Injector Output Signal "inverted"?

[jmp88]

Hi all,

I've been working on this project for a long time but it's been on the back burner for months now. I never was able to get my car to start up before so now I'm starting from square 1 and testing everything with an oscilloscope before I even tough the car again. While going over everything on the JimStim I noticed something weird about the INJ1 and INJ2 outputs. The voltages seem to be backwards from what they should be - but the timing is correct. i.e. the voltage is 'high' for most of the cycle EXCEPT the 16ms per cycle where it should be high (at which time it goes down to 0V.)

Is there some obvious setting that I'm missing here? or some resistor that I forgot to attach or something? I can't really immagine where that problem would come from.

I attached a picture of the oscilloscope reading, the yellow channel is the tach input from the JimStim and the blue channel is the INJ1 output. The voltages are lined up at the bottom (0V). I already checked if the signal was just inverted but it really is 0V for 16ms per cycle and about 4V otherwise. (which brings up another question about what voltage the injectors need, is 4 enough at all? but I'll address that one later)

Thanks in advance!

Cheers,

Jonathan pH

Oscilloscope Test (Tach vs INJ1)

IMAG003.BMP (47 KiB) Viewed 477 times

[SymTech Laboratories]

The injector drivers switch the injectors' path to ground, so what you're seeing is correct. When you install your MegaSquirt ECU, the INJx wires connect to one side of the injectors, and a switched and fused +12V source go to the other side.

[jmp88]

Ah, good call! It's been so long away from the actual car that I forgot how the harness was wired up.

I guess this next question must be unrelated then, but at some RPM those ~16ms grounded intervals approach each other until the injectors are constantly open throughout the engine cycle (this happens around 4500 RPM) and shortly after that (around 5000RPM), certain aspects of the MS seem to completely stop working. The software (I'm running MegaTunix) continues registering an RPM but the pulsewidths and dutycycles immediately drop to zero. Also, the INJ 1 and INJ 2 outputs become un-grounded (go back to 4V or so) and of course the injector indicator LEDs turn off.

Any ideas what to do about this? Am I completely wrong about my injector timing? I have 19lb/hr injectors on a 4cyl/4stroke engine. Thanks again!

Cheers,

Jonathan

[SymTech Laboratories]

There's a direct relationship between engine RPM and the injector pulsewidths, so the on/off ratio per unit of time for the injectors will increase as the engine RPM increases. If the engine RPM continues to increase, the injectors will eventually reach a 100% duty cycle (in other words, they'll be held open). If this occurs before you think it should, then either your injectors are too small or your req_fuel setting isn't correct.

[jmp88]

Alright, as long as the MS is doing what it should here then it's okay for now. Although I guess I'm gonna have to upgrade those injectors eventually. 4500 RPM is enough for most daily driving but it sucks that the injectors would be the limiting factor.

Thanks for your help!

[SymTech Laboratories]

If you haven't changed any settings yet, it's more likely that the req_fuel setting is incorrect. There are many online calculators to help you determine the proper size for your fuel injectors, like this one here:

http://www.symtechlabs.com/support/know ... :injectors

[jmp88]

I think I've got req_fuel set up correctly. I used this equation:

Code: Select all

REQ_FUEL*10 = 36,000,000 * CID * AIRDEN(100kPA, 70°F)/(NCYL*AFR*INJFLOW ) * 1/DIVIDE_PULSE

from the megamanual and got 16.8ms for a 2L engine with 19lb/hr injectors. With that pulsewidth there's already a hard limit at 7143 RPM where the duty cycle would be 100%. Then, as you pointed out, the pulsewidth increases with engine speed (it looks like it goes up to about 23ms at high RPM) which puts that 100% duty cycle limit at 5217 RPM. I guess the limited resolution on my scope makes it look like that limit is earlier than it is.

Let me know if I messed something up in that req_fuel calculation. I really don't know if 16ms is a reasonable number, I'm just abusing the equation and I really have no way to reality check myself on it. Is 16ms a reasonable req_fuel time or am I an order of magnitude off or something?

It's actually really good to know that this is a logical problem to have with this setup and it's not a problem with my MS. Thanks for pointing it out to me!

Cheers,

Jonathan

[jsmcortina]

Have you tried using the ReqFuel calculator that is built into the tuning software?

James

[jmp88]

Yes, I have. It gives me the same result. The only place I could imagine messing up would be converting from liters to cubic inches, I got 2L = 122in^3. I guess the fact that nobody has scoffed at my 16ms number yet means that I probably got it right though. That's really all I need, I was just hoping to get a reality check on that number.

Thanks!

[jsmcortina]

It likely means that your injectors are too small for you engine. The ReqFuel calculator allows you to use litres and cc/min flow numbers.

James

[rickb794]

19# injectors should be fine for a NA 2.0l motor @ 45 to 55psi
Dodge even uses 19# injectors on 2.4L motors...PT Cruiser for example @ 58psi

[jmp88]

Can someone explain how the flow rate of the injectors varies with fuel pressure? How can I find out what maximum pressure my injectors are rated for and what the corresponding flow rate is? If the injectors say 19lb/hour, at what pressure is that actually true? I somehow always assumed that 19lb/hour meant that they would always output 19lb/hour but that doesn't really make sense now that I think about it, it must go up with fuel pressure.

If the req_fuel is really over 16ms then I don't see any way around that RPM limit. The full engine cycle just cannot take less than 16ms in that case (or 23 ms or whatever with the PW increase at high RPM). That seems like as big of a physical constraint as I can imagine. Do PT cruisers only rev to 5000 RPM or do they push their injectors past their flow rating?

James: I've been using MegaTunix lately and I can't find a way to input CC rather than CID there. Google is definitely invaluable for conversions like that though. Typing something like "2L in in^3" will give you the answer at the top of the search result page. I think you're right though that my injectors are just too small. I'm gonna hold on to them for now while I get the engine running but I'll definitely have to upgrade them once I get everything up and running. Realistically, 5,000 RPM is good enough for a daily driver but I'd rather not have that arbitrary limit hanging over my head.

[dontz125]

Flow through a fixed orifice (which is what your injector acts as once the pintle is fully opened) varies with the square-root of the change in pressure drop across the orifice. To double the flow, you need to quadruple the pressure differential across the orifice. MPI injectors are typically rated for 3 bard (43psid). If you want to increase your flow by 10%, the new rail pressure is 43 x (1.10 x 1.10) = 52 psi.

Note that the rail pressure must account for this dP and any intake boost - if you are running 7 psig of boost, you need 43 + 7 = 50 psig in the rail. If you want to increase your flow by 10%, the new rail pressure is 43 x (1.10 x 1.10) + 7 = 59 psig.

The boost adder is normally handled by the pressure reference tap from the fuel pressure regulator to the intake plenum (or intake runners, for MPI), and isn't anything you normally need to worry about. If you are adding a turbo to a car that does NOT have a pressure reference tap, you definitely need to worry about this or you will blow up your engine REAL quick!

[rickb794]

It will run fine on 19# injectors like I said.
Your ReqFuel is only 1 point off mine, so that is sane.

Are you sure you don't have some WE or other trim driving your PW high?
Perhaps you are over analyzing it on the stim plug that thing in and make your contribution to global warming!