Project Electronic Fuel Injection

[zombie]

The O2 sensor will change everything. I didn't realize you had it out. It will never run right without it. Like taking the float out of a carb.

[baddi]

The O2 sensor will change everything. I didn't realize you had it out. It will never run right without it. Like taking the float out of a carb.

I didn't have a wealding machiene, so so far it have just been running open-loop with the standart calibrations.

[zombie]

That will change everything for you.

[baddi]

Hey guys, i am an unlucky idiot once again.

Here the other day, i got a piece for the injection, which i had broken in the mail. (not enough clearance between saddle compartment and injector, so a piece of plastic broke).
I installed the EFI kit along with the O2-sensor, but it kept reading lean, no matter how much fuel i added, even to the point where it couldn't rev properly, which was double the fuel than standart calibration, so this confirmed my earlier fear of an air leak. All the times i've checked for air leaks before by putting my finger in the exhaust and and turning the crankshaft by hand or spraying air at the engine while running, but the engine felt like always and didn't react to the water, so i pulled myself together and opened the engine for the first time ever and saw what is on the pictures.

When i was removing the bottom gasket, it started raining and i had to get the scooter into the garage. I then saw that i was going to be late for work, and quickly assembled the engine, forgetting to remove the last of the gasket, so i drove 200 meters before it had thrown all of the oil onto the ground and the engine stopped.  

Fortunately i've got another cylinder + piston, which has only driven 70 km, so tomorrow i'm going to change to that, and hopefully i will experience a better and more stably running engine.

The piston/cylinder might have been broken when i had only driven a couple of hundred kilometers, as an idiotic salesman sold me 15W40 oil in the winter instead of 10W40, and i had to drive it like that for a whole month due to lack of money. I have also driven with the standart #80 jet, as when i changed to #82 after driving a month or like that(cant remember when i did it, but it was cold as hell) it wouldn't run very well, so the engine might have been ruined back then already. :/

Anyway, i'm going to fix it tomorrow, and when it runs well, i will try the 7g dr. pulley i've just gotten in the mail and i will see what happens.


As an effect of me ruining an engine, i will take back all i've ever said about small jets, as i have possibly also ruined my first agility, when i were driving an open air filter with the #84 jet, even though it was the jet it would run with the best back then, so now the only active statement about jets and open filters are that you should use 100+ jets, as you then won't burn the engine!

[streido]

Bummer. Just when you were getting there. At least you got some spare parts already to replace them, hopefully that works.

I was surprised when you said before about the 84 jet on an open filter, that def seemed way too small. Mine is running good on the 92, seems almost perfect. Think a lot comes down to the filter type/size, your altitude, the humidity and temp when it comes to the jetting on these open filters, thats why no 2 folk on here use the same size.

[baddi]

Bummer. Just when you were getting there. At least you got some spare parts already to replace them, hopefully that works.

I was surprised when you said before about the 84 jet on an open filter, that def seemed way too small. Mine is running good on the 92, seems almost perfect. Think a lot comes down to the filter type/size, your altitude, the humidity and temp when it comes to the jetting on these open filters, thats why no 2 folk on here use the same size.

Yeah, but i've never used anything but standart filter on this engine, so the engine being like this must be because of some oil, fuel or cooling problem, but there are one thing i have to get straight now before it hurts another cylinder/piston:
When i tap all the fuel off so it only drips once every 2 secs, i can only fill 0,6L oil on it, before the dipstick shows full. Should i fill the 0,7L oil the manual say or the 0,6L the dipstick say?

[streido]

I never need the 0.7l of oil either, i use about the same as you did. I put a mark on my oil filer can so i just use that now, that takes it up full. Remember too much oil is can bejust as bad as too less oil.

[baddi]

I never need the 0.7l of oil either, i use about the same as you did. I put a mark on my oil filer can so i just use that now, that takes it up full. Remember too much oil is can bejust as bad as too less oil.

Yeah i know, the pressure will be too high and the gasket/sealing between the two halfes of the engine will work its way out over time.

But my thought was, that less oil will be able to absorb less energy, and the engine will be hotter. Just to narrow down what could have given me the cylinder/piston damage that i've got now, because if it is too thick oil at too low temp when it was new, i am going to complain to the people who sold me the oil and told me that that oil was what i needed, and hopefully get something for free or on discount or whatever. I just have to be sure what possible causes to the engine damage could be, because the first agility i had have been driven through 2 very tough and much colder winters than this with the standart #80 jet without being damaged when owned by previous owner. :p

[LoveMyKymco]

I would say an air leak is seriously the problem, or your running very lean(usually caused by air leaks), I would also assume the score marks near the skirt on the piston are from seizure when you lost your oil.
I would not think 15 weight would cause any of these scenarios. Did it have problems starting? It would have low compression I would think since so much blow by.

[zombie]

The blow by is what caught my eye. The rings were never set before it was run at wot. That will prevent the rings from ever seating... I ran about 250 miles on the 87cc kit before ever opening it up. The kit now has about 1000 miles, and you can still see most of the cross hatch, and there is no apparent blow by.
The oil is not the culprit. Where did you find the air leak? I would also invert the head, and fill the combustion chamber with Brake clean to be sure your valves/seats are in good shape. The leak MUST be in the intake tract. There is no where else that could induce air.

[baddi]

The blow by is what caught my eye. The rings were never set before it was run at wot. That will prevent the rings from ever seating... I ran about 250 miles on the 87cc kit before ever opening it up. The kit now has about 1000 miles, and you can still see most of the cross hatch, and there is no apparent blow by.
The oil is not the culprit. Where did you find the air leak? I would also invert the head, and fill the combustion chamber with Brake clean to be sure your valves/seats are in good shape. The leak MUST be in the intake tract. There is no where else that could induce air.

I've changed cylinder/piston and gaskets, but the problem keeps on. I've checked for air leaks between the throttle and the engine over and over again, but never found any.

The RPM are constantly high, but today i found a way to keep them down. If i cut the fuel to half the standart calibration and screw the idle screw out, the idle will be at 1200-1600, but the O2-sensor show lean mixture. If i then screw the idle screw in a little, the rpm will slowly rise until hitting some critical point, where it rises uncontrollably to 4-5000 rpm and even with the same amount of fuel, the O2-sensor now shows rich. It is like this all the way, that no matter how little fuel it gets, it shows rich when the rpm are high.

I am going to buy new piston rings next week, when i get my paycheck, and then it will hopefully work.

I would say an air leak is seriously the problem, or your running very lean(usually caused by air leaks), I would also assume the score marks near the skirt on the piston are from seizure when you lost your oil.
I would not think 15 weight would cause any of these scenarios. Did it have problems starting? It would have low compression I would think since so much blow by.

The pictures are from before the oil loss. I lost oil because i stressed and f***ed up, when putting it back together, because i was late for work and the oil loss is fixed when i installed new piston/cylinder and gaskets.

[zombie]

Step by step this... What is the idle screw? Is it the primary air break adjustment for the butterfly? ... Think I have the issue... The set is callibrated for a 125. The butterfly needs to be EXACT. I think your leak is the butterfly. There is an air hole in it right? Try using some TINY rubber grommets to restrict that hole. Then you may be able to compensate with the fuel pressure/map. If that gets you in the right direction have the hole welded/brazed, and drill it to the appropriate size.

Also just to be sure try the o ring around the injector. I have cut a few in the past when seating them.

[LoveMyKymco]

by cutting fuel flow you are changing its pulse width, eventually the injector will jam by doing this. too little or too high of fuel pressure can jam pintle style injectors. The ECU will still try to cycle it the same regardless of pressure, it is the pulse length (the width) that you are changing. It is no longer spraying a fine mist but more of a dribble. I would definitely recommend against doing this again. You are not having a fuel pressure issue. Your O2 should show lean condition when it is idleing, it should hunt between 12-14 AFR then when under a load (moving) it will balance itself to stoich unless it needs calibration.

1) You may have an injector issue from the start
2) May be bad ECU
3) May be bad user input interface with ecu
4) May be bad fuel pump
5) May be compression issues and the only way for the ECU to think its ok is at a high rpm (very unlikely)
6) Still air leak probable

What I see as an outsider who hasnt seen it run is that it runs but rich at 4-5k rpms, so I would start adjusting your fuel table (VE) at those rpms. And go back and check to make sure the ecu retains those values you set.
do not use the fuel pressure as a means to lower rpm. The injector will get ruined.
a 4-5k rpm would mean a massive air leak somewhere if it were an air leak. I would replace every O ring and gasket on it, not just inspect it. This is an engine that can be ruined by failure to tune properly. But honestly for that much rpm difference I would think the leak would have been found by now, who knows maybe zombies on to something with the air bypass on the flap. Most kits dont have an air bypass hole though and thats why they have a screw/bleeder style bypass.
Any way it goes it still comes back to the ecu not performing as it should. There should be a self learn cycle on it, if not I would be concerned of the quality of the program. Since you have had this issue from the start it would be safe to assume there could be a hardware or software issue, if all mechanic parts and seals check out. It is not an intermittent issue so it HAS to be one of the constants.
Have you contacted the company that produced the ECU? I would try to directly contact them or the company that provided the software.

If you can upload the software program to some upload site i can take a look at it. Or if theres a site for the DL already link me to it.

[LoveMyKymco]

If you are the first to buy this kit and install it on a 49cc then you may just be in for a long head ache, but dont worry it will get figured out.
can your ECU tell you voltages of the MAP sensor and the TPS? I would have tried to adjust the tps first really now that I think about it. lol. But either way see if on the data logging part of the program what the voltages are and cross reference them to what you know they should be. IE. if throttle at full closed the TPS should show 1 or 5 volts depending on style but it should not read 2 or 4 volts, you know what I mean?
If you can run a data log session and post it on here it may help us. In your video it did not show much other than the movements, I need real value of what is what and what it is doing. The real time graph sucks compared to numerical reading IMHO, so if you can show that rather than the graph it would be best. Make sure to label or tell us EXACTLY what reading we are seeing and what is the max and minimum reading value possible, what it should be based off of mathematical logic. Sorry I cant explain it better, i tried 10 different ways and ended up erasing them all lol.

I need these values. Actual reading (real time) what it reads when idling, minimum and maximum for what it could be 1-5v 5-1v 1-12v etc., what it should be, not what you are seeing, for the following
 fuel pressure reading if possible
 timing values at 1k rpm and at 4k rpm these are in the data log
 MAP sensor values (bar #)
 TPS values (they should stay constant and not fluctuate as you are NOT touching the throttle)
 VE table %s at 600-6000 if you can
 A/F ratio at idle give detail of full fluctuation in real values not description ie. at 1080 rpm it is 12.7 fluctuated to 13.1 or if it stays at 13.1 most and fluctuates down instead of up At 4k rpm it is at 15.1 going down to 13.1 or vice versa.
 make sure the lambda is running in its loop also. I saw it data logged for you but make sure its not just operating as a data log only.
Once again sorry just not sure how to explain this all online.

This should be done at an idle only and absolutely no touching the throttle, we will figure out what your idle issue is, and go about it this way, I would rather work down than up but gotta do this first this time to rule out any issues.

[baddi]

Zombie: The idle screw is a screw that bypasses air so it goes past the closed throttle.

LMK:
- The Lambda-sensor is a narrowband, so i cannot see the AFR, but only if it is lean, rich or just right.
- The fuel pressure is generated by the fuel pump and kept constantly by a fuel pressure regulator, which lets fuel go back into the fuel tank when the pressure rises, so i am not touching this. I am only manipulating the pulse duration.

If you have any uploading site (damnit that megaupload were closed) i can upload the tuning guides and the program together with logged data, so you can play it back. Both voltages and what they mean are logged but can only be shown by the analyser program.

There are a self-learning part of the program always active, but i cant really see it doing anything and i suspect that it will only learn and calibrate for small changes, which i suspect that this isnt. :/