Hard start has me thinking

[jaeger22]

Quote from: autoteach on November 24, 2014, 10:55:47 pm

my chevy blazer, my subaru, kawasaki watercraft... I have seen this occur on far more fuel injected vehicles than you can imagine.:http://forums.pelicanparts.com/uploads15/throttle+response+cam1250899088.jpg




Apparently you're just not living right. I drive cars and motorcycles too and I've never experienced this problem with any of them except the Super Tenere.

Exactly right! It is true that almost any system, EFI or Carb, can be made to screw up either on purpose or by accident. But that is NOT what is happening here. I am really sick of those that have not ever even seen the problem implying or outright saying that is is an operator error. That is pure BS. Let me put it in perspective:
I have put somewhere around 400,000 miles on trucks and cars with EFI. Zero hard starts.
I put 93,000 miles on my V-Strom. Zero hard starts.
I even have 14,000 miles on my home made EFI system on my DR-650. Zero hard starts.
I have 45,000 miles on my S10. 5 hard starts!
So did I suddenly get stupid when I bought the S10? And did all the other knowledgeable and experienced owners that have experienced this on ONLY this bike suddenly get stupid also?
Yes some of my hard starts have been after a prior short run and shut down. But some were not. I just pushed the button like I always do and got the hard start. I never give it any throttle unless it fails to start right away. Hundreds of crisp starts in a row and then bingo. Hard start. Every time but once it eventually started with WOT. One time it ran the battery down completely. Fortunately I had an Antigravity Micro Start battery with me and jumped it with that and got it started eventually with more WOT. (I was 350 miles from home).
I still don't think this is a MAJOR issue with the bike, all bikes have some warts and this is really a relatively minor annoyance in the big picture. But PLEASE don't be so arrogant as to tell those of us us that HAVE experienced it that it is our own fault and not the bike. It just isn't true.
I have made my share of mistakes and have admitted some of my biggest ones here on this forum, but how I push the magic button is not one of them.

 

[EricV]

Pay attention, it's worth it.

You may believe you did nothing unusual or out of the ordinary. However, it's far more likely that you did, but you simply don't realize it. Or you simply have a habit that doesn't work well for the Yamaha. We often forget seemingly minor details when we come back to the bike a few days or weeks later and it doesn't start. If you examine the times you've had a hard start, I honestly suspect you will find some common aspects. That would be your first clue.

And yes, I have experienced the hard start issue, both in the garage and on the road. And I experienced it with the FJR before too. Exact same causative reasons. For me, finding those common parameters meant being able to mostly avoid them and the hard start. I have on occasion simply forgotten to follow the correct procedures and it's bit me with a hard start.

Is it an issue we wish Yamaha would spend some more time examining and hopefully correct? You bet! Would you go buy a new bike to get rid of that 'feature/problem"? Probably not, but to each their own.

 

[yz454]

Had hard start 3 times in the first 5 thou now with 51 thou I can,t get to happen . And two of those were playing with the key . All my fault . I don,t baby my bike like most , it,s always out side in rain or shine and snow . Stock battery still ,go push button , go through warm and go .

 

[autoteach]

Lets be clear here, again... I for one am not implying that anyone has turned stupid and THAT is why they are having hard starts. I am implying that you are stupid IF you disregard how a FI system operates, the inputs, the algorithms, the outputs, the environmental factors, as well as the mechanical design of the engine and fail to make it not start within ten seconds by following industry wide standard operating procedure.

If it rains outside, grab an umbrella. Quit trying to figure out how to stop the rain.

 

[Checkswrecks]

Mark -


I've not only read the posts for a very long time, I've intentionally caused mine to hard start and not said a word. (It was difficult and took a combination of keying the fuel pump AND lightly blipping the throttle.)


I said "most common" and gave one example of how there could be others. But I've got 40 years as a licensed mechanic, have taken the classes, trouble-shot and torn down injectors, and know more than enough history for me to be comfortable in using the words "most common" when generically relating the operator to hard starting of fuel injected engines. Autoteach sounds beyond my experience with these smaller engines and is somebody I listen to.


What I did not say in my post was that what I posted were the ONLY things which could cause the hard start. In having read these posts for a long time, I do believe that hard starts seem to be over-represented on this model. So even if "most common" is the rider, I'm wondering if a smaller number have a combined stack of factors. Potential contributors that I've wondered about are the big engine's starting shake dumping the rail in lieu of the operator's hand, a weak internal spring, 'Teach's thoughts on the compression release, a bubble in a fuel line forming after shutdown, or even a brand of gasoline which affects the O2 sensor.


Until somebody really has solid evidence to move our collective knowledge further along, the little amount of prevention stated and WOT when the problem does happen is still the same and all we have.

 

[Tremor38]

At the risk of adding to the operator error/you dissed me/it's all yamaha's fault discussion, I'd like to add some practices that avoid exacerbating the situation if you are already in it. The best way is to continue cranking and go to WOT. If you stop cranking prior to trying WOT immediately hit yhe kill switch. If you turned off the key intead. Hit the kill switch before you turn the key back on and wait 5sec before returning the kill switch to 'on.'

 

[AVGeek]

I know this is a contentious subject, mostly because the issue is so seemingly random. Let's focus on the circumstances surrounding the hard start, and collect hard data (ambient temp, fuel used, etc.) to see if we can definitively answer the question as to why this occurs.

 

[Mike B]

Was talking to a Harley rider at work one day about this problem he told me his bike did the same thing.

He said when he had to turn his bike off before warming up he would use the kill switch so the ecu will remember
where it was at and what it was doing at the time before turning it off with the switch. and never had a hard start
when doing that.

I have never tried that because I just don't turn it off before the fan comes on or go on a long ride.

 

[autoteach]

"Hard Data" will have to include all data points that influence fuel injection. If you dont report all of it then it is merely circumstantial evidence. So, if anyone wants to collect the data so that we can get to the bottom of it we need the following from the bike:
Throttle position
throttle position desired
engine temp
intake air temp
cranking rpms
actual crank position at beginning of cranking sequence
manifold absolute pressure
barometric sensor
Oxygen sensor- which isn't active on the bike, but maybe data from a 5 gas analyzer would be nice.
Desired injection in milliseconds
fuel pressure
fuel being used
any modifications made to the bike

Environmental data points:
actual air temp
actual baro
humidity


And then any human issues such as thumb slipped off the start button, and then all the data from previous run cycle including run time, etc.

Once we have this...then magic might just occur and we will solve this better than those total wankers that programmed this originally.

 

[arjayes]

"Hard Data" will have to include all data points that influence fuel injection. If you dont report all of it then it is merely circumstantial evidence. So, if anyone wants to collect the data so that we can get to the bottom of it we need the following from the bike:
Throttle position
throttle position desired
engine temp
intake air temp
cranking rpms
actual crank position at beginning of cranking sequence
manifold absolute pressure
barometric sensor
Oxygen sensor- which isn't active on the bike, but maybe data from a 5 gas analyzer would be nice.
Desired injection in milliseconds
fuel pressure
fuel being used
any modifications made to the bike

Environmental data points:
actual air temp
actual baro
humidity


And then any human issues such as thumb slipped off the start button, and then all the data from previous run cycle including run time, etc.

Once we have this...then magic might just occur and we will solve this better than those total wankers that programmed this originally.

autoteach - Why don't ECUs keep a log of all sensor inputs? Let's say that all of the ECU inputs comprise 512 bytes. Storing 512 bytes once per second for one hour comes to 1.8MB. It would add very little cost to the system to have, say, 128MB of nonvolatile storage for data logging. Then the information could be accessed and issues like the hard start could be diagnosed. Of course the information could be used for other purposes as well:

http://www.usatoday.com/story/money/cars/2014/05/25/black-boxes-for-cars-raise-privacy-concerns/9556615/

I personally would be fine with having this info logged on my bike, even knowing that it could potentially work against me in a legal case. Not everyone would feel that way I'm sure. But the benefits of having that information available for diagnosing problems are obvious.

 

[racer]

I had an 04 FJR that would occasionally give me a hard start, which is part of the reason I got rid of it. It would usually happen after sitting for a few weeks, even on a battery tender. My S10 gave me one hard start after the one and only time I filled it with high octane gas. I burn regular now, with out any problems. One time when it was really hot outside, it acted liked it wouldn't start, but I went WFO and it fired. When its hot out this is my starting procedure. Kill switch off, turn key on, flip kill switch to run and immediately hit the starter button. When I do this, it turns over a couple more times than usual before firing, but when the injectors fire, the engine is already turning and will fire before they have a chance to flood the cylinders.

All of this could be BS thinking on my part, maybe a few of the more knowledgeable readers will chime in.

 

[autoteach]

autoteach - Why don't ECUs keep a log of all sensor inputs? Let's say that all of the ECU inputs comprise 512 bytes. Storing 512 bytes once per second for one hour comes to 1.8MB. It would add very little cost to the system to have, say, 128MB of nonvolatile storage for data logging. Then the information could be accessed and issues like the hard start could be diagnosed. Of course the information could be used for other purposes as well:

http://www.usatoday.com/story/money/cars/2014/05/25/black-boxes-for-cars-raise-privacy-concerns/9556615/

I personally would be fine with having this info logged on my bike, even knowing that it could potentially work against me in a legal case. Not everyone would feel that way I'm sure. But the benefits of having that information available for diagnosing problems are obvious.

There is storage available in the computer. So far this has been mainly "failure data" where a snap shot at the time of a code based failure is saved and all engine and measured environmental data is stored. This can be helpful if a sensor is only occasionally giving a false reading but doesn't help in the instance of a bad alternator (reg/rec portion on a motorcycle), for example, where the vehicle receives an instantaneous voltage spike and the computer self protects and shuts down. Also, some companies like Bombardier are using the memory to store information such as time spent at WOT. This comes in handy for "research" purposes when you have engine failures and you are trying to duplicate running conditions. We had a customer that had over 80% WOT for all run time on his sled with the motor finally giving up at 6k miles. The dealer that scanned it wanted to know what the hell he was doing to amass that much WOT.

 

[Checkswrecks]

I had an 04 FJR that would occasionally give me a hard start, which is part of the reason I got rid of it. It would usually happen after sitting for a few weeks, even on a battery tender. My S10 gave me one hard start after the one and only time I filled it with high octane gas. I burn regular now, with out any problems. One time when it was really hot outside, it acted liked it wouldn't start, but I went WFO and it fired. When its hot out this is my starting procedure. Kill switch off, turn key on, flip kill switch to run and immediately hit the starter button. When I do this, it turns over a couple more times than usual before firing, but when the injectors fire, the engine is already turning and will fire before they have a chance to flood the cylinders.

All of this could be BS thinking on my part, maybe a few of the more knowledgeable readers will chime in.

I've seen fuels and definitely fuel additives change O2 sensor outputs, so it's not necessarily BS. In the 1970s there was an additive called Bug Juice for racing which was high in sulfur and it was absolutely horrible on sensors. Your one-time issue was more likely the station's tank, the batch, or possibly the brand you filled with, not necessarily that it was premium.


But as 'Teach and AvGeek wrote, we do need real data.

 

[EricV]

"Hard Data" will have to include all data points that influence fuel injection. If you dont report all of it then it is merely circumstantial evidence. So, if anyone wants to collect the data so that we can get to the bottom of it we need the following from the bike:
Throttle position
throttle position desired
engine temp
intake air temp
cranking rpms
actual crank position at beginning of cranking sequence
manifold absolute pressure
barometric sensor
Oxygen sensor- which isn't active on the bike, but maybe data from a 5 gas analyzer would be nice.
Desired injection in milliseconds
fuel pressure
fuel being used
any modifications made to the bike

Environmental data points:
actual air temp
actual baro
humidity


And then any human issues such as thumb slipped off the start button, and then all the data from previous run cycle including run time, etc.

Once we have this...then magic might just occur and we will solve this better than those total wankers that programmed this originally.

While I don't disagree with your list, most owners are not going to be able to easily provide a data point for everything you have listed. It might be more constructive for the forum members if you trimmed that down to a simple list of questions, and where necessary, an explanation of what data you were looking for, if there is a serious attempt to gather data from multiple hard start events. Starting a Sticky for just that data gathering might help too.

As to the wankers that programmed this originally... Don't forget human nature. They didn't start with a blank slate, they took 90% of what "they always do" and tweaked that to get the end results. Their process is likely part of the problem, which is why this issue crosses platforms.

I don't expect the limited amount of data to provide a true answer. I do expect that the commonalities will start to become apparent at some point, which may help people to understand and avoid the problem more. No issue experienced by so many people is truly random.

 

[markjenn]

Good discussion. While having every piece of data would be ideal, you might be able to gather some useful data just with the things that the operator could easily log like ambient temp, starting process, and previous shutdown circumstances.

I've never heard anything definitive, but I doubt the Yamaha ECU records anything affecting startup between key cycles....when you cycle the key, the ECU reboots and "wakes up" fresh, reading all inputs from scratch. Assuming this is true, any "memory" of the previous start and shutdown is being recorded "physically" (e.g., extra fuel in the inlet tract), not in the ECU. I personally don't think key cycling or use/non-use of the kill switch is likely a factor - when someone can cycle the key 50 times with and without throttle input and the bikes starts fine and do this over and over, this pretty much eliminates this as the smoking gun in my book.

One idea I had.... what if the inlet air temp sensor could somehow "stick" or otherwise malfunction in an odd way such that its resistance goes "off scale high" at startup? The IAT sensor is a thermistor which changes its resistance with temp. If the thermistor occasionally (and randomly) sticks as an open circuit, this would be be interpreted by the ECU as off-scale low temp and the ECU may just then assume it is as cold as it can possibly get and make the mixture accordingly rich. And nearly all the time, even this mixture will fire, but if the previous startup/shutdown cycle left the plugs even slightly wet-fouled, this is the straw that breaks the camel's back and the bike floods. This would provide both the random element that can affect bikes with no startup/shutdown anomaly, but make such anomalies a factor. Having said this, thermistors are pretty simple and foolproof devices and I don't know if they have any propensity to fault randomly. I would sure love to have a log of startup IAT readings for a bunch of bikes.

Someone mentioned that maybe Yamaha's "process" of developing motor tuning code is flawed. This is an interesting idea. I certainly think Yamaha has had more issues delivering bikes with odd EFI fueling issues over the years than almost any other mfg. Certainly the S10 EFI has received its fair share of criticism. The FZ1 problems were epic and the FZ09 is also receiving a lot of flak.

Just speculating.

- Mark

 

[Koinz]

One idea I had.... what if the inlet air temp sensor could somehow "stick" or otherwise malfunction in an odd way such that its resistance goes "off scale high" at startup? The IAT sensor is a thermistor which changes its resistance with temp. If the thermistor occasionally (and randomly) sticks as an open circuit, this would be be interpreted by the ECU as off-scale low temp and the ECU may just then assume it is as cold as it can possibly get and make the mixture accordingly rich. And nearly all the time, even this mixture will fire, but if the previous startup/shutdown cycle left the plugs even slightly wet-fouled, this is the straw that breaks the camel's back and the bike floods. This would provide both the random element that can affect bikes with no startup/shutdown anomaly, but make such anomalies a factor. Having said this, thermistors are pretty simple and foolproof devices and I don't know if they have any propensity to fault randomly. I would sure love to have a log of startup IAT readings for a bunch of bikes.

- Mark

I was thinking along the same lines except it might be more feasible that the coolant temperature sensor might have more to do with it or maybe it's a combination of the two. Both provide signals to the ecu to control fuel injection duration (rich or lean conditions).

 

[autoteach]

While I can understand that you want a simpler list of things for the consumer to grab, that is not how you diagnose a problem. When was the last time that you went to your mechanic with a problem on your car and he asked for some data points and you told him that you ordered a green tea for breakfast the day it started, it was snowing, and you put in 87 octane and he could diagnose the issue. I keep saying it, and all y'all keep ignoring it, this is a tougher problem then fuel octane or outside temperature. As for the wankers, I was being sarcastic. The likely culprit is a variance in measured and actual values or a compounding error in fuel map corrective factors. Who am I, though? Maybe someone can solve this with fuel octane and I am just being an obnoxious jackass.

I will say, lower octane will tend to help this scenario...as I think it is highly correlated to the decompression. This I base from combustion theory and knowledge.

 

[EricV]

While I can understand that you want a simpler list of things for the consumer to grab, that is not how you diagnose a problem. When was the last time that you went to your mechanic with a problem on your car and he asked for some data points and you told him that you ordered a green tea for breakfast the day it started, it was snowing, and you put in 87 octane and he could diagnose the issue. I keep saying it, and all y'all keep ignoring it, this is a tougher problem then fuel octane or outside temperature. As for the wankers, I was being sarcastic. The likely culprit is a variance in measured and actual values or a compounding error in fuel map corrective factors. Who am I, though? Maybe someone can solve this with fuel octane and I am just being an obnoxious jackass.

I will say, lower octane will tend to help this scenario...as I think it is highly correlated to the decompression. This I base from combustion theory and knowledge.

I used to tell customers that premium was like cardboard, regular like paper. More octane, harder to ignite, (duh, resistent to pre-ignition), lower octane, easier to ignite.

My point to your list was that some data is better than no data. Few people are prepared to generate much of the data points on your list. Some data points, might start to point in a particular direction, which may or may not help owners.

Understanding how things work gives you an intuitive ability to make a jump when hints point in a particular direction.

Personally, I think the cylinders are getting washed down and that's the primary loss of compression, but I can't tell you what factors cause that to occur.

 

[KCW]

Our Yamaha dealer on the island is very bad. HOWEVER, I mentioned the "hard start" to their tech one day, and he told me that he's heard of it discussed before during dealer conferences, and it's caused by a mixture rich condition caused by a bug in the ECU. He claims the fix that Yamaha has asked dealers to suggest for it to get it to start right away is the following:

(1) Identify hard start and stop cranking
(2) Turn kill switch to "kill" position
(3) Turn off key
(4) Wait 5 seconds
(5) Turn on key
(6) Turn kill switch to "run" and IMMEDIATELY start cranking
(7) Bike starts

The two times I've experienced the hard start the above has given me an immediate start, almost like a normal start!

 

[autoteach]

I used to tell customers that premium was like cardboard, regular like paper. More octane, harder to ignite, (duh, resistent to pre-ignition), lower octane, easier to ignite.

My point to your list was that some data is better than no data. Few people are prepared to generate much of the data points on your list. Some data points, might start to point in a particular direction, which may or may not help owners.

Understanding how things work gives you an intuitive ability to make a jump when hints point in a particular direction.

Personally, I think the cylinders are getting washed down and that's the primary loss of compression, but I can't tell you what factors cause that to occur.

Sorry, I was/am just getting tired of the armchair mechanics who know it has to be the fact that they looped the block counter clockwise before they pull in and got off the right side of the bike and if they just do that opposite next time it will be fine. As for the "bug", read my description of the factors, that is what contributes to "misfueling", not to be confused with missed fueling.