DIY Fuel injection thread.

[shrike 451]

Thanks for taking the time, ill do some more investigations and report back, also expect some Speeduino updates when I get my car going

[chris r 9,148]

So I've  finally had a play with the EMS, Same issues as @Transom. The ecu will register RPM input but seems to hold the ignition either low or high constantly. I haven't checked to see if injection side of things behaves the same

Has anyone had a play with ems or come across that?

 

[NickJ 4,056]

Finally got the machining completed on the throttle body - twin carb adapter plate, now to figure out all the nitty gritty details, timing triggers, fuel delivery etc.

[shrike 451]

Anyone know if putting a wide-band just before the cat will be an issue? trying to keep the factory O2 in the dump pipe at this stage

Something like this, will be shortening the 90 degree pipe however

 

 

[Yowzer 34,811]

Before the cat will be mint

[shrike 451]

Just now, Yowzer said:

Before the cat will be mint

So in the bung on the cat before it goes through the cat material? I can add another bung before the flexi or further up the pipe if need be, its a 200 cell cat so hopefully not to much restriction

[ajg193 9,656]

So my 14point7 Spartan2 lost it's heater output a few months ago.

These things have a 2 year warranty (I had about 4 months left), so I contacted Alan at 14Point7 and he sorted out a replacement.

Just received it today. Good customer service from him, would recommend.

 

As a side note I cut the old one open and found the housing is simple 3d printed plastic

[yoeddynz 22,989]

That is good service. I think I'll be getting one of them thank you please.

Is the chinese bosch sensor still going ok?

 

[kpr 6,497]

Old China sensor not drawing too much current for it? 

[ajg193 9,656]

In other news the old controller wasn't actually dead. The new one behaved exactly the same as the old one so alarm bells went off. Turns out I just had the heater ground wire disconnected... Man I feel bad about doing the warranty claim now. Should probably send him a donation

 

Yep, chinese sensor still works. I ordered a second as a spare but it never turned up. 

[yoeddynz 22,989]

I  buy controller off you.  give you monies i do. Five dollar. Buy now. I buy. You controller you not need. I buy. 

[ajg193 9,656]

I cut the old controller up and bodged on a replacement mosfet. I'll have to test if it still works after my bodgery and make a new case for it.

[h4nd 6,967]

Pics

[ajg193 9,656]

I'm actually extremely surprised that the chinese oxygen sensor is still outputting somewhat reasonable looking numbers after sitting in the exhaust without being powered for over 5,000 km.

I'm certain it has lost some accuracy, next time I get some motivation I will through a legit sensor in the exhaust again to see how it compares. My suspicion is that it always says stuff is a little closer to stoich than it really is. Engine seems to run fine with how much it is leaning it out at any rate so I'm not going to complain just yet.

[Roman 14,427]

I've been spending some time with ITB Alpha N setup on the Echo. 

It drives much nicer than I'd expect, and stays near to tune much nicer than I'd expect too. 

I've just tried something different though. Which is instead of having throttle position as the load axis on both fuel and ignition tables. 

I have TPS as load axis on fuel, then the fuel table calculations work out the airmass in the cylinder. (so this incorporates barometric pressure, IAT, etc automagically)

Then I use air mass in the cylinder as the load axis for ignition timing.

So this means if you are at 100% throttle but the airmass is less, it'll automatically move a little further up the ignition table and add timing. 

Then if the air is super cold and dense, meaning a fuller cylinder it'll pull timing to suit. 

Still needs a compensation table to trim timing for likelyhood of knock at high ECT and IAT, and then add timing for low ECT and IAT when motor is cold and combustion isnt great. 

But at least it's seperated out these variables from trying to trim for airmass as well in the compensation table. 

 

[Roman 14,427]

Ahh I forgot to say that one of the benefits of this, is that the shape of your ignition table doesnt have to follow the weird shape of the fuel table, that is weird because airflow through a butterfly valve is weird.

Because its just looking at how full the cylinder is, it doesnt care what the throttle angle is (which makes sense) 

So the ignition table becomes a lot more linear, it doesnt need an aggressive ramp up of timing around the "knee" point where load drops off really quickly at a particular throttle angle.

It also means, for better or worse that if you make any changes that allow your engine to consume more air (or less) when you update the fuel table the ignition table "should be" corrected to it as well without doing anything.

(Double edged sword obviously)

Top one fuel table, bottom one ignition table that's been translated across from my previous TP based one that was a bunch bendier:

[ajg193 9,656]

Similar to the GM approach?

[Roman 14,427]

1 minute ago, ajg193 said:

Similar to the GM approach?

Similar to any MAF based approach I guess, with my other car using a maf both of the tables have "air per cyl measured" as the load axis. Id just never considered using a different load axis for each of the tables. But it makes some sense for Alpha N and maybe MAP too.

[ajg193 9,656]

I wonder if I can make the microsquirt concert map/rpm into mass per cylinder

[Roman 14,427]

What are your current options for load axis? 

You could use effective injector pulsewidth as a similar-ish thing.