Have you measured when the injectors...

[djuosnteisn]

The fuel system (injectors) is not adequately sized for producing more than around 350-380whp. At that point you're also looking at very poor combustion and efficiency in the upper RPMs.


Here is a summary of this thread

We have measured the injectors for the mzr disi: This thread is littered with videos and pics, and the first 3 pages are definitely a good read if you bored, and a must read if your going to be measuring the injectors yourself.
Some facts:

We can update the pinouts for the MS3 if anyone wants to provide me with that data.
Pinouts on the ecu for a MS6 (you can use the wot box install guides for pinout locations)
Ignition coils:
cyl 1 - 3W
cyl 2 - 3X
cyl 3 - 3Z
cyl 4 - 3AA

Crank sensor - 2C

Cam shaft sensor (i didn't use it) - 2E

ECU injector control line to injector driver
cyl 1 - 3B
cyl 2 & cyl 3 - 3E & 3H (haven't confirmed exactly which is which)
cyl 4 - 3K

Injector wire colors on the injector plug i keep probing in the various pics and vids
cyl 1 - red
cyl 2 - brn/wht
cyl 3 - grn/wht
cyl 4 - blue


The crank:
Confirmed it is a 60-2 wheel. So 60 teeth with 2 missing. (The regular mazda 3 duratec has a different crank).
See the picture below. The engine rotates clockwise. So that means the missing tooth/gap is 120 degrees BEFORE TDC on cylinder 1 compression.


Firing order is 1 3 4 2.

The 4 injectors are controlled by both a + & - control rails. The - controls are specific to each cylinder. The + controls are shared between cylinders 1-4, and 2-3.


The injectors are likely controlled with logic like this:


If you monitor one of the injector lines (not ecu control line, but direct line to injector) you will see both the actual opening event, and also the + switching between 1-4 & 2-3 connection. The injector is not spraying twice though.


Spark:

The spark actually happens at the end of the pink trace (falling edge). When the signal is high, it is charging the coil. When the edge falls, the spark happens. I counted about 18 crank pulses from the missing tooth to the end of the spark event in this picture. Every crank pulse is 6 degrees. So 6*18 = 108 degrees. Since spark happens at 10 degrees before TDC, that means it is 108+10 = around 118 degrees from the missing tooth to TDC. Which is correct. Nice capture!

During shifts, the injector may sometimes spray after the spark event.


HERE'S THE JUICE OF THE THREAD
When fairly bolted, and near redline, the injection cycle will extend all the way through the intake stroke and into the compression stroke, and depending on the fueling requirements, it'll get damn close to the spark event. This has some serious implications:
- Fuel injected from the injector depends not only on the fuel rail pressure, but the differential pressure between the fuel rail and the combustion chamber. As you start spraying further and further into the compression stroke, the injector requires more and more time to spray a given amount of fuel. You can almost think of this as a redline for our injectors, higher rpm's may not result in higher power beyond a certain point.
- For proper fuel atomization, time & fresh oxygen is required. We have a swirl type injector injecting at around 12MPa in a wall guided system. This is pretty much AS BASIC as it gets with direct injection.

I have an SAE paper in front of me quoting 3ms for complete vaporization. At 6000RPM 3ms = around 110 crank degrees. If we are still spraying let's say even 40 degrees BTDC, that means unvaporized fuel will exist well past 70 degrees after TDC and by that point it's too late to get power out of it. We get soot.

- Mazda's Direct injection is limited to a maximum of 50% duty cycle, consisting of the intake stroke and compression stroke.


Potential fixes for a maxxed out DI fuel system:
1) Bigger injectors
2) Secondary fueling
3) Increasing fuel pressure



-----------------Here's the original thread, enjoy!--------------------




Okay. So today i attempted to measure when the injectors open. I still have no clue when they open.

Time for lex and 06speed6 to chime in. I'll leave all my wires in place so i can continue measuring any additional shit anyone wants, just need some help gaining my bearings.

From the beginning:

I wanted to measure 3 different events, Crank Shaft sensor, ECU injector ctrl line for cylinder #1, and ECU spark ctrl line for spark #1.

For the crank sensor line, i used the exact same one as used in the wot box, here's their install pic:


For the spark line, cylinder 1, i used pin 3W from this install pic for the wot box:


The injector circuitry was a bit more difficult, and i got a schematic from Jon at N2MB performance (Jon rocks the shit, good guy).


You can see the injector driver module in lower left hand corner. I assumed, that the indicated lines were control signals from teh ecu, because they map all the way back to it.




I made up some handy probe tools to stab the wires, saving me the trouble of splicing and stripping (but i do like to strip ).



I had no clue what ctrl line controlled what injector, so set up my scope and mapped them by monitoring both the actual drive circuitry with the ecu control logic. *edit* I had incorrectly mapped the ecu control signals. I had chosen cylinder 4 for 1, read below posts for clarification.
All these pictures look the same, but basically it was me matching injector driver line to ecu control line.





So thought the universe was great, and lady fate was on our side. I called my buddy adam (koukis14) to come over and video the scope for me. I didn't really take a hard look at any of the traces on the scope yet, just knew that everything was hooked up and working right, car started, scope powered with the inverter etc.

Once he got here, we started studying the traces. Shit didn't seem to "add up".

important stuff
Here's a pic of the scope trace we were staring at:


Ch 1 is the crank (yellow), and as i suspected, it has data for many many sub-degrees.
You can see the "gap" section in, which maybe corresponds to the missing tooth lex mentioned.

Ch 2 (blue) is the ecu's ctrl line for the injector on cylinder 1 (assuming the pdf i got from jon is accurate, which i'm not 100% sure since it's from a source other than mazda).

Ch 3 (pretty pink) is the ecu's ctrl line for spark for cylinder 1, base on above wot box schem.



The pic shows something that is not possible (or at most, completely illogical), here's why
-Why would spark happen before the injector?
-Why would spark happen after TDC, assuming the gap in the crank signal = TDC? That means it'd either be advanced by like 360 degrees or retarded.


So we decide to start probing the actual injector driver along with the rest of the stuff we had from the ecu, and we realized a few things.
- The actual voltage signal on the injector lines (the ones i used to correlate to ecu control logic) was happening twice as often as we thought, but it was outside the span of the scope window. So basically there was the spike like in the pics above, TWICE as often as the control signal from the ecu. I'm not sure if this has to do with the inductive load, or some circuitry in the injector driver module, but it was definitely happening twice the rate of the ecu logic.

- We also noticed that the cylinders 2, 3 & 1, 4 seemed to have injection events at nearly the same time. But maybe since the scope was picking up double the real injection event, 2 & 3 are actually 360 degrees out of phase (crank degrees), and 1 & 4 also 360 degrees out of phase (crank again), and the two groups out of phase by 90 or so. I'm not sure.

I can think of two ways the motor would run (i'm probably showing alot of noobness here), one where cylinders 1 & 4 and 2 & 3 fire together, out of phase from each other but firing in groups of 2.

or one where each cylinder fires 180 degrees (crank degrees) out of phase from all other cylinders, with only one firing at a time. If this is the case, then maybe i correlated the wrong injector to the wrong ecu event, and there should have been a correct ecu control event that makes everything else make sense.

Here's a video of the mayhem, taking forever to upload: In the vid you can see how there are 2 injector events for every 1 pulse from the ecu ctrl line, and how things would work beautifully with 2 assumptions:
1) the gap on the crank signal represents BDC instead of TDC. I have no clue on this
2) I had incorrectly mapped the wrong ecu ctrl line to real life injector pulse because there are 2 pulses for each ecu event, so you can assume the odd pulse is the legit one. Also if you go back and look at the 4 pics where i tried to match ecu logic with injection pulse, there are overlaps of the curves, with one jagged looking blue curve and both a flat and notched yellow curve. This was the double pulse event, but i didn't recognize it.

SANY0038.flv video by djuosnteisn - Photobucket

SANY0038.flv video by djuosnteisn - Photobucket

If anybody has actually read this far, then i know they care, lol, and i'll do whatever i can to fucking measure this right, but i need some directions from the seasoned folks here.
Take your time, digest what i posted, and ask any questions you want. I'm going to leave all the shit hooked up till we get the right measurements, and then i'm gonna go rip around like samsel with a scope on my forehead, lol.



Right now i'm going to go out and solder what i think is ecu control line #4, cause if i mapped them wrong originally, this would be the correct one. If it works, and shit looks dandy, i'll let everyone now and clean up this fail ridden OP.