Injector wiring

[Vertigo]

Question: why would anyone wire four low impedence (~3.3ohm) injectors together to a single 4AGE ECU pin? Found that kind of horror in the loom after I went digging.

[GDII]

The wire to the ECU from the injector is the earth pin. Quite often they are wired in pairs but a single wire is rather odd.
Wiring in pairs makes the injectors fire twice per engine cylce rather than one. Wiring all to one would make them fire 4 times per cycle. That makes it very inefficient.

[Vertigo]

Exactly - the No10 and No20 pins are grounds for the two injector pairs. However, all injectors were tied together on No10. I dont understand why anyone would do this, and will be fixing it soon, but just wondering if there was any reason, like perhaps to install different impedance injectors.

In addition, using a calculator shows the injectors have a combined parallel resistance of around 0.82ohms, which seems exceedingly low for a loom with no resistor pack. Surprised the driver FET hasnt exploded long ago.

Also, the power feed wire for the injectors seems to split and head to a pin on the ECU as well as the injectors, which it shouldnt be doing. Im trying to understand if this is typical for a NZ new AW11 or whether some cowboy decided to play autoelectrical engineer.

[GDII]

Does it look hacked up or factory?
Did this come from a running car?
Do you have the ECU that this loom came from?

Looking at different 80s and 90s Toyota 4cyl diagrams, there are a few variations.
You either have 1 wire to the ECU per injector with power directly from the battery or 2 wired together like you describe then to 2 pins on the ECU.

[Vertigo]

Looks factory, though cant completely rule out that somebody has modified it. The tie point for the wires especially looks like nothing Toyota would do, surely - its just four wires meeting at a point and another single wire of the same gauge heading to the ECU, with some heat shrink over the top. I havent opened up the heat shrink yet, but I will eventually. Car is running perfectly fine. Using either the factory or a Megasquirt ECU, it runs just fine. Its completely bizarre.

Edit: Btw, all of this completely explains why the engine ran really badly when I attempted to try alternating injector fire mode in the Megasquirt - the second output disappears into the loom and connects to nothing at all.

Edit2: There must be a resistor wired in there somehow. Ill find it!

[GDII]

Odd. Not smelling rich? Maybe it is factory. I'd been keen to know more if you find anything else.

So changing the injector mode made it run on 2 cylinders? Or just leaner because less fuel is injected to each cylinder per full cycle. See below for my hopefully sound theory.

Based on my limited knowledge of ECUs (or at least limited compared to what I want to know) each injector would fire as cylinder 1 is on the intake stroke therefore filling up the closed intake valve side of cylinder 2, 3 and 4. Then when cylinder 3 opens it would squirt in fuel to all 4 again filling up cylinder 1, adding more to cylinder 2 and 4. Then when 4 opens it would squirt in more into 1, 2 and 3. So at the end of one full engine cycle of 1,3,4,2 you get a very full cylinder 2 as it is the last to open from the initial start up. Over the next 4 cycles all cylinders would get 3 pulses of fuel before opening. So in theory you get 3 squirts worth of fuel pooling on the valve then one squirt 'atomized' in the intake charge.

That's how I see it. Maybe it's completely wrong though.

[Vertigo]

AFAIK, normally the ECU will fire the injectors twice between each cylinders spark events - likely during intake and power strokes.

The Megasquirt allows this to be user set, and runs best on 2 squirts per spark event. You can also set whether to use alternating or simultaneous squirting, and when I try alternating, it just doesnt run well - obviously now I see because the second control pin isnt hooked up to anything, but when using simultaneous mode, its still squirting the same frequency, just all on one wire. So no extra fuel. That would only be the case if both of the control wires were wired in to the tied together wiring abortion.

I think I read somewhere that the factory ECU uses simultaneous also, and the only reason for the two control pins is to split the impedance load across two FET's. Considering that the injectors in it currently are low impedance, and they are all tied together, that results in a resistance that should blow the FET's, especially with the life the car has had. The only possible out that I can think of is the existence of a hidden resistor in the loom somewhere. Im yet to find it, and running out of places it could be hiding. Its also possible that the ECU itself has been modified with extra resistance. Actually, I wonder if that could explain the weird injector power wire splitting off to the ECU....

[Vertigo]

Dug deeper, near the ECU. Found the both injector drivers are being used after all - except tied together in the same way the injectors all are.

Heres a diagram:

Code: Select all

                             /-- Inj 1
                            /
Pin 10 --\                 / \-- Inj 2
          >---------------<
Pin 20 --/                 \ /-- Inj 3
                            \
                             \-- Inj 4

Yep - single wire for everything. What in the hell. Somewhat explains why the drivers arent exploding, as they are sharing the load evenly. I still suspect there may be a resistor in play somewhere, or the power wire splitting off to the ECU somehow reducing current.

[whynot]

I have come across this wiring in the older aw11's before, they use multi point injection but batch fire the whole lot at once. No idea why but I think it is the motors that have the 4 tooth. Crank sensor in the dizzy which have this injector configuration.

[Vertigo]

Very interesting - thanks for that! I have that exact distributor. So this really might be factory wiring, after all. How bizarre!

Ive put my 440cc injectors in now, and the Megasquirt runs the engine fine, though a little rich, even with the VE at zero around idle. I measured the resistance across all injectors through the wiring, and found 1.8ohms. Would adding a resistor help reduce the fuel flow through the injectors at all? Id like to be able to raise the VE values a bit.

[whynot]

Not sure which megasquirt you have but I thought they did need resistors on the injectors. Would pay to check, your one may have peak and hold injector drivers which means you can run low impedance injectors.

Does the megasquirt have settings for minimum effective pulse width? If that is set too high then you won't be able to lower the fuel rate enough to get your desired afr's either that or try adjusting your fuel master settings if you can't get a decent afr at idle.

[jondee86]

Download this and have a read...
https://github.com/sparkiedk/Toyota-PCM-hacking/raw/master/Lifting%20The%20Lid%20on%20the%20mk1%20MR2%20ECU%20(Jeremy%20Ross).pdf

It might help you figure out the AW11 injector wiring

Cheers... jondee86

[matt dunn]

I have seen inside some toyota ECU's where #10 and #20 are actually linked inside the ecu too.

Was probably an early 4age ECU as I have not had many other ones apart.


Also if you remove the heatshrink and there is solder there,
it is not factory, as all factory joins will be crimped only.

[Vertigo]

Good info, thanks guys.

I think I will wire in the resistor and leave the rest of the wiring as is. If it doesnt run well, Ill just remove it.

I have installed my 440cc injectors now, and in the process of getting the idle right with the Megasquirt. It wants to run really rich, even with the fuel turned all the way down. I suspect my earlier assumption that it runs best on 2-squirts/cycle was a simplification. Testing it earlier this morning resulted in the AFR heading up to 16-17:1, rather than 10-11:1 on 2-squirt. It may well run well on 1-squirt if I tweak the fuel parameters enough.

[Vertigo]

1-squirt works perfectly - AFR back to normal, on the 440cc's.

Ill leave wiring the resistor box in for a day when I have nothing else to do. Ta for your help guys.