EFI for 50cc-250cc Scooter! (UPDATES!!)

[mainepeace]

Bench testing the EFI unit.

Attachments:

[90GTVert]

Just a note to anyone else looking into this stuff, read reviews on the USB to serial adapter if at all possible. I'm not using it on a scooter, but I was having a hard time with a cheap no name USB to serial adapter hooking up a Megasquirt in my Mustang to my laptop with Win7. It's not supposed to be as big of an issue on other operating systems, but if you have Windows 7 for sure, don't just go and buy the cheapest USB to serial adapter you see on eBay. It's not saving you any money when you have to buy one that works later.


Also, and I see this is an old point from page 1 but I had to comment, raising the idle does make a nice difference in charging if you have any problems. I installed a stereo on one of my scoots that draws up to about 1.5A as I use it. I raised the idle until my headlight was no longer dim, which was about 2000-2100RPM on my 2T and the battery stays charged very well.

[mainepeace]

Fuel return line nipple.

Attachments:

[mainepeace]

Need to make sure fuel line clears the seat bucket attachment points.

Attachments:

[mainepeace]

See the huge difference in the diameter of the intake port and the spacer for the new intake plenum. The plenum is a little bit smaller. Ideally there could be a port match between the two, but I'm interested in evaluating the kit as is and am trying to abstain from making airflow mods like that. Maybe I'll do that later if the kit works out well.

Attachments:

[mainepeace]

Looking down into the exhaust header at the placement of the oxygen sensor. The exhaust is almost too small for the sensor to fit. I backed out the sensor as much as I could to try to keep airflow up. I would guess that a larger exhaust pipe is in the future plans. For the moment the restriction could actually be a good thing. It would keep up the back pressure on the exhaust to keep up V.E. (volumetric efficiency) since the throttle body and intake plenum are so huge. They would tend to slow down the intake charge which could actually hurt overall power.

Attachments:

[mainepeace]

Funny you should mention the wiring harness pinout. Matt just send the the schematics this morning.

I got some more info from him so now I hope all issues are resolved. More coming soon.

Greg

Attachments:

[D-cat]

Do you mean less wattage? I imagine the supply voltage will remain +12.

Anyway, yeah, +12v is straight black. As someone who is color deficient something like that could easily bite me too. At least you caught it. Looking forward to the running review.

[mainepeace]

Yes, I mean it uses less wattage power and should be fine at idle. Now the pump should be fine over 2k rpms or so.

Greg

[mainepeace]

With the foam filter it is easy to just poke whatever you want through the foam. This is the crankcase ventilation and the IAT sensor.

Attachments:

[mainepeace]

Here is the Performance switch and the Error LED. I had planned on having the LED in the switch light up when Performance mode is activated, but it's a 3 wire connector on the back with a shared ground. The switch is 5v so it won't work. I might get another switch that has 4 wires so both circuits are isolated.

Attachments:

[mainepeace]

More photos of the install.

Attachments:

[mainepeace]

.

Attachments:

[mainepeace]

My DIY manual petcock.

Attachments:

[tvnacman]

seems like the ecu needs to learn , just a thought could the air filter be affecting the MAP and inturn causeing a too rich or lean .

[justbuggin]

cool i want to get this system for my 150 anjd see if it will wake it up in the pwoer band as i would like to climb hills and not lose speed at the same time

[mainepeace]

I got the new manual petcock and throttle cable today. I installed the throttle cable and it fit perfectly... maybe it could be about 1/4" shorter on the inner cable (and overall 2-6" shorter), but with adjusting the three adjustment points I was able to get it situated properly.

One thing about the new throttle body is that the return spring for the throttle blade is much stronger than on the carburetors. It takes more effort to turn the throttle. Not a big deal, just something to get used to, and make sure your hands aren't slippery from hand creme or the like.

Tried to start the scooter up and it kept dying. It would idle fine but would die when touching the throttle even 1 degree. After a few minutes of idling, it would throttle up fine. I guess the "cold" fuel tables need to be modified a bit.

The scooter ran well, the engine strong. It did keep on bucking occasionally, especially after running for a while. It's definitely the clutch that needs to be replaced. I'm probably going to get a performance clutch from Scrappy.

This morning Matt from Ecotrons sent me another updated settings file to upload into the ECU. He really seems on the ball in helping me get this thing running well.

Once I get everything all tuned up and the settings perfect, I'll make these available to anyone who purchased a system.

I spent about 4 hours today putting all the body panels back together and detailing every crack and crevice on my scooter. It glows now! I also went to Radio Shack and bought a new switch that lights up Blue when the lights are on. Kinda cool.

Greg

[noday]

sounds like your O2 sensor took a while to come up to a functional temp to send a signal back to the ECU.

idle is open loop until the O2 sensor heats up

in open loop the ECU does not know how much fuel to inject to to satisfy an increased throttle .

new tables will not compensate well for no after fuel burn info from the O2 sensor.

the problem yesterday with lag at the stop sign sounds like the same thing.

this system appears to require fuel burn to have the O2 sensor heat enough to come online.

I expect a significant part of the problem is the volumetric disparity between intake and exhaust, along with different materials ?

small diameter exhaust with steel will heat up slowly. hot exhaust gas is ejected rather than retained to heat the sensor..

once the O2 sensor is hot enough to send a valid signal, the ECU responds to the throttle signal.

you had constant fuel with a large air surge when the throttle was advanced.

leaned to death.

do you know the O2 sensor type ?

[mainepeace]

Hmm, interesting, because in tests before sometimes I didn't have an issue, but then again I don't remember if those times the engine was cold.

It comes and goes which is why I think it's part of the fuel map or calibration. I'm still working on an issue with a code being thrown for the MAP sensor, so it's probably related.

When it died at the stop sign the engine wasn't cold. I had ridden for 10 miles then stopped for about 3 minutes before getting back on the scooter.

The MAP sensor functions like the MAF, and since they provided the TB they know it's flow characteristics. It *should* be able to provide a somewhat accurate open loop fuel map. In fact, they sell the kit without the oxygen sensor too, using only the programmed fuel maps and temp sensors and MAP sensor.

AFAIK, this is the first Chinese made 50 (63) scooter to have this particular kit. I'm dealing closely with Matt, the owner of the company. The original config files are somewhat "generic" that probably originated from their mathematical model and test bed rather than a 139QMB engine.

I'm not sure what oxygen sensor it is... all I know is it's a common one for small engines and it's not a wideband.

I'll know more when I do another cold start with the new config file. This time I'll record the session so I can analyze it.

Greg

[D-cat]

I'm thinking hide a netbook under the seat with a screen-cap app running so you can play back later and see how the numbers change as you are riding. It might help you come up with better cold numbers.

I was also thinking about the placement of the MAP sensor. Is it literally just hanging in the flow? I'm thinking that it might work better if it can actually be secured to something right at the inlet opening, even a paperclip might be enough.

[noday]

why I asked about the sensor.

1st generation have 1 or 2 wires & depended on exhaust gas alone heat to heat the sensor to a functional level.

2nd gen have 4 wires & and a heater built into the sensor. they come online faster. the heater circuit is on when the key is on, kind of like a glow plug.

why ?

1st gens have longer times to come up to operational temp.

if you have a 1st gen sensor, part of this "bug" might be easily corrected by using a 2nd gen sensor.

for about a month I have been searching and reading to build a cheap a/f mix monitor to use in tuning a scooter.

the cheapest wide band o2 monitor is about $180 discounted price.

can put a narrow band gauge monitor together for about $40-$50 Plus the bung.

1 drawback is, it is only good for steady state & only says rich or lean unless you are very, very close to the ideal a/f mix ratio. no real help in how much to change carb settings or jets. only tells you that the mix is off and which way.

2nd is how to do a temporary o2 sensor mount in the exhaust stream.

currently I am thinking of a 1" long manifold spacer with the sensor in place that can be mounted between the exhaust port and manifold.

with no dyno, carb settings must be road tested at multiple rpms.

any way, I applaud your efforts and look forward to more of your reports.

Dan

[mainepeace]

This oxygen sensor has 4 wires. I've been told it takes 2-3 minutes to warm up. What you say makes perfect sense. From my work on supercharged Mustangs, it's always been a "rule" to never place the oxygen sensor too close to the exhaust port, as the turbulence will screw up the reading. It's recommended at least 6" down the header to put the oxygen sensor. For the scooters, I'm not sure if it matters that much, but that was my thinking when I put the sensor in the first bend, a good 4 inches from the port, and in a place that is somewhat protected from getting banged.

I flashed the ECU with the new code I got from Matt. Then I loaded up the new profile settings. It took 4-5 minutes to do all this, with the ignition key on, so I would guess the oxygen sensor would have time to warm up.

Anyway, when I started it it started right up. No issues with touching the throttle or even gunning it. Absolutely no bogging anywhere. It rode like it came from the factory this way. Everything seems great. No codes whatsoever. Sometimes it looked like a code flashed on the light, but it ended up being glare from the beautiful sunset. I've sent the data log over to be analyzed but it seems we have a winner now!

Throttle is crisp and immediate. Engine is strong at every RPM and every throttle position. Compared to another fully stock 50 that I just got working, the difference in acceleration is noticeable. I even almost popped up the front tire from a stop sign. Also, I noticed that my 2009 Peace Sports 50 is MUCH heavier than the 2003 Tank 50 I have. Not just a few pounds but more like 20-30. It's also almost 2" taller.

Seat of the pants tells me that the overall power gain is as much as going to a free flowing filter. I can't give MPH numbers, as those are very dependent on individual and situational factors. Also, acceleration and top speed are very dependent upon the drive train itself. I will say that the spots where the scooter would tend to bog before, like slowing down for a sharp curve, later into a hill when the RPM's are going down, it wouldn't. I'm going to analyze my a/f readings and see if there is anything that can be tuned. But what's so nice is that I can tune for a specific RPM and a specific throttle position and don't have to get my fingers all wet with gasoline to do it. That's the main reason for this upgrade, and not because it gives an immediate power boost. It does let the engine give you the most power it possibly can for whatever mods you put on.

I filled up the tank with 91 (at $4.19 per gallon!! so at least it was only $4.30). The odometer read 733.1 miles. When I fill it back up, at the same pump, I'll read back the mileage.

My very first tank fillup I got 110 mpg if I remember correctly. After doing some airbox mods and re-jetting and ONLY riding at WOT, it went down to ~70mpg. I don't log my mpg anymore so I'm going from memory. We'll see how it runs and what sort of mileage it's getting.

I've been writing a novel here and have only gotten a few comments. I'd like to know that this information is at least helping someone.

Greg

[mainepeace]

May 13, 2011 16:58:11 GMT -5 D-cat said:

I'm thinking hide a netbook under the seat with a screen-cap app running so you can play back later and see how the numbers change as you are riding. It might help you come up with better cold numbers.

I was also thinking about the placement of the MAP sensor.

I ended up putting my netbook in the rear luggage compartment. I have the serial connector attached next to the fuel pump, so that it is accessible under the side fender. I'll eventually make a cap to keep out the dirt. I think I have a cap that came on a VGA connector that should work. I wrapped my netbook in a big towel to try to keep down the vibrations. It has a physical hard drive which is why I'm worried about the vibrations under the seat. When recording, the ProCAL software does write to the hard drive so that needs to be turned on. Possibly, the software could be configured to run from a USB stick, or even booted off one using Linux or winboot. That's if you wanted to configure your netbook to be somewhat dedicated to doing tuning. Big issue is that you need to manually hit "connect", then "record". Of course you could setup a macro using winbatch or something to do it.

The MAP sensor is attached to the vacuum tubes. As you can see on the attached image there is only one vacuum attachment. The MAP isn't in the airflow like a MAF would be. I have as short a vacuum tube attached to it as I can have and still be able to mount the MAP to the frame where it won't jostle around. I have the emissions junk still attached as I like a cleaner burning engine. With the new settings file the MAP works fine now. After I get through this first tank of gas I'm going to do some mods like putting JB Weld on the intake to port match the runner to the intake manifold. Right now there is almost a 1/2" discrepancy in the two. That definitely makes for a power stopper there. Also, I plan on making a velocity stack for the intake and extend the runner at least 6" if I can. I'm sure there's lots of power to be gained from that. And last I am going to replace the exhaust with something a bit bigger.

Greg

Attachments:

[mainepeace]

You can see the placement of the vacuum tubes and the MAP in this pic.

Greg

Attachments:

[mainepeace]

I got some more settings from Matt from Ecotrons to change. The top end is better. Basically, I adjusted more fuel for WOT (or thereabouts as my WOT = 82% of the TPS voltage). From a stop the acceleration is much better. With more tweaking the power keeps coming. Not that it's an enormous amount, but it's noticeable.

It took 3 tries to start from dead cold. But each try was no longer than 1 second. I don't like to pump the starter. It's possible that even with the fuel pump turning on for 6 seconds, the fuel system might not be pressurized enough. Or it could be something entirely different. After the start there was no issue so I'm not concerned at the moment.

I can't seem to get the data analyzer to run so I can't go over my data logs. Well I can, in Excel. However, the numbers have multipliers and offsets so it would take a bit of programming to get a meaningful graph.

If I can get it to run I'll post some graphs to show what's going on. There are dozens of data variables that are tracked. All the data you need to be an armchair tuner.

I would say overall that the increase is much greater than when I first put on the free flow filter and up-jetted.

Going from a completely stock system is much more noticeable, but that's because you NEED to get a freeflow filter with the EFI. I'll try to get them to include it in the kit.

But it's not just the power increase but where it's distributed.

Greg

[D-cat]

Maybe the full article when done could get a permanent residence in the How-To section.

Meanwhile I'm wondering how hard it would be to engineer a new snorkel to match up with the stock air box. Sure not as much air flow but would probably be more reliable in wet weather. If they engineer it, it may also solve things like where to mount the MAP sensor and such. I suppose the coffee can trick works too though.

[justbuggin]

this makes me to want an efi system for my 150 that much more and yes this is helpful

[mainepeace]

Theoretically it shouldn't be *too hard* to make an adapter for the stock snorkel as it's just rubber. I guess if you wanted to you could cut the snorkel midway and put on a stiff rubber hose that has a 44 ID or so. I originally didn't want to do anything *weird* to initially put on the system so I could evaluate the system itself.

Since I have several airboxes floating around now, from my rental scooters, I might try modding one in that way. I don't mind too much the added noise from the free flow filter, but my parents will definitely mention it. And the drone can get to you after several miles if you're at WOT. I'm not sure about other scooters, but on mine it sounds like a chorus of frogs croaking. That's one reason I was thinking of the velocity stack. Anything that can cut down on the noise. When I rode one of the completely stock 50, which also had a 10" long snorkel that fed from the hole next to your feet, it was very quiet. I'd forgotten how quiet a stock airbox was.

Greg

[D-cat]

I've forgotten how quiet a stock exhaust is!

[mainepeace]

Here is the new switch I got. It has a separate power supply for the light. It glows blue when the lights are on. This one sticks out further in the back so the switch won't sit flush. Damn, I don't want to move it because there will be a hole there I'd need to fill it with something. Can't think of another switch I'd need. Oh well. It looks cool.

Greg

Attachments: