DIY Fuel injection thread.

[Roman]

Hmmmm okay.

 

How come it only seems to be effective at higher rpm then?

 

I'd have thought if it was primarily evaporative cooling making the difference, it would be just as if not more effective at low rpm as well.

 

Maybe airspeed something something.

Either way, will be interesting to see how it goes.

 

I've got the loom completely back out of the car now and tearing it to bits. Lots of junk in there still from the factory car's loom that I dont need.

 

Trying to get my head around the wiring required for setting up E-throttle. It says it needs two TPS singals and two FPS signals to work.
I'm guessing this is talking about two of voltage wires on the same TPS, rather than two whole seperate TPS units... Has anyone set this up before?
 

Will be cool to make the car a little less jerky to drive at low throttle by giving a non linear opening rate.

And cruise control

[Rookie]

I can't really remember what it said, but I think it is to do with the air speed and the travel length, so at low rpm / low throttle the fuel can fall out of suspension but at high rpm it doesn't get a chance.

And yea +1 to 15 bar fuel pressure.

[Lith]

Cant remember exactly about drift, but the thing is that the calibration for some of my sensors was a bit of a bodge because I just wanted to get the car going.

I just guessed that my car would idle at about 40kpa, and that was my map sensor calibration haha! So this time I want to do it properly. (First time with an aftermarket ECU)

 

Okay soooo regarding staged injectors. The idea is that for an NA motor you have a set of injectors higher up the intake runners, that come on at high rpm. Gives more time to mix air/fuel at high RPM and so a little more power for 'free'. (yet to be confirmed...)

 

 

Sorry I wasn't asking for an explanation about what having injectors higher up the runners is or why - I totally get it, I was expressing enthusiasm that you sound like you're going there and it's awesome  Be keen to hear how it all comes together when you've got it nutted out.  As mentioned elsewhere, evaporative cooling SHOULD allow for some power gains.   I understand the effect and concept of injection timing well too, was just curious about how you were doing it as it sounds like you were trying to guess it and it can be quite a beast... especially when playing with different injector positions etc, I've played with checking lambda for given PW using different injector timings as well but yeah - dyno tuning ftw in this kind of thing.  Be interesting to see if you find injector timing having a big influence on power at high load/rpm with the outboard injectors as it seems that the longer the duty cycle the less it seems to matter.

Have helped a mate a little with some interesting playing around using MoTEC M1 Build regarding injection timing with splitting between port and direct injection and working out how to bias between the two and how to determine the injector timing once the bias is determined in order to make sure that the fuel calculation model stays "accurate" is a potential headf**k when it comes to implementing and tuning to the whole thing.  I am sure doing what you are doing is not necessarily going to be a lot easier :/

 

In terms of the tuning drift, I guess if the MAP sensor calibration is off then that could explain a lot, depending on how "off" it is and how the map has been put together.  It's potentially possible for it to be off but still do a suitable tune, bearing in mind that to a degree the numbers themselves are more for the benefit human part of the mapping equation and the exactly manifold pressure value is just part of an equation... but this comment sortof moves in the direction of something that made me ponder on your second set of injectors:

I can't really remember what it said, but I think it is to do with the air speed and the travel length, so at low rpm / low throttle the fuel can fall out of suspension but at high rpm it doesn't get a chance.

 

 

Have you considered the influence of the fuel's state change/wall wetting?  I can only imagine that the further away the injectors are away from the valves etc the more subject to changing state based off a mixture of the surface temp and the air temp - so your IAT sensor calibration and correction tables being correct may not be enough if your setup is actually more vulnerable to charge temp changes than most would have to deal with.  Not sure if I'm overthinking it or not, be interesting to see how you go when you get everything configured more accurately.

[Roman]

Oh hey yep cheers, sorry if it came across as I was trying to teach you how to suck eggs. I know you (and others) have done a lot more of this kind of thing than me!

Regarding what you're saying about temperature stability.
My plenum is thermally isolated from the head by a rubber spacer, the only way it really gets hot is from radiant engine bay heat.
For sake of keeping temps downstream of the air temp sensor more consistent, I experimented with wrapping the manifold with foam / aluminium foil to reduce radiant heat increase.
Which (according to the hand-ometer) seemed to keep the plenum cooler after a long run, presumably at a more stable temperature.
Hopefully meaning the IAT readings are more consistent with what's happening closer to the motor. But either way, it never got to the point where it was too hot to touch. Maybe 30-40 degrees max. So even with evaporative cooling bringing the temps down a little it's not going to see big 80 degree swings or anything like that.
 
I've still got a heap of spare inputs and outputs so I might chuck a surface mount temp sensor on one of the runners for curiosity's sake though. To see if those injectors are actually cooling the runner walls by any measurable amount.

Situations like this are where a MAF sensor earns their keep! Because it automatically accounts for all of these variables which are difficult or impossible to build compensations for, into a MAP setup.

Hopefully in another 2-3 weeks I'll have a clutch and can start putting the car back together.

[Bangbug]

I find temp of the manifold is only relevant when hot starting or when sitting in traffic etc.

[Lith]

Oh hey yep cheers, sorry if it came across as I was trying to teach you how to suck eggs. I know you (and others) have done a lot more of this kind of thing than me!

Regarding what you're saying about temperature stability.

to see big 80 degree swings or anything like that.

 

I've still got a heap of spare inputs and outputs so I might chuck a surface mount temp sensor on one of the runners for curiosity's sake though. To see if those injectors are actually cooling the runner walls by any measurable amount.

Situations like this are where a MAF sensor earns their keep! Because it automatically accounts for all of these variables which are difficult or impossible to build compensations for, into a MAP setup.

 

 

No didn't feel that way, and I'm not precious nor am I sure that I do have any more experience or know any more than you here.... I've never done what you're doing with this, part of the reason for the interest!  Guessing we all may have something to learn from your experimenting here if you share it  I'm just trying to make sure we're talking on the same page as there is some interesting stuff that could be explored here, again I've not tuned a setup like this so quite interested to see what challenges and advantages you see.

 

So you mentioned Link earlier, but MAF now - not sure I follow, you are running a Link or something MAF based?

 

In regards to the wall wetting, again not quite sure if we're on the same page or not but just checking (apologies if I also am wandering into "suck eggs territory).... you are familiar with the consideration of "tau" in terms of fuel delivery calculation?  Imho the Link software isn't that well setup for accomodating for it even though they're sortof indirectly dealing with it (kindof) with their charge temp correction tables and acceleration enrichment logic and they've introduced to get around the effect of the evaporation factor.

 

There will always be some kind of fuel filming when the fuel comes into contact with the intake tract, "tau" describes how long it takes before the film starts evaporating again and it is influenced (like most things in this topic) by pressure and temperature.  What it effectively means is that the temperature of the air doesn't just affect the effective air/fuel ratio due to air density, but there is also the matter of how much of the fuel vapour actually makes it into the cylinder for burning.  Part of the reason for ECT correction tables (ie, cold start enrichment) is because of how poorly fuel tends to evaporate at lower temperatures and we end up will a crapload of liquid fuel in the inlet path - or for a given amount of fuel injected, a much lower percentage of it actually useful for the combustion process.   This phenomenon never goes away, it's just the "tau" effect greatly reduces as the temps get up... I'm just saying I wonder if a longer path for the fuel flow to travel will increase the tau effect to the point where there may need to be more consideration of it's effect will need to be made with tuning than normal.

 

Check this article out if it's not something you've seen or familiar with it, it's pretty good, albeit mainly focussing on transients: http://www.megamanual.com/ms2/xtau.htm

[Roman]

So you mentioned Link earlier, but MAF now - not sure I follow, you are running a Link or something MAF based?

I'm using a Link G4+ Xtreme, the wire in version that does not have an internal MAP sensor. (Has baro sensor instead) So I can run MAP or MAF, or whatever.
 
I put an initial tune together using MAP sensor, but wired in the MAF for datalogging purposes for setting up VVTI etc.
Then for interest's sake I decided to put together a MAF only map, with limited success so flagged it. (My fault, I selected wrong type of temp compensation)
However the IAT in the MAF was so terribly slow to respond so I replaced the whole thing with a fast response IAT. Which made a world of difference!

With the MAF's IAT I could be sitting at staging at the drags heat soaking, and it wouldnt even start to register a temperature drop until I was half way down the return road.

The new IAT changes value with a snap of the throttle, it's awesome.

In regards to the wall wetting *snip*

Hey wow, thanks! I've not heard it described in that way before, that's an interesting explanation and article. I've only really considered it from within the context of cold start, but it's never really made much sense to me why you needed to tip a bunch more fuel in to get the engine to run.

[Lith]

 

I put an initial tune together using MAP sensor, but wired in the MAF for datalogging purposes for setting up VVTI etc.

Then for interest's sake I decided to put together a MAF only map, with limited success so flagged it. (My fault, I selected wrong type of temp compensation)

However the IAT in the MAF was so terribly slow to respond so I replaced the whole thing with a fast response IAT. Which made a world of difference!

With the MAF's IAT I could be sitting at staging at the drags heat soaking, and it wouldnt even start to register a temperature drop until I was half way down the return road.

The new IAT changes value with a snap of the throttle, it's awesome.

 

Hey wow, thanks! I've not heard it described in that way before, that's an interesting explanation and article. I've only really considered it from within the context of cold start, but it's never really made much sense to me why you needed to tip a bunch more fuel in to get the engine to run.

 

 

Ahhh that makes sense, cool - having a MAF wired in would make for some generally really interesting comparisons and data  

And gotta love (/need) a quick reacting IAT, I find them (and baro pressure too, actually) quite mesmerizing to watch and compare when dyno tuning and road tuning.

No worries, the effects of wall wetting & evaporation and X-Tau is a handy thing to know about ... it's interesting too, when people outside of tuning seem to talk about the black art I think they imagine certain areas as where the wizardry is and not realise how much of the art is really just trying best to model the ECU with what actually is going on in the real world.  Finding power doesn't tend to be the hard part, the devil is in the detail

[Roman]

Hey people,

 

I'm redoing my wiring loom at the moment, as it's still had craploads of extra wires etc that are now redundant in it.

 

Previously all of the wiring bunched up over the injectors area, to one side of the engine. and then ran individual wires to the top of the motor for the coilpacks.

 

However now it would be desirable to move the bulk of that wiring to the top of the motor, so it's all under the plastic engine cover and just comes out the back as a single bunch.
but it would be running over top of the hot motor and the Toyota individual coil on plug coilpacks.

 

Is this likely to cause me issues, either from heat, or electric interference from the coils going into everything else? As it includes the crank trigger wire. (That's sheilded)

[downtrail]

Errr will probably be ok but alwsys best practice to keep stuff away from coils, ignitors and alternators. Vr sensors are quite good with noise however.

[Rookie]

I would second that being a dumb idea.

[Roman]

Setting up some new fun bits at the moment:

 

-A 6 switch panel for turning things off and on

-proper knock sensing

-ethrottle

-Redoing the 8 injector thing (Although still waiting on better firmware to get it running properly)

-clutch switch to ECU

-brake switch to ECU

-Cruise control switch

-Second map sensor and IAT for datalogging stuff

 

E throttle is so friggen greedy for inputs/outputs!

For safety reasons there needs to be two TPS signals and two FPS signals. Plus the + and - for the E throttle motor. And a PWM output for an electric clutch that's in it. And an output to switch a relay to cut power to the system if there's a fault code.

 

So 8 inputs/outputs hogged by just that haha. And then one more for a cruise control multiswitch, because why not

 

Still got inputs and outputs to spare though which is cool
 

Looking forward to experimenting with the e-throttle. The main thing I'm after will be having a non linear opening rate so that it's less jerky to drive around at very low throttle.
 

And some launch control shenanigans for cheating at OS drags haha.

[Rookie]

Keen to see you 2 stepping and flat shifting Roman, make it happen son.

[Roman]

Same cant wait

[Roman]

Got my E-throttle setup today, which has been interesting.

 

Basically there are a bunch of values you can adjust which determine how responsive/laggy it is.

 

If you set these values too high, the throttle plate flutters because it overshoots the value and then overcompensates back the other way, and so on.

 

So took about 1/2 an hour to fine tune the variables to give the best response without any flutter.

 

It's a bit of a laugh seeing your throttle plate flap around crazily when nothing else is going on haha.

 

Very dangerous if you drove on this setup without taking the time to fine tune for your individual motor, the suggested settings were way off what works on mine. I might need to end up fine tuning it a bit more later on when the motor is running, because at the moment its running at 12v supplied to the motor instead of 14 when the alternator is going.

[yoeddynz]

Interested to see how you get on. I like the idea of changing curves because the Viva has a very snappy throttle which coupled with a super lightweight flywheel makes for a sensitive throttle.

Im used to it and most people get used to it quickly but still wouldnt mind more of a rising rate setup instead.  I looked into making a bigger TB pulley but I cant because of my enclosed cable routing through the plenum. I cant gear it down at the pedal end because design wont cope with that much travel. Thought about offset pulley at TB which would make it rising rate but cable will rub at extremes.

 

E throttles do scare me though. I picture my ECU having a spaz suddenly just as I slow down at a crossing and the Viva lurches forwards to take out an entire family. Scary. Its one place Im happy to stick with a wire and some decent TB springs.

[Roman]

Yep I hear you, I've tossed it up a bit, pros and cons etc.

 

It only takes a split second of the throttle running on, and then you're flying off the side of Takaka hill or Scenic Drive which is less than ideal!

 

It's fairly robust with its safety features though... The TPS outputs two seperate signals, and so does the FPS (foot position sensor) So if one of them goes faulty, it sees that it doesnt match the 2nd signal anymore and shuts the system down.

 

It also has a relay controlling power to it, and an electronic clutch to disengage it both as safety features if the TPS readings arent as expected etc.

I guess another way to think about it is, a cable pull throttle can still jam open as well. And some people have had that happen too.

 

The Atlezza TB has a cable pull system, that if the electronic junk shuts down you can still open the TB half way by cable. So can limp home if need be.

 

There are zillions of Altezzas out there which all seem to be running fine with their factory fitted E-throttles...

 

Lots of people moan about E-throttle being mushy and unresponsive, but I reckon it's all about how the factory ECU is controlling it, not a physical throttle lag.
Looking at the datalogs the physical delay in throttle angle is near negligable,

An interesting use of E-throttle that I heard of, is a guy who's in a race series where there's a "Breakout" laptime that you're not allowed to beat. He sets his E throttle to limit throttle opening so that when he's going full throttle, he goes just over that laptime every time. But then when he gets stuck in traffic, he flips a switch and it reverts to full throttle angle and he roars past everyone. Flips the switch again and goes back to "laptime" mode.

Or the cars with big horsepower, having a "rain map" that makes the bulk of their pedal angle more usable in the wet.

Most of these things dont apply when I've got a peaky motor with only 200hp, but will be fun to play around with anyway
 

[yoeddynz]

Please do post photos on this thread of your setup. Yeah I like those above safety designs in a E throttle you mentioned.

 

Plus you could fit a bluetooth dongle thingee to a MS equiped car running a E throttle and using MS droid rev the engine whilst standing away from the car !

[RevDWC]

Or your kids could hack in via Bluetooth when they think you're driving too slowly.

[Lith]

@yoeddynz - I thought MegaSquirt weren't going to develop/support E-Throttle?