Ecotrons Trike.

[bigbadbob]

Hi Guys.
I've been reading through a lot of the Ninjette Ecotrons thread on here and got loads of usefull info so thanks for that.
@RacerX suggested I put up some stuff about my project here so here goes-

Basically i'm using the Ecotrons EFI kit designed for a Ninja 250 on this-



It's running a citroen 2cv6 engine. flat twin (boxer), 602cc.
Ecotrons do a kit that works for this engine but the cost of shipping and import taxes put it out of my price range. So I got a 2nd hand Ninja 250 kit and a BMW F650 throttle body and got it to work with my engine.

Naturally, as I didn't buy the kit from Matt, I don't expect him to support a project that it isn't designed for and I respect him for that. he's running a company not a charity.

What it does mean is that in order to understand how the Ecotrons ECU works I have to ask on forums like this handy place where guys have first hand experiance of the kit.
I'm a complete tech head so apologies in advance if i go a bit ott with the tech stuff.

anyway. here's a couple more pics of the project...

F650 throttle body mounted on my inlet manifold, the long manifold splits the intake air flow out to each cylinder.



And here's the trigger wheel that controls the injection timing. Thanks to @speedpro for his suggestions with this.



As I'm using one throttle body and one injector to supply two cylinders and have to fool the ECU into thinking it's on a Ninjette, I have to use an 8 tooth trigger wheel instead of the Ninjette's 4 tooth flywheel.
All so it fires the injector every revolution of the engine. (it fires into the manifold and whichever cylinder is on the intake stroke at the time draws the charge)

was that a snore I heard? anyone nodded off yet.

The rest of the kit is wired and plumbed just as it would be on the Ninjette.
The maps are from a guy in Italy who used a different Ecotrons Kit.
I've imported them into my ninjette cal file.
And Yes.. it runs really well so far. it idles, revs, survives throttle blips etc.
I've not had it out on the street with EFI yet as i'm also rebuilding the front end, but things look good so far. a bit on the rich side but I can do some tuning to bring it down once I get my head round what the ECU is doing in learning mode and Rich mode. and what all the advanced calibrations do.

And finally... Many folks ask WHY??????
There's several answers to that-
1) to see if I can.
2) cos someone said I couldn't.
3) cos I'm mad.
4) why not?

So if this kinda thing rocks your boat please
If it doesn't, thats fine. I won't get upset. hahaha. (much)

[micoulisninja]

First time I ever post on anything other than the "tech" section...
I love that thing Bob !! Two questions though...
1. why the 2cv engine instead of something... well... more updated ? was it just sitting around or was for reasons of easy maintenance, room on the chassis etc ?
2. Since the 2cv is also a 2cyl, why not use both TBs ? Wouldn't it be easier to tune it up ?

p.s. Is the front wheel off a suzuki bandit or what ?

[bigbadbob]

Thanks for your interest
Why 2cv?.... well I bought the trike off a guy here on the island, allready road registered. it was rotting in a shed and i wanted it. haha. I like a challenge.
here's how it was when i bought it-



I stripped it down to bare steel and re-built it how I wanted it.
Now i'm updating the engine a bit with EFI to get the best out of it.
I'm an electronics engineer by trade and an engine tinkerer by habit.
I had thought about using two throttle bodies but they'd probably have to be mounted out on the cylinders and the linkages would be a nightmare, what with it being a flat twin 'n all.
The other option would be two injectors, one TB. but I don't think i can get the Ecotrons kit I have to do a 360 degree injector sequence, the ninjette is 180 degree as you know. the trigger wheel settings in the nijette ECU are greyed out and are fixed in firmware I think.
The front wheel is either bandit or gsx600f, definately suzuki though.
Back wheels are original '70s chevy slot mags.
The tank came of my rotten Kawasaki 250 Scorpion and was sprayed by a mate.
My other ride started life as a 600 bandit and got a top end transplant from a gsxr750, that was another good project.

[Racer x]

I built a VW trike with a guy when I was a kid. The single TB is good . Performance is not what you want. Just smooth. The thing should do some great wheelies. We used shopping cart wheels as wheelie bars.

I can vision some steam punk design elements on that would look cool.

[allanoue]

[bigbadbob]

Yeah. vw ones are pretty smart. My mate has one he's supercharging as we speak.
Yes, mine will loft the front in first so i never use first unless i want to wheelie, pull away in second is fine. just don't snap the throttle open pulling out of junctions or you end up on the far side, good fun though.
Yep, i'm not looking for huge power gains just nice smooth running and accelleration.
p.s. I use steel trolly jack wheels to stop the frame hitting the deck.

[Somchai]

Nice

[dcj13]

The trike is cool! And very clever rigging an EFI set-up using stuff intended for vastly different applications!

[bigbadbob]

Thanks guys.
Clever or stupid... time will tell.
The project is on hold as i bust an oil cooler pipe, new cooler in the post, more tuning to do once i fit that.

[Snake]

Awesome looking trike.

[bigbadbob]

Well... I bust an oil cooler pipe and had to get a new cooler so the project has been on hold...
New cooler now fitted so we're back on track.
Something's up with the O2 sensor though as it's reading ok for a few seconds then drops to 14mv once it warms up.
I have a spare sensor but want to be sure i'm not flooding it with unburned fuel before i replace it.

[Racer x]

Pull the sensor and heat it with a torch. It should put out voltage. Definitely be sure your not flooding it.

[bigbadbob]

Cheers, i'll try that. on the graph it seems to put out voltage for a fewseconds before it drops to near zero. will see what it does with the torch.

[Racer x]

A propane torch not a welding torch. About 700-800 degrees is good

[bigbadbob]

Originally Posted by Racer x

A propane torch not a welding torch.

[Racer x]

I'm used to dealing with youngsters at work.

[bigbadbob]

Well, I finally got time to tinker in the garage...
Pulled the O2 sensor and it was black and sooty. a tad rich methinks.
Hit the sensor with the propane torch and once the soot burned off it started showing milivolts. was glowing red at 800mv.
I'll do a global lean setting on the ecu and try again.

[bigbadbob]

I got some more time to tinker in the garage tonight, been busy lately, the local bike rally was on... lots of beer and playing air guitar with a lady in a lycra catsuit to great rock/blues bands. life is tough huh..

anyway.. back to the trike.... I did a global lean and the o2 sensor is still dropping out. i gave good readings for a while as the warmup fuel factor ramped down to 1, i could see the mixture leaning in line with the warmup factor.
Then O2 voltage just dropped to almost zero which pushed the correction factor up to max.
Has anyone seen this before?
It did seem to run for longer than last time but it's also a lot colder out there so warmup took longer.
here's the plot anyway.
Does anyone know what PwrLsbHtNorm does? normalised heating power of lambda sensor????



There's a few things I can try next time-
1-Run in rich mode, it should just read the O2 voltage but not use it.
2- unplug the O2 sensor. but then i won't know what's happening.
3- lean it down even more in case the sensor is sooting up again.
4- any other suggestions?

[bigbadbob]

I had another play this morning and have found out what PwrLsbHtNorm does...
The ECU sends pulses to the O2 sensor heater to warm it up gently rather than hitting it straight with 12v.
I put an LED on the spare sensor cable so i can see when it's pulsing and when it's full on.
Basically.. when PwrLsbHtNorm is low it's warming up and when it goes high it's getting the full 12v and goes into open loop mode.

Playing with global fuel factors suggests i'm actually running lean so the O2 sensor voltage might be real. In which case i'm VERY lean.
The readings before PwrLsbHtNorm goes high may well be false readings.

I tried running in open loop/rich mode and it runs better so I either need to richen everything up or use a bigger injector.

Getting there though and learning a lot along the way which is kinda the whole point of the project.

Edit: I just pulled the O2 sensor and hit it with the blowlamp after it had warmed up and it's reading fine, so i'm definately running lean.

[Somchai]

Bob, even when I know that I like many times talk 'against the wind' let me say the following.
O2-sensor and catalytic converter are not good for power, they are emission-crap only and if you want to tune your engine and fuel-system in the best possible way then don't use this (if possible, idk the Ecotrons-system).
All positive settings you're doing will be cheated by the o2-sensor (within its working range).
You can have emissions or power, but not both together and also please forget about the worthless talk from auto-tune functions.
Anyway good luck for your project and here's as bonus

Link to original page on YouTube.

[bigbadbob]

Thanks for your comments and the bonus Somchai.
Yes I agree. luckily the Ecotrons system has a "power" mode that defeats the O2 sensor.
O2 sensor is good for getting the base settings right, then you can switch to a map that works well for power. the switch is mounted on the bike so you can run in eco mode around town and power mode on the highway at the flick of a switch.

I just played around with global enrichment and found a factor of 1.9 gets the O2 sensor swinging around as it should. acceleration is better too.
woohoo!!!!

Now I need to richen things up in the maps as global fuel factor is just a quick and dirty way to do it.

[bigbadbob]

That's better.

I've multiplied the map by 1.9 to correct for the global enrichment.
It's made the warmup blend into the steady state running much more smoothly.

[bigbadbob]

Multiplying the map was not a good idea, it gave me Volumetric efficiency over 100% in some areas. while this would be nice it's not right.
Welding up the exhaust leaks was a better fix.lol, Air was obviously getting in and upsetting the O2 sensor.
Now she runs just a little rich with the original settings.
I'm getting there slowly, one day at a time like a recovering alcoholic.

[bigbadbob]

I've been making minor tweaks to the settings and things look good.
I put one wheel up on my DIY rolling road, lol, it's just a frame about a foot square with about 8 bearings running on bike wheel spindles.
Works ok with one wheel on it as the diff takes care of the other wheel. I just tie the back of the trike to the towbar of my truck in case it jumps off the roller.
Anyway.... I ran it up and slowly let the clutch in.... revs up ok and the sensor readings look good, a little rich but if I lean it up she stumbles so i'll leave it as is for now.

[bigbadbob]

Just for shits n giggles I scoped the injector signal and compared it's timing to both MAP pulse and Ignition coil pulse.
Using my ancient analogue oscilloscope rather than a nice new digital one meant the readings were kind of hard to read accurately but it was enough to show that changing injection angle in software calibrations didn't change anything, I guess this is fixed in firmware with the ninja kit.
I did notice that injection timing advances during cranking and initial start and then is retarded once revs exceed end of start revs. (1krpm)
It also showed that my injection event is happening while the inlet valve is still open, but one or other inlet valve is open for most of each revolution so this is hard to avoid with my setup. doesn't seem to matter tho with my long inlet manifold, it acts as a plenum to hold the air/fuel mixture.

[bigbadbob]

I'm still not quite there with this. very close though.
I'm going to try a smaller injector as my current one seems to spend a lot of time at minimum opening time but is still rich.

One question for the experienced guys on here... do you have a stream of small bubbles in the return pipe from the regulator?
I don't think these come from the pump as they only show after the regulator, looks like cavitation in the regulator. probably not an issue but I thought I'd check with you guys.

[speedpro]

The stream of bubbles from the regulator is normal for me. I have had regulator problems which I only just discovered recently. Over the last 2 years as I've made progress I've been adding fuel to the VE table and eventually increased the injector size to 128g from 80g. It still leaned out as the revs and load increased. Turned out the regulator was on the way out. When I measured it I found the fuel pressure was only 25psi, should have been 45psi or so. I found the little o-ring inside was damaged allowing fuel to bypass the regulator capsule. The replacement o-ring didn't last one day but the new regulator is working fine. A very simple test is to remove the regulator and try blowing through it. Obviously you should not be able to. My mates spare one had the same problem. Now I'm back to the 80g injectors and the motor is a little rich at low revs/load even with the injector pulse width at a minimum of 1.4mS. I can now load it up so I get 1050mbar on the MAP and hold 12,000rpm and it is still a little rich at only 4mS injector pulse. It revs clean to 16K without leaning out as well on the inertia dyno. I highly recommend putting a fuel pressure gauge on it at least until you have it all sorted.

[bigbadbob]

Thanks for that Mike, I was tempted to replce the regulator with a more robust one anyway so actually measuring the presssure and not just assuming a cheapo chinese plastic regulator is going to be spot on makes good sense to me.

[speedpro]

Originally Posted by bigbadbob

Thanks for that Mike, I was tempted to replce the regulator with a more robust one anyway so actually measuring the presssure and not just assuming a cheapo chinese plastic regulator is going to be spot on makes good sense to me.

A couple of quick checks are, 1 - run the pump then turn it off. Fuel pressure should only decrease in the lines slowly. You can check by squeezing them. 2 - remove the regulator and try blowing through it. If you can then it's failed, same if it doesn't hold fuel pressure after the pump is off.

[bigbadbob]

Thanks Mike.
Mine does hold pressure when the pump is off, I nearly got it in the eye when I changed the injector.
I didn't measure the pressure though. haha.

[spooph]

Super cool project! Please keep posting updates!

[bigbadbob]

Thanks for your interest Spooph. :-)
Nothing much to report as I need to get my ass in gear and get the rest of the trike roadworthy before i do some on-road tests.
Will keep you posted when that happens.
BBB.