Episode 23. We tune and road test the EFI 420 cc engine in our Honda Insight.

hello hey jimbo how are you doing oh hey yeah i'm doing fine great great so uh i hate to bother you but you know it's been a while since you put out your last video oh yeah i know it's just that hey it's cool man it's just that i'm getting kind of itchy you know for another video you know kind of jonesing like sure i get it you know i think i got one right here here it is i knew i had around somewhere wow man you're the greatest do you think you know make some magic so i can see it real soon no worries it's on its way thanks man talk to you later welcome back to robot cantina in today's episode we find out that the cheapest efi kit on ebay may not be the best deal you get what you pay for we were able to get mixed results over at the hillbilly proving grounds during the tuning and testing disaster let's take a look [Music] let's take a look at what we got under the hood so we can identify what we're working with and then go ahead with some sort of plan and having a plan is good but not everything goes to the plan so obviously this is a single cylinder engine and only has one throttle body and that's a really good indication of what we're dealing with i know it sounds simple but in order to get maximum performance from this engine we're gonna need to tune it like it has itbs so an itb is an independent throttle body and this is the sort of fuel injection you'd find on extreme high performance cars like this porsche setup right here you can see each cylinder has its own throttle body and that's cool the problem is we don't necessarily have a real race car the honest truth is we have shitbox honda with a cement mixer engine i know the truth hurts but that's what we got and the rules we need to follow are for high performance itbs and the tuning method we're going to be using is alpha n or sometimes referred to as a n now before you start pounding words out on your keyboard the reason we're going with alpha n is because speed density will not work for low speeds on an itb or a single cylinder engine the best we could possibly hope for is alpha n with speed density blend and that's where speed density takes over at say 3000 rpm and up and normally that would be the goal but we're dealing with the cheapest efi kit on ebay as we'll find out soon enough it's not going to be up to the task let's take a moment and talk about why alpha n is used to tune itbs so in order for electronic fuel injection to work correctly it needs to rely on information collected from the sensors one of the main sensor signals is the map sensor and as you may recall this is the manifold absolute pressure sensor and it's responsible for reporting the vacuum being generated by the engine and normally with this signal we can extrapolate the load the engine is under the problem with itbs or in our case a single cylinder engine with an itb is it won't generate a constant vacuum signal you see the vacuum signal on a single cylinder engine is too uneven and electronically speaking the signal is trash at low rpms now as the engine speed picks up the signal gets a little bit better but it's always going to be flawed now there are plenty of tricks used on multi-cylinder engines with itbs but those don't apply on a single cylinder engine plus the ecu we're using doesn't have the capacity to mathematically average signals as it turns out the ecu we are using is severely crippled but we'll get into that a little bit later anyway alpha n is the method of tuning where we use the throttle position sensor and the engine rpm to calculate the fuel required it's not perfect and as a matter of fact is a bit of a hack now in theory in our application we would only use alpha n for idle and part throttle so it's not that bad but as i mentioned a few moments ago our ecu is crippled and it won't allow this and that's kind of a deal breaker but since we're here let's try to make the best out of the situation of course all this will be futile because we're going to be switching to a different ecu as soon as we can as you may recall in the last video we got the engine up and running with no difficulties at all sure it was way out of tune but it ran and it didn't take much to get it to idle it's easy to think the hard part's over you might think all jimbo needs to do now is fiddle with a few settings and we're off to the races well not exactly and i'll tell you what you can get an engine to run pretty damn good with all the settings completely fubar so even though the engine ran in the last video we still need to set some baseline settings otherwise i'll be fixing problems with solutions outside standard practice or for the metro crowd that's what we call half-ass fixes so the first step in getting our engine to run in proper tune is to find out as much as we can about the fuel injector the injector operating characteristics is a big part of the formula used to calculate fuel delivery now in our case all the injector info was loaded in the msq and it was completely wrong and yet the engine was still able to start and run and while it's true you can more or less trick the ecu to get the engine to run but if you have bad settings the engine will never run quite right the injectors used on our engine is an efij series but unfortunately we could not identify the exact model the good news is one of the critical specifications was found online and that was the dead time of the injector so this injector has a one millisecond opening time and that's often referred to as dead time and one millisecond can also be expressed as 1000 microseconds anyway we need to let the ecu know that for the first microsecond the injector is powered no fuel will flow so if the ecu calculates the engine needs a score to feel for three milliseconds it knows the injector needs to be powered for four milliseconds to compensate for the delay in the injector opening so right here we need to put the correct opening time or dead time now this kit was pre-programmed with 2.3 milliseconds opening time and that's way out in left field so the next thing we needed to find out was the impedance of the injector so the impedance is kind of like the resistance but it's a bit more complicated than that but generally all you need to do is use an ohm meter and measure between the injector terminals now we didn't need to do this because all the efij injectors are high impedance the primary concern with impedance is if the injector is of the low impedance variety now if that were the case it would take a lot more power to drive the injector and most aftermarket ecu's need to be modified for low impedance injectors anyway a low impedance injector will measure about 4 ohms and a high impedance injector will measure between 10 and 16 ohms the next question we had is how much fuel will flow through the injector at 3 bar or 42 psi so the flow rate on this injector was unknown and the specs online vary between 100 ccs per minute up to 330 cc's per minute now this is something that's important and guessing is not an option so without any real confirmation on the flow rate it was time to run an experiment so in order to determine the flow rate we're going to use the brute force method and that is to remove the injector and power both the injector and the fuel pump at the same time and see how much gasoline flows out in one minute now unfortunately this ecu does not have a utility to turn on the injector and the pump so we're gonna have to hot wire them now i remember the polarity of both the injector and the fuel pump from the bench test video so connecting them is easy with jumper wires so both the fuel pump and the injector are hot wired and all we need to do now is turn them on and let the fuel flow wow and that injector flows plenty well let's go ahead and fast forward this all right now let's measure what came out now i fool around with this stuff on a regular basis so i have some scientific type measuring containers this stuff is relatively cheap and has multiple purposes now normally gasoline volume is measured at 60 degrees fahrenheit or something celsius and the day i did the flow analysis the temperature was in the mid 60s so we're in the ballpark and there you go 305 cc's per minute now that doesn't match anything that was published on this series of injectors so i went ahead and did the test a few more times and my best guess is this is a 310 cc per minute injector and that's what we're going to go with so the flow rate of the injector is used to determine the required field calculation the required fuel number represents the longest time the injector will remain open at full throttle during peak horsepower normally you plug in the flow rate number along with some other data and the software will generate a required fuel number and of course with this software package the required fuel calculator is broken and it spits out very large numbers so there's that fortunately we have another option mega squared's mega manual has a flawed but good enough version of the required fuel calculator that we can use of course all the data we need to input is in imperial units so we'll need to convert some of our information let's see the 420cc big block works out to be 25.63 cubic inches now that seems kind of small you know back in the day i had a cadillac with a 429 cubic inch engine which works out to seven liters now at idle that car would burn a gallon of gasoline every 20 minutes anyway let's get back to whatever i was talking about let's plug in one for the number of cylinders number of injectors let's go with one again number of squirts one sounds about right okay injector flow rate now 310 cc's per minute converts to 29.54 pounds per hour injector staging well the default is simultaneous and that sounds about right for one injector the fuel type is of course gasoline or if you like petrol for the metric crowd so here we can see the required fuel number is 9.1 now from past experience i've learned on smaller and more or less stock engines the required fuel that's calculated is often on the rich side so i derated the number a little bit i guess they call this tribal knowledge and so we put that number here now at this point i could generate a ve table that would put us in the ballpark as far as setting the airfield ratios but once again the software is a bit janky so in this case i pretty much wrote the whole ve table by hand and it's rough the only area that works well is the idle region and everything else is more or less a guess normally i would generate the rough ve numbers and then use a wideband sensor to auto tune the ve table but once again we encountered a flaw in the cheapest efi kit on ebay now let's look at the instructions i finally found tucked away somewhere on the internet now these instructions had to be translated and it translates pretty good all things considered and here we go it says here wide band oxygen sensors are expensive and mainly used on high-end cars the current version of the ecu doesn't support wideband oxygen sensors then it goes on to say the software will allow the settings to be changed but it will have no effect on the ecu so basically you can set up the ecu for wideband but will not do anything um so this ecu will not accept a wideband signal and the only feedback it will accept is from the narrowband sensor and it will not auto-tune okay so not all of you folks know the difference between a narrow band and a wideband o2 sensor well the purpose of the o2 sensor is to monitor the exhaust and report the air fuel ratio back to the ecu if the engine's running rich or lean the ecu needs to know this in order to correct the air fuel mixture the narrowband sensor is common and is used on pretty much all cars since the 1970s the problem with the narrowband sensor is they can only measure accurately in a very narrow range or band and that's perfectly fine however when tuning an engine for more power the air fuel ratio will be outside the region that the narrow band sensor can accurately measure the narrowband sensor is pretty much useless when the air fuel ratio is above or below 14.65 on the other hand a wideband sensor can accurately measure the air fuel ratio between 10 and 20 so it can help the ecu calculate the air fuel ratio with precision there's a little bit more to it than that but that's the basics now pretty much all dui ecu's will accept the wideband signal that's because it's the preferred way to measure the airfield ratio and with the wideband signal the ecu can accurately self-tune the airfield ratios self-tune or auto-tune are really handy features and it makes tuning infinitely easier wide band sensors are expensive and that's why they're not typically found in the average car but recently more and more manufacturers are equipping the cars with wideband sensors anyway since we had the manual handy i continued reading and learned a whole bunch of other stuff so furthermore it will only do alpha n on a single table no dual tables are allowed and the manual also indicates many of the other features in the software can be enabled but they will have no effect so it looks like this ecu's crippled i think we'll do a basic tune for funsies and then upgrade the ecu to a speed duino n02c i think the speedwino will give us the functions we require plus it comes with a map sensor that can read boost and it's cheap anyway i had the engine up and running fairly fast and tuning the ve table directly was a bit of a challenge it took a few hours to get the car drivable enough to head over to the hillbilly proving grounds but before we go let's measure how much current this system draws because you know we ain't got a working charging system and oh boy do i have a whole video coming up on various charging systems so stay tuned for that so in order to measure the current that the fuel injection draws we're going to use this calibrated shunt now a shunt is pretty much a low ohms resistor and the current passing through it creates a slight voltage drop anyway with this shunt the voltage drop is proportional to the current and we can use the millivolt setting on the meter and not the amp setting i find it handy to use a shunt for most of my current readings because using the amp setting on the meter has the potential for disaster and it's all hooked up now i have everything electrical passing through the shunt everything except the starter because we really don't care about that in today's experiment now as you can see i have the shunt connected to the voltmeter and we're pretty much ready to go well with the ignition switch turned on we can see a slight jump in current as the fuel pump cycles then after that the current settles down to about 700 milliamps and that's just slightly less than one amp not too bad let's see what happens when we start the engine so when the engine starts the ecu turns on the fuel pump and it also turns on the heater for the oxygen sensor and after a little bit when the oxygen sensor gets warmed up the current drops back to 1.4 amps okay so this oxygen sensor ain't going to do us any good because it's not very accurate for what we wanted to do so let's take a look at a wideband sensor now we won't be able to connect it to the ecu but we can use it to edit the be tables by hand so this is an aem 30-0-300 and you can generally pick one of these up for 180 dollars they ain't cheap now this one i converted into a portable sensor and i use it to tune carburetors and whatnot the gauge is mounted in a sturdy plastic box and all the connections are made in the aluminum box it's got an on off switch a nifty handle and right here is the output wires that we could connect to the ecu but as you know our ecu is crippled and it will not accept the wideband signal anyway these are of course the battery wires and i have them set up with alligator clips and this of course is the connector and harness for the wideband sensor which we already have installed in the exhaust let's hook this up and see how much power it'll draw so we'll connect this here and now we can give it some power let's go ahead and start the engine now i have the narrow band oxygen sensor disconnected because it ain't going to do us no good with the alpha entune and at this point it's just a waste of power [Applause] so it looks like we got a little bit more tuning to do so the total power draw of the efi with the wideband sensor works out to be 2.7 amps so we should be good to drive out to the hillbilly proving grounds and have plenty of power in reserve just in case we'll bring some extra fuel this efi system doesn't have an active idle adjuster and that's fine anyway to start the engine you have to crack the throttle ever so slightly yeah go ahead and raise your hand if you can pull start your car yeah i thought so let's see we got a fuel injected cement mixer powered street legal race car with both electric and pull start and of course the pull start option as you know is to impress the chicks so right here is where the car stalled and she started ride up but that was completely unexpected [Music] now later i would discover that as the underhood temperatures rise the engine would start running lean and i think for good reason now we're running the ecu in open loop and there's absolutely no feedback i don't think that's supposed to happen with an alpha entune but there's a lot of goofy stuff going on with this ecu and it's hard to say what to expect anyway all the testing and tuning we did at the hillbilly proving grounds was done with the hood off the car and it ran fine so one of the things i noticed in the software is there's no setup screen for the air intake temperature now typically you have the option to the ignition timing when the air temperature rises i have no idea what this ecu is doing with the air temp data and after all the ecu is made for a motorcycle perhaps this is not even an issue on a motorbike it's hard to say so running lean is a problem and for most of the trip i had to keep an eye on the wide man gauge whenever the air fuel ratio went above 14 i had to find a different throttle setting and that seemed to work but unfortunately i wasn't able to keep a constant speed now the fastest i got the car to go was 60 miles per hour but keep in mind this is the first real time the car was ever driven with the efi so now on the way i did stop at the drag race portion of the hillbilly proving grounds and i did a quick 0-30 run yeah that's not too good but once again the airfield ratio was a bit janky so we'll take what we can get and when i arrived the engine was making a terrible noise and it turned out to be a broken bracket on the fuel tank and that's easy to fix so with the fuel tank bracket fixed we were back at it trying to sort out a few small issues now this is weird with the timing light i can see the ignitions right on but i can also see the coil is firing twice the thing is the ecu is reading the correct rpm and i verified that with the laptop but for some reason it's firing the coil twice now here you can see on my cheapo tachometer that it's picking up double spark you see the attack should be reading 1400 rpm and it's reading 2700 now the cheapo tachometer is directly connected to the spark plug wire so i don't know about that one and the truth is i really don't care we're going to switch the ecu for a speed we know anyway ready so with most of the issues sorted out chuck and i set out to do some on the road tuning now chuck drove as i edited the ve tables now this is really hard to do during the day because the sunlight washes out the laptop screen but somehow we managed to get a decent table set up basically i watched the airfield meter on the dashboard and when i see it go out of range i edit the offending cell like i said it's hard to do but being the engine so small it usually runs wide open that sort of limits the amount of cells that need to be edited now the ve table is not perfect but it was good enough to get some performance data [Music] [Applause] so in the end the fuel tank bracket broke again the thing is this fuel tank has always had a mounting problem the brackets that hold it down are stout but for some reason it keeps breaking something i'm gonna have to figure out a different way to mount the tank so this is a long video and for the folks who made it this far i have a little surprise so as you know we're going to upgrade the ecu and unfortunately we're going to ditch the turbocharger but in its place we're going to have an amr300 supercharger so things will continue to get interesting and hopefully we'll get into the supercharger stuff soon enough now let's talk about what we did manage to learn at the proving grounds [Music] well the trip over to the proving grounds netted us some fuel economy data and we managed to get 29 miles to the gallon of course the engine wasn't exactly in its best tune anyway as far as fuel economy goes unfortunately that's the only data point we managed to get now top speed with two people in the car and the windows open and the hood off was a brisk 60 miles per hour or 90 kph for the metric crowd now i'm not sure we can compare that with the 64 miles per hour or 103 kph we did previously but it looks like the potential is there the best 0-30 time we got was 10 seconds and that's with one person in the car and the hood off but i don't think the lack of the hood made a difference because that sort of stuff matters above 40 miles per hour so with some basic street tuning we were able to shave one second off to zero to 30 times ah it sucks but i'm really looking forward to getting some real tuning done with the new ecu now overall the efi certainly was not worth the trouble given the results we got but this is just the beginning and unfortunately we more or less have to start all over again but that's just the way it is now the car is going to be down for a few weeks as we go through it again but don't worry i have some small engine experiments we're going to be doing in the meantime well that's about it for today and if you made it this far thank you very much and i'll see you next time do [Music] you

Channel: Robot Cantina

Views: 75,489

Keywords: $99, 212 cc, 420 cc, 420 hemi, alternative transportation, custom, cvt transmission, diy, documentary, efi, electric car, electric vehicle, engine mods, engine swap, fabricate, fuel economy, fun, funny, go cart, go kart, gocart, gokart, golf cart, goonzquad, great idea, gx390, hack, hand made, harbor freight, hemi, hho, high performance, home made, homemade, hp, hybrid, lada, lawn mower, lawnmower, modification, modify, motor, motor swap, motorcycle, mpg, predator powered car, Fuel injection, fart

Id: SUHbsrEJkrY

Length: 25min 20sec (1520 seconds)

Published: Sat May 29 2021