Why would anyone in their right mind replace a perfectly working stock ECU?
First I need to explain Japanese car classifications. There are two tax classes of consumer cars in Japan: normal, with a white license plate, and “kei”, with a yellow plate. The “kei” means lightweight; it’s designed mostly for people who need a small cheap commuter car, or a small truck for hauling around bags of rice or whatever. Kei cars must fit certain manufacturing criteria: they have to be no larger than 3.4 meters long and 1.48 meters wide, and no taller than 2 meters. These dimensions lead to many manufacturers making a very boxy looking car to maximize interior space. Their maximum engine displacement is limited to 660cc. Finally, they are restricted to making no more than 64 hp. And of course, while all Japanese cars are limited from the factory to go less than 180km/h or 160km/h, keis are usually restricted further; my governor kicks in at about 135km/h with a disorienting and rather abrupt jolt.
The engine size and power ratings dovetailed pretty well until about 1990, when manufacturers started fitting fuel injection and turbochargers to their engines. Those tiny 660cc engines started putting out more than 64 horsepower, and so the manufacturers detuned the fuel and spark mappings and added turbo bypass valves to meet requirements. As a result, the power curves for the cars started to look really weird. Whereas normally the power of a combustion engine usually increases until just before redline, my car hits 64 hp at about 4000 rpm and then just stays there until the 8500 RPM redline.
I figure if I derestrict things it should make 90hp peak quite handily. That might not sound like much power, but considering it’s coming from a tiny 660cc 4-cyl engine running regular 87 octane fuel and powering a car about the size of a go-kart, it’s enough. And as a nice side effect, no more nanny speed governor either, which means… track days!
According to my buddies, the power restriction is effected through a turbo bypass valve or two, controlled via the ECU. Obviously modding the valve to stay closed all the time won’t work, because the stock ECU won’t deliver enough fuel and things will go boom. On top of that, I don’t know whether the speed governor retards the spark or cuts fuel or both. In short, I don’t trust the damn thing. So to make things simpler (ha!) I figured replacing it with a custom fuel and spark computer would be ideal.
There are a lot of cool things about this approach. First of all, I get some hands-on experience building it myself. Secondly, it has a serial port on it so I can hook it to my laptop and adjust fuel parameters and spark advance through software. Of course, it’s cheaper than an aftermarket piggyback ECU. On top of that, if tuned properly the economy and emissions may actually be better than stock. If I do the wiring harness properly I can just pull my ECU out and reinstall the stock ECU when it’s time to go home. And finally, it’s a chance to learn about engines and apply my engineering background to them– a few weeks ago I had no idea what spark advance or fuel injector pulse width was.
I got ahold of a wiring diagram and have been poring over it (and translating it) for the past couple days. The stock ECU controls two fuel injector banks, two spark plug coils, an idle air stepper motor, and two boost control valves. Information about the crankshaft position comes from what I’m guessing is a variable reluctor sensor, but I don’t know yet how many teeth the stock timing wheel has. A good ECU automatically controls how much extra air comes into the engine by means of a valve or stepper motor, which helps when starting cold. As for those two solenoids… they regulate boost pressure, but as of right now I don’t understand their default operation… by default they’ll restrict boost pressure a lot, which I guess is fine when I’m just starting to tune the engine, but I’ll need to get these working properly. There are already stock O2, coolant temperature, and air temperature sensors installed; I believe I’ll be able to calibrate those and not have to replace them.
The only other oddball piece of hardware is a wacky throttle position sensor; it has just a switch for idle and a switch for “power”. This piece of junk obviously needs to be replaced, hopefully with a junkyard part that I can just bolt on. Hopefully.
I am all for doing nifty projects yourself, especially for the sake of advancing personal knowledge. That being said, I would be VERY careful with this project. I would also urge you to look into a replacement ECU, because chances are someone else has already done this before and better. Also, it may be possible to remove the EEPROMs (or equivalent) in the ECU and get them re-flashed with a custom fuel map, etc. By the way, the booster solenoids restrict boost at low RPMs so you don’t create too much power before the engine is spinning at the proper rate and cause catastrophic detonation.
This has nothing whatsoever to do with the content of your post, but I was wondering if you’d see Yongfook’s April 4th post? Might be something you’d be interested in seeing as you fit the description perfectly.
http://yongfook.com/2007/04/04/are-you-an-english-teacher-in-japan-with-a-blog/
Eric: this seems like the best existing option that doesn’t cost a fortune. It’s been tested on several hundred vehicles, and the source is open so I can fix it if it goes wrong. There are aftermarket boost controllers but they cost in the neighborhood of $200, and aftermarket ECU replacements are nearly $1000, which is more than the car is worth.
The good news is that I figured out what the two VSVs do. One of them is a VSV to hold the wastegate closed (to build boost until just before the wastegate would open on its own, basically helping with the problem of leaky and early-opening mechanical wastegates) and the other allows the blowoff valve to close, allowing the system to make boost without the pressure bleeding off. Both fail to a safe configuration, so if I leave them unpowered for now I’m not going to break anything. I need to test with minimal boost for initial fuel maps and advance anyway.
Don’t worry, I’m not doing this alone– I’ve got a couple of my bike shop buddies helping me out with this project. They tune bike carburetors a lot; one of the guys just finished re-carbing an old Yamaha 80s sportbike that he picked up used, and the power curve is the smoothest I’ve ever seen from a carbureted dyno run. And I know I need to run rich when under boost, around 12:1 instead of stoichiometric.
There’s a good set of instructions on the site, and I’m going to follow them to the letter. I also bought (not assembled) one of those nifty stim tester thingies, so I can check my soldering before I plug it into something that squirts and ignites flammable liquid.
Hmmm, so I think I understand why my max power is 264ps@5500rpm when the redline is 6700rpm (and is restricted to 180km/h when it can do 254km/h)
To their credit, I’ve heard that when they have restrictions like that they can sometimes add a little bit of low end torque and still keep the high end pretty flat, thus improving drivability. Sacrifice a spiky 20ps at the top end and you can get 3 or 4ps on the low end, or make the power curve smoother overall, or even make the throttle a bit less twitchy if you’re using fly-by-wire. Of course, that’s assuming their power curve is anywhere near normal to begin with.
That, and I don’t think most Japanese drivers would know how to handle that much power. The ones that do generally derestrict their rides, but it costs a pretty penny.
I am appalled to hear that aftermarket ECUs are so costly. It is really good that you have knowledgeable friends helping you, that could prove to be your biggest asset. Don’t forget, you might need to switch to a higher octane gas as you increase the power (maybe an octane additive? I don’t know how gasoline grades are in Japan). If you have ready access to a dyno, you should be golden to tune your project. Keep my informed, this should be a sweet experiment.
Yeah, like you mentioned, I think mine does fly-by-wire and they chopped off the top end on the power curve. It does help the drivability, because I’ve started to notice that on power take-off, it does try to rip the steering out of my hands. Also, since it’s driving the front wheels, it’s probably likely that any more torque and it won’t be able to transfer it onto the ground without traction control shutting off boost.
It’s also possible the stock clutch/transmission/tires simply can’t take any more torque without slipping or shredding gear teeth. On a stock car, sacrificing a bit of torque or horsepower in order to double the lifespan of critical parts is usually a smart move.
First things first: Sorry about dropping off the face of the earth when we talked last. Right about the time that you got in touch with me, The Maternal Unit started a very long discussion with me.
Um, I wouldn’t do it. I haven’t read through the comments yet, so it’s possible that someone else has said this already, but there’s more to meeting the legal requirements of a 軽自動車 than simply meeting the requirements. Remember, when trying to make a profit, you don’t build any more into a product than is required. As such, some components on your car (e.g. clutch, transmission, drive axle, steering rack, etc.) may not be able to handle a 25-30% increase in power as is.
Another big concern: How well programmed will the new ECU be? I know you’re a computer engineer by training, but honestly, this isn’t something you want to do yourself on your only car. Remember that at some point you will have to have 車検 renewed. I don’t know what the Japanese requirements are, but in Houston at least, they will check for emissions, and if you fail, it’s gonna cost you. Also, remember that even if the original ECU wasn’t programmed all that well, it was at the very least programmed by someone who, theoretically at least, has a much more complete knowledge of that car and what it needs than you do. My big concern: that in your lust for power, you’ll permit an ECU design that will allow your engine to overheat, with catastrophic failure as a potential result.
Remember, though, that you’re also hearing this from someone who has had a hell of a lot of trouble resulting from work done to my old car by supposedly qualified professionals.
As for the “too much power for the drivetrain” argument, you’re probably right. There’s likely a piece that won’t be able to take that much strain, and the best I can do is replace something if it breaks.
As for 車検 emissions, I’m not worried about that at all. Worst case scenario I just plug the stock ECU back in, a 3 minute swap and nobody is the wiser. I’m trying to find plugs so I can wire up a harness because when I finally leave this country I want to take this box with me. And I certainly don’t want to do wire crimping in the car. Ick.
Finally, I’m using a wideband O2 sensor to measure the fuel-air ratio at the exhaust. With this I’ll be able to determine the stock fuel-air ratio and match it before I start doing tweaks, and also make sure I’m not going too lean at any point, including acceleration enrichments. I do know that running 14.7:1 on boost is a quick way to melt pistons, a mistake I intend not to make.
Finally, although this is my daily driver I do have a backup 2-wheel commuter, so even if the whole thing goes to hell in a handbasket I won’t be stranded.