Making the Dog run as it should

[kiwiape]

Hello Mark! I am a Swedish seaman and I live in Donsol Sorsogon. I bought one Husqvarna terra 650 and I have had it for little over a year with plenty plenty problems with misfire, backfire and it stops in the middle of nowhere many many times and it drives me insane!!! I just want to ask you how's your Husqvarna before I take it to court to get my money back cause I had it with that shit after they tried to fix it for me three times and the last time they didn't even let me testrun it unless I paid for the five months i refused to pay before it runs like clockwork!!! Please reply if you have time!!! Thanks a lot!! Regards Ted Litens

Mine use to run like yours (very bad when cold) ....I put the Plustars in and now (when cold) its 95% better...it just splutter a little bit and and I have slight miss on/off throttle sometimes...did 430 yesterday open road and slow bush track, it missed three times over the whole day....no big deal. I believe when I put the bigger injector in it will be 99.9% , but for now it is ok and not really a problem, when it was a REAL problem before. It also pulls much better with the Pulstars.

Do you know what ECU map your bike is on...? Get the print out from your dealer and let me know.
http://www.cafehusky.com/threads/random-stall-–-a-silver-bullett-or-as-close-as-we-may-get-tr650-husky.80593/#post-539119

 

[engineerk9]

Mine use to run like yours (very bad when cold) ....I put the Plustars in and now (when cold) its 95% better...it just splutter a little bit and and I have slight miss on/off throttle sometimes...did 430 yesterday open road and slow bush track, it missed three times over the whole day....no big deal. I believe when I put the bigger injector in it will be 99.9% , but for now it is ok and not really a problem, when it was a REAL problem before. It also pulls much better with the Pulstars.

Do you know what ECU map your bike is on...? Get the print out from your dealer and let me know.
http://www.cafehusky.com/threads/random-stall-–-a-silver-bullett-or-as-close-as-we-may-get-tr650-husky.80593/#post-539119

What Dealer?

I think WayneC has all those details, taken from his GS911 readings of a week ago - as far as I understand the ecu firmware was the very latest available, by Dahlitz in Canberra

I've ordered both Pulstar plugs & large injector

It would appear that the stalls are at the point from transition from Open to Closed Loop at 52degC - it only does it once - had a great run on Saturday up into the Blue Mountains & back, mix of riding conditions - ran really well

The occasional, infrequent miss-fire is a mystery which has only occurred since the ecu upgrade

The ecu chasing the 14.7 appears to be the "killer"

 

[WayneC]

Keith, I emailed you the information last week, you have the same version of firmware now as KiwiApe, the cold start change over from open loop to closed loop temperature I referred to was 42c not 52c. Roger has found with tests they did on DRZ Charlie's machine that on his machine it is too rich initially rather than too lean but unfortunately there is no information as to which version of firmware is in that machine

The bottom line is the simple things need to be done before anything else and these are

1/ Ensuring the machine has the latest version of firmware, it is very different to earlier versions
2/ Ensure the idle actuator and airways are cleaned. Many machines have had dirt through the airbox and inlet manifold. We should thank 650x guys for this one
3/ Ensure BMS-E adaption resets are done after any sensor/air filter/exhaust changes

There are misunderstandings on adaptions with the programmable BMS units, there are multiple tables, 02 and idle actuator are separate but BMW reset them together. GS911 separates them. Consider them long term settings. They marry electronic and mechanical components together taking care of tolerances, not my words by the way, they are BMW's words (my paraphrasing though )

There are likely other short term trims of the devices but further real time testing, not just with GS911 is needed to really understand it all

It really would be worthwhile starting the machine and then putting on the helmet etc and allowing it to warm a little more, others have commented it likes it

 

[Roger 04 RT]

Keith, I emailed you the information last week, you have the same version of firmware now as KiwiApe, the cold start change over from open loop to closed loop temperature I referred to was 42c not 52c. Roger has found with tests they did on DRZ Charlie's machine that on his machine it is too rich initially rather than too lean but unfortunately there is no information as to which version of firmware is in that machine

The bottom line is the simple things need to be done before anything else and these are

1/ Ensuring the machine has the latest version of firmware, it is very different to earlier versions
2/ Ensure the idle actuator and airways are cleaned. Many machines have had dirt through the airbox and inlet manifold. We should thank 650x guys for this one
3/ Ensure BMS-E adaption resets are done after any sensor/air filter/exhaust changes

There are misunderstandings on adaptions with the programmable BMS units, there are multiple tables, 02 and idle actuator are separate but BMW reset them together. GS911 separates them. Consider them long term settings. They marry electronic and mechanical components together taking care of tolerances, not my words by the way, they are BMW's words (my paraphrasing though )

There are likely other short term trims of the devices but further real time testing, not just with GS911 is needed to really understand it all

It really would be worthwhile starting the machine and then putting on the helmet etc and allowing it to warm a little more, others have commented it likes it

Wayne, You are right on. While dissecting the fueling for several F800, R1150 and R1200 models the GS-911 and LC-2 made a powerful combination that allowed accurate assessment of how the ECUs worked, and what changes worked & didn't. Rider inputs are valuable, especially in the context of good data. BTW, the GS-911 guys have done a great job of improving the GS-911 Wifi software for the newest models, fixing several bugs and adding a great set of features. Kudos to them.

I will go so far as to say that GS-911 and LC-2 provide much more information to the tuner, than an inertial dyno. More of the bikes performance is measured, the measurements are made on the road and therefore are much more relevant to the rider.

 

[Mark_H]

We finally managed to get together and do some real time logging with a GS911 from a Terra with the latest firmware.................
Secondly the data indicates the temperature from the AIT is very stable and a reasonable indication of actual air temperature. The AIT is mounted further back than the AIT on the 650GS BM and I expected to see some engine heat influences but saw less than I would normally expect from the GS in city traffic at the start and end of the 30 mins riding which was logged. It is admittably winter here at the moment so some summer data would be worth logging

Interesting information, but I find the above regarding the AIT sensor difficult to understand. OK it was winter but you did not give any actual temperature readings or info.

The TR650 OEM sensor siting is:

a) directly above the engine and exhaust;
b) in the flow of the cooling fan

I can confirm that in heavy traffic on a hot day the readings are far from correct or stable and are very much influenced by the engine heat.

Stationary in traffic on a hot day my AIT sensor suffered from the engine heat and fan airflow to give readings as high as 20c over and above actual. This is fact based upon measurement of actual air temperature against dash reading. ie 53c shown as opposed to 33c it should read.

With the dash reading high was when the stalling issues were at their worst and the bike was crap to ride in traffic.

Re-siting the sensor first to the front of the bike gave much more stable and realistic readings, however I felt there was room for more accuracy. That is why I placed my sensor in the right front beak at the air intake so it is measuring the temperature of the air entering the airbox. The new position means that the sensor is no longer affected by exhaust and fan air and reads much nearer to the real air temperature, particularly in traffic.

Many other TR650 owners have done the same and have gained similar improvements in the way the bike runs on a hot day.

Your post implies that there is no issue with the OEM siting of the AIT Sensor, when in fact for most of us there certainly is.

 

[Mark_H]

The bottom line is the simple things need to be done before anything else and these are

1/ Ensuring the machine has the latest version of firmware, it is very different to earlier versions
2/ Ensure the idle actuator and airways are cleaned. Many machines have had dirt through the airbox and inlet manifold. We should thank 650x guys for this one
3/ Ensure BMS-E adaption resets are done after any sensor/air filter/exhaust changes

How does one go about ensuring point 1?
ie How does one find out if the bike has the latest firmware and if it does not, how do we obtain it?

Remember that for many of us, there is no MOSS and no dealer support whatsoever!

 

[WayneC]

Mark

As I said in the earlier post, put a call out for BMW owners in your area with a GS911 to run an autoscan to determine the firmware revision, you could also listen to what KiwApe has been saying, he has some answers on it as well

KiwiApe is going about things the right way, he determined his firmware version first then looked at ways of improving combustion with the Pulstar plugs and the injector. He has then tested each component individually with proper adaption resets to ensure his results are valid, he has been even more pedantic than I in the process

With all the other testing etc which has been done because we dont know firmware versions on each machine tested/measured we cant compare apples with apples and then know what settings to apply to the next cab off the ranks and it is all based on the feed it more fuel mantra which is not always correct

Re the air intake and AIT, the first time I saw the air inlet system I thought it would be affected by engine temperature due to the position and our experience with the 650GS. That is the reason I commented on it following the real time tests we did. The fan was not active during our tests. I expect that in our summer conditions there will be a difference but until we can do a real time capture in summer I wont jump to conclusions

As for lack of Dealer Diagnostics on these models, until owners unite and bash heads at BMW/Husky/KTM for commonsense to prevail so BMW dealers are authorised to use the BMW Dealer Diags Systems for programming/diagnostics you will always be in a bad position. It is merely permissions in the Online System and some petty bureaucracy which is preventing it.

Enough said, let us not hijack this thread from discussing the Injector/Plugs

 

[engineerk9]

Interesting information, but I find the above regarding the AIT sensor difficult to understand. OK it was winter but you did not give any actual temperature readings or info.

The TR650 OEM sensor siting is:

a) directly above the engine and exhaust;
b) in the flow of the cooling fan

I can confirm that in heavy traffic on a hot day the readings are far from correct or stable and are very much influenced by the engine heat.

Stationary in traffic on a hot day my AIT sensor suffered from the engine heat and fan airflow to give readings as high as 20c over and above actual. This is fact based upon measurement of actual air temperature against dash reading. ie 53c shown as opposed to 33c it should read.

With the dash reading high was when the stalling issues were at their worst and the bike was crap to ride in traffic.

Re-siting the sensor first to the front of the bike gave much more stable and realistic readings, however I felt there was room for more accuracy. That is why I placed my sensor in the right front beak at the air intake so it is measuring the temperature of the air entering the airbox. The new position means that the sensor is no longer affected by exhaust and fan air and reads much nearer to the real air temperature, particularly in traffic.

Many other TR650 owners have done the same and have gained similar improvements in the way the bike runs on a hot day.

Your post implies that there is no issue with the OEM siting of the AIT Sensor, when in fact for most of us there certainly is.

Hi Mark

I agree - my AIT is now located in the RH beak as far forward as is possible - I also use a scientific remote temperature senser to monitor actual air temperature

Even with the latest firmware the ecu will always be chasing the 14.7 air fuel ratio, whish for most of us appears too lean fuelling

 

[Roger 04 RT]

One of the reasons to have the temperature sensor in the airbox is that you want it to capture the temperature of the air entering the engine, not the temperature of the air entering the intake.

There can be a big difference. For instance, if you're riding along and the actual air temperature is 70F, but you stop in traffic for a minute or two for a red light, the engine often warms the air to 100-110F. You want the sensor measuring that temperature, not actual air temp.

 

[mag00]

So what happened to the "learned out" theory?

 

[Roger 04 RT]

So what happened to the "learned out" theory?

Mixture adaptation isn't a theory, it is they way modern ECUs work.

However, it isn't the air temperature sensor that gets "learned out", it is errors and wear that get learned and then corrected. Even a deliberately induced error (like an IAT shifter) gets corrected by mixture adaptation.

 

[mag00]

Just remove the AIT and problem solved. Not needed at least from what I hear you saying.

 

[hasenpfeffer]

That's not what he said at all.

My understanding is the AIT is used in conjunction with other sensors and adaptations to help reference a particular fuel map point. I don't believe this fact ever changes (except maybe during fault conditions).

What changes is during closed loop operation, the O2/lambda sensor can be used as a feedback mechanism to figure out how well the fuel map references are working. In a simplistic example, say AIT_X, TPS_X cause FUELMAP_X. This choice causes O2_X and thus LAMBDA_X. Well, maybe it turns out that LAMBDA_X is not ideal, for whatever reason, doesn't matter. To fix the problem, the ECU can throw in an adaptation correction.

So, the ECU is actually taking AIT_X, TPS_X, RPM_X, MAP_X, possibly others, and modifying with ADAPTATION_X, which causes FUELMAP_Y. Hopefully now the choice results in O2_Y and thus LAMBDA_Y, which is ideal.

So, the AIT is always used as an input, but the adaptation mechanism can correct for consistently incorrect readings.

 

[mag00]

So if the consistently incorrect reading of NO AIT, will be adapted for, right?

 

[hasenpfeffer]

No AIT could easily be a special fault condition which is likely to force a different running mechanism. I could also imagine that the adaptation mechanism(s) have limits to them, so I'm sure if you push inputs past the engineering design assumptions then "all bets may be off", so to speak, and the theory of operation could no longer apply.

I also edited my post above a little based on feedback.

 

[mag00]

If the ecu is going to disregard the AIT based on o2 output, it is not needed. It's just that simple.

Maybe my logic is simplistic, however, it is based on the statements of the ait spoofs getting "learned out".

 

[hasenpfeffer]

It's not disregarding it. It's using adaptations to compensate for a non-ideal lambda. Think of it more in terms of what one of the adaptation's real purpose is. To compensate for variance between sensors. For example, a TPS sensor may be mounted or manufactured slightly differently from sensor to sensor or from bike to bike. So, a particular TPS reading on 1 bike does not likely represent the exact same position on another bike. To help compensate for this, the ECU uses adaptations to shift which fuel map points it picks for particular sensor readings. The AIT spoofing simply takes this concept to an extreme, where the variance in the AIT reading is really big compared to nominal.

 

[hasenpfeffer]

Another example:

if the fuel map was designed to expect point X to be used when AIT was 20 deg C, then that point may be expected to produce an ideal lambda (or at least an expected lambda). But say the air temp was actually 21 deg C simply because that particular AIT sensor is a little off calibration from manufacture. Then the AIT reading and fuel map selection might be slightly off, producing a non-ideal lambda feedback. The ECU adaptations allow for automatic compensation. The AIT spoofer scenario is simply a more extreme condition of this.

It's not disregarding the input, just adjusting to the new nominal.

 

[mag00]

So then are you saying the spoof works?

 

[hasenpfeffer]

Don't know. Just trying to explain my understanding of the ECU's theory of operation. That theory would suggest that an AIT spoofer could get adapted out. I'm well aware of the empirical evidence that suggests otherwise and do not understand the discrepancy.