Jump to content

is it Reset of the ECU ?


Guest MotiAlfia

Recommended Posts

I think Steve meant a different FI system from the FI system described in the article.

Link to post

Basic principles will be more or less the same, but 15yrs is several generations on in ECU/engine management terms, just like PCs, TVs and mobile phones.

  • Like 1
Link to post

The original post referred to the fact he has changed his exhaust system - does this have a bearing on matters?

 

Link to post
wingrider.steve
On 12/19/2017 at 01:02, SteveThackery said:

What nails it for me is this: there is absolutely no mention of it AT ALL in the official service manual.  Trust me, if there were anything in it, it would DEFINITELY be in the service manual.  There's no way Honda would leave out such an important procedure.

 

....But it is normally in the setup instructions that Honda dealers get with bikes fitted with the PGM-Fi system in order to ensure that base settings are established for the particular location (i.e. altitude and ambient temp) where the bike will be first used.

The ones I've seen are similar but simpler to that in the original post, basically start the engine with out touching the throttle ((and do NOT touch the throttle during the procedure) and let it idle until 90 secs after the engine is up to normal running temperature then switch off the ignition.

Doing this procedure at any time CAN help sometimes if for whatever reason the PGM-Fi system has lost or needs the base settings re-establishing, for instance on trips to the Alps my first 1800 Wing would benefit from but doing this procedure when we reached the Alps let it adjust to the high altitude. I've found the same thing with a 700 Deauville.

Honda's PGM-Fi system WILL gradually adjust itself in these circumstances but not necessarily quickly enough at times.

Like Steve I'm not sure why it is not in the Honda service manuals but it does work.

Edited by wingrider.steve
Link to post
SteveThackery
14 hours ago, Andy m said:

Every approved ECU uses EEPROM. It may be just a way of avoiding testing tests, but they do.

 

I'm happy to stand corrected.  I thought I'd read somewhere that a lot of modern microcontrollers are actually implementing the EEPROM function with flash memory, on the grounds that it simplifies the die design bearing in mind there's a load of flash memory already in there for the program memory.

 

But I can't remember where I read it so I'm happy to admit that I'm probably wrong!  :)

Link to post
SteveThackery
10 hours ago, djsb said:

So why does it have a section called PGM-FI SYSTEM in the workshop manual (page 4-1 onwards)? Granted it's probably been improved a lot since that article was produced as well as being integrated with the DCT system where appropriate.

 

Because "PGM-FI" is simply a Honda brand name that they've applied to all their fuel-injected bikes for the past 15 years.  It doesn't mean anything in terms of the actual technology used.

Link to post

It stands for ‘ProGraMmed Fuel Injection. Honda use that name for all their products that don’t have carburettors. Cars, motorcycles, mowers, agricultural machinery. 

  • Like 1
Link to post
SteveThackery
29 minutes ago, wingrider.steve said:

....But it is normally in the setup instructions that Honda dealers get with bikes fitted with the PGM-Fi system in order to ensure that base settings are established for the particular location (i.e. altitude and ambient temp) where the bike will be first used.

 

OK, so can someone explain this to me, rather than just making assertions?  My knowledge is far from complete, and I am genuinely interested in learning.

 

So what exactly are these "base settings" that everyone keeps mentioning?

 

I've mapped an ECU from empty before, and it involves building two fuelling base maps.  One uses the MAP sensor (manifold absolute pressure) and engine speed (for light loads and cruising); the other uses the throttle position sensor (TPS) and engine speed for the higher performance ranges.  The ECU switches between the two maps depending on a complicated algorithm based upon the speed of throttle opening, etc.  Anyway, each location in the map has a fuel injection pulse duration value, which is arrived at by estimation first to get the engine started, then by dynamometer and exhaust gas analysis.  A lot of work goes into fine tuning the maps to give the best combination of economy and performance.  Finally you "plug in" the lambda control loop regime, which is complicated but essentially uses the lambda sensor output to "pull" the mixture away from the mapped value towards stoichiometric (assuming it's not already stoichiometric from the map).

 

Now, for an FI system to work perfectly it needs to know the mass of the air flowing into the cylinder.  Unfortunately mass air flow sensors (MAF) are usually too bulky to package well in a motorcycle (and they work better when feeding multiple cylinders, hence not experiencing a "pulsey" air flow).  That is why motorcycles usually use the MAP/rpm combination or the TPS/rpm combination to estimate the air mass, rather than measure it directly.

 

And this leads to a particular requirement: to make a good job of estimating the air mass, it helps a lot to know the barometric pressure.  According to my Advanced Powertrain lecturer, a standard technique in motorcycle ECUs is to use the inlet manifold pressure sensor (the MAP) to measure the ambient barometric pressure when the ignition is first switched on, prior to cranking over the motor (so the inlet is at atmospheric pressure).  

 

What else might be "variable"?  Well, the voltage span from the analogue sensors (such as engine temperature and throttle position).  I can't think of anything else.

 

Back to the question: what are these "base settings" people keep mentioning?

 

I can think of only one, and that is what the TPS reads when the throttle is closed.  I would think it trivial to discern that without a special warm-up regime, but nevertheless an enforced idle period could be used to determine the TPS reading for a closed throttle.  The TPS reading for fully opened isn't checked, of course, during a warm-up idle.

 

Perhaps the span of the temperature sensor could be measured?  But in that case the ECU would need to know the "real" temperature and it can only get that by using another temperature sensor.

 

Apart from the lambda sensor, the engine has no built-in dynamometer nor a built-in exhaust gas analyser, so it has no data available to reprogram its own fuelling base map.

 

I'm out of ideas.  Can anyone give a convincing story of what, exactly, is going on when an ECU is allegedly establishing its "base settings"?

 

29 minutes ago, wingrider.steve said:

Like Steve I'm not sure why it is not in the Honda service manuals but it does work.

 

How do you know it works?  (Not a snidey comment - a genuine question.) 

 

Link to post
wingrider.steve
19 minutes ago, SteveThackery said:

How do you know it works?  (Not a snidey comment - a genuine question.) 

This comment related to my experience with 1800 Wings and 700 Deauville sometimes they can idle and run less smoothly than normally (when I've been at high altitudes particularly) which can be cured by doing this "reset" procedure. The Honda Fi system will "learn" and compensate for this itself but in my experience its not always as quick as maybe it should be.

Also if an 1800 Wing at any time develops a slightly uneven tickover it can be cured by doing this procedure. I'm not sure why it works but it does and as I said it is in the dealer setup instructions of which I still have the ones for the '05 Wing I used to own.

 

Edited by wingrider.steve
Link to post
SteveThackery

Thanks, Steve - very interesting.

 

I'm hoping between us we can cast some light on what's actually going on.  :)

 

Also, tomorrow morning I'll try the cold start warm-up thing and will report back.

 

Edited by SteveThackery
Link to post

Car ECUs have a barometric sensor in them usually for altitude compensation, don't know about bike systems. I would nave thought they'd use a baro sensor in the ECU by now. In theory it could use the pressure sensor in the throttle body at ignition on but before crank to establish an atmospheric pressure at start but as WRSteve says it won't then know when you've gone up a mountain unless you stop the engine and restart. The idle speed function should still work OK but the idle quality might suffer (mass airflow varies with pressure difference between intake/exhaust, exh is effectively atmosphere at idle so fuelling changes a lot). The feedback should cover it but it'll be offset. Bear in  mind the effect of altitude is quite dramatic, going 2000m up a mountain reduces density by about 20%, it takes a temp difference of around 100degC to give that sort of effect.

air-altitude-density.png

air_temperature_density.png

 

If doing a set procedure from cold to normal full warm it can (in theory) use the idle speed controller position vs idle speed to establish where it should be for a cold start crank (e.g. so many steps further open than hot idle on a stepper motor). It can also determine whether the closed loop feedback is centralised (or offset if so intended) and adapt accordingly. Temp thermistors are usually pretty precise and accurate, certainly accurate enough (working with absolute temperature for density determination, +/- a couple of degC is fine) which is why an AIT sensor in the airbox is perfectly adequate.

If someone (Steve?) is sufficiently enthusiastic you could record the feedback (EGO sensor output) at hot idle, do the procedure and record it again, see if there's any difference. Injector pulse width might show it too, depending on the ramps used. In theory the average pulse width wouldn't change (all other things being equal) but the lean/rich swings might be different.

Link to post
SteveThackery
8 hours ago, embee said:

Car ECUs have a barometric sensor in them usually for altitude compensation, don't know about bike systems. I would nave thought they'd use a baro sensor in the ECU by now.

 

Actually that makes far more sense than what my lecturer said - baro sensor chips are dirt cheap and would work all the time, not just when the engine isn't running.

 

8 hours ago, embee said:

If someone (Steve?) is sufficiently enthusiastic you could record the feedback (EGO sensor output) at hot idle, do the procedure and record it again, see if there's any difference. Injector pulse width might show it too, depending on the ramps used. In theory the average pulse width wouldn't change (all other things being equal) but the lean/rich swings might be different.

 

Dammit - you may not know this but I developed a comprehensive data logger for my Enfield as my final year project at uni.  It monitored and recorded loads of stuff, including the lambda sensor and the FI injection pulses, engine temperature, etc.  Unfortunately I've taken it off the bike as I'm planning on selling it and getting a new one next year, otherwise we could have tried this cold-start thing on the Enfield.

 

I could develop a new one aimed at the NC, but I didn't really want to mess with it as I'm thinking of selling that, too.

  • Like 1
Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...