proximity of hydraulic line and throttle line

2. The electrical sensor on the clutch is connected to your ECU via the clutch / stand safety circuit - the same ECU as the TPS and TB.

So please check you do not have trapped cables around the headstock - in idle - what happens if you turn from end to end on headstock lock

What happens if you put it in gear on centre stand - with side stand up and down? Then try headstock movement - end to end in both senarios.

The other thing you could do is call Don Guhl in USA and ask him about the TBP and TPS and the idle speed - as he has been inside the workings of the ECU - he may know how its supposed to work.

[Mohawk]

Try this trick. You'll need a helper or tape. Disconnect power to bike. Open the throttle on bar approx 10% then lock in that position, tape or helper.

Reconnect power & turn on ignition, observe butterflies with previous known power off / on position.

If position is more open, then its NOT the throttle, if position is the same, release throttle & do butterflies close ? If yes, then this may be it fixed !

Turn off & on again, do butterflies hold the closed position ? If yes, open throttle fully, do butterflies reach full open position ? & then return to fully closed when throttle returned to stop ?

Worth a try.

[SimonAudi]

In Neutral and lower gears that wont work - 5th is OK - the TPS allows the throttle bodies to open fully in 4th and 5th - in 1,2 &3 the wide open throttle does not equal wide open Throttle bodies - Honda limits the power in 1,2,3 and 6 (top speed) in Europe bikes - via Throttle bodies - spoke with Don on that yesterday.