Does anyone know a lot about ground-switched injectors?

My 3rd gen RX7 has ground-switched injectors. Does anyone know if, on cars that have ground switched injectors, it would be typical for there to be continuity from the injector plug to the chassis, on both leads, when the car is off?

If that is typical, then the ECU must have the ability to break the continuity on the ground side in order to pulse the injector.

I am asking because I suspect that I may have a bad engine harness that's causing an injector to see a continuous ground, thus holding it open. I'm trying to figure out how this is all supposed to work in order to diagnose.

thanks, James

It is pretty common to switch the injectors to ground. With the motor off, you should not have continuity to ground on the injector plug. Try it again with the harness unplugged from the ECU. With the connections pulled from the injectors and the system powered on (key on) do you have +12v on one side of the plug and zero on the other? Are any of the fuses blown? With it all hooked up and when cranking, can you hear the injectors clicking with a mechanic's stethoscope?

What problem are you having? I'm not a magic spinning triangle guy, but injectors are injectors.

JamesMcD wrote: My 3rd gen RX7 has ground-switched injectors. Does anyone know if, on cars that have ground switched injectors, it would be typical for there to be continuity from the injector plug to the chassis, on both leads, when the car is off? If that is typical, then the ECU must have the ability to break the continuity on the ground side in order to pulse the injector. I am asking because I suspect that I may have a bad engine harness that's causing an injector to see a continuous ground, thus holding it open. I'm trying to figure out how this is all supposed to work in order to diagnose. thanks, James

Ground switching is pretty standard - some Bendix ECUs in the '70s switched 12 volt power, but pretty much everything since then is ground switched. Some ECUs have a driver circuit that can go to ground when powered off; cars like this will be wired so power to the ECU and the injectors is cut at the same time. If you suspect the ECU may ground the injector wires when powered off, just unplug the ECU. If the ground side of the wire is still grounded, your harness is damaged.

What it's doing is, I believe, dumping fuel. It's making a non-coolant smoke screen. Engine is new so I don't think it's oil. Injectors have been cleaned. I'm trying to figure out if it's the harness or the ECU that is causing an injector to hang open, if that's what's going on...

Hmmm. I've learned a bunch about 2nd gen FI but not so much on the 3rds. As others have said, the injectors will run on a +12V on one side, then the ECU pulses the ground to control the opening time and thus the fuel/air ratio. Honestly, if it runs well but just smokes it's probably not the ECM or injector harness.

If you unplug the ECU harness on a 2nd gen, there should be no continuity to ground when testing the injector harness. Again I am not positive about 3rd gens, but on 2nd gens the injectors get their 12V from the ECM. So, on a 2nd gen, with the ECM harness unplugged there should be no continuity to ground on either side of the injector connector. If the injectors on a 3rd gen get their power from somewhere other than the ECM, it's possible for the positive side of the connector to have continuity to ground when testing with a meter even with the ECM unplugged. So if your harness passes this test you need to see if the ECM is doing its job.

To see if the ECU is pulsing the ground, get a 12V LED, unplug a primary injector (that's the ones under the intake) and connect the LED across the harness plug terminals. (I wouldn't recommend using a light bulb due to the possibility of damaging the ECM with too much load.) Then spin the engine over (of course the ECM harness has to be connected for this to work!) and you should see the LED either pulse or flash. No light at all probably means the LED is connected backwards, so switch the LED connectors and try it again. Constant brightness while the engine is spinning over could indicate a problem in the ECU. You can also do this by sticking a stethoscope on the primary injectors and having a buddy spin the motor over, clicking = OK, no click = possibly bad. Again on 2nd gens, the secondary injectors do not open below 3500 RPM so they don't click or show a pulse with the LED till then.

What color is the smoke? Black which hangs in the air = fuel, white which dissipates quickly = coolant, bluish white which hangs in the air = oil, grayish white which hangs in the air = ATF.

IIRC there are coolant passages in a 3rd gen intake, if those are leaking (such as between the intake gasket and the side of the motor) it's very possible to have it steam like crazy yet not have the other coolant leak symptoms such as overheating etc.

Thanks, that is excellent information.

The car seems to idle pretty decently. I don't know if that is possible with a stuck open injector. Perhaps I have a bad oil control ring after all. Fooey.

That was my original diagnosis, but a guy on the RX7 forum had smoke that looked identical to mine (as seen in his Youtube vids) and he fixed it when he had his injectors cleaned out.

This car is very frustrating. I've been building it for 7 years and have never driven it.

I stayed in a Holiday Inn Express last night. Plus, I managed to recently rebuild a 13B NA and (wonder of wonders) it runs.

What I learned about rotaries during my rebuild: the oil control rings themselves generally do not wear, but the O rings behind them get worn and/or hard. Again on 2nd gens, that's the 'thermal pellet' which is usually at fault. That damn thing is an oil thermostat which shuts off the oil sprayers behind the rotors during warmup. They fail in the closed position, causing the rotors to absorb a lot of combustion heat which makes the O rings brittle. Atkins and some others have a bypass to eliminate that problem.

It's also possible to put the springs in backwards. If that happens, the oil control ring can spin on the rotor and that will eat the O rings up very quickly. Either way, the engine will run fine but smoke like there ain't no tomorrow.

On 2nd gen turbos and 3rd gens, the turbos use engine oil to lubricate the shafts (duh) and if the shaft seals are bad the same thing will happen, it smokes like hell but still makes good power. It's pretty easy to see whether it's the oil rings or the turbo: first, to cut down on disassembly see if you can borrow a borescope. If it's the oil rings there will be very little oil in the intake manifold because the oil will be coming from the rings past the side seals then straight out the exhaust. If it's the turbo(s) leaking, everything in the intake will be wet because it will come from the turbo seals on the suction (vacuum) side and get sucked into the intake manifold. That would be most evident just past the vanes inside the 'snail' on the intake manifold side. That's called the 'compressor cover' in this pic.

I'd be more inclined to suspect the turbo(s) if the motor's new.

Again, this is all good info. But I am not good with electricity.

The turbos I'm using are low mileage ones bought from a reputable guy on the Rx7 forum, and my intake tract is dry as a bone, so I'm discounting it being the turbos. That was one of the first things I checked.

On the 3rd gen, the ECU pulses the ground but the injectors get their power from the EGI main relay.

I checked continuities again today and I am getting continuity on ALL injector leads, both + & -, to the chassis. This was with the battery disconnected and ECU plugged in. I can see this as being plausibly correct, although I'm not sure I understand why I get continuity on the + side to the chassis.

I also checked resistance on all injectors leads, from the connector to the chassis. They all read as being the same value. I found that to be odd, because the ground sides go to the ECU, and the voltage sides, again, go elsewhere to the other side of the engine compartment. So it seems weird to me that resistance would be the same on both the + & - leads.

I am more confused now than when I started.

Hmmm. Are you getting injector pulse with the LED like I mentioned? Reason I ask, if it's pulsing then the resistance readings mean zero. FWIW, you can get 'noid' lights at Harbor Freight pretty cheap, makes it easier to test for pulse. http://www.harborfreight.com/11-piece-noid-light-and-iac-tester-set-97959.html

When checking for resistance to ground, it makes no difference whether you are checking the positive or ground side of the circuit: if there's a diode or etc somewhere in the power side of the circuit that will allow even a tiny amount of the voltmeter's power to get to ground, you will get a reading. That doesn't necessarily mean anything is wrong.

Since the 3rd gen injectors get their power from the relay, pull it and also disconnect the ECM plug (this should remove the ground side continuity) then do the test again. Unless there is something really odd about the 3rd gen harness, when that is done you should have no continuity to ground between either injector lead and the car body.

In truth, if the car runs well but smokes it's very doubtful the injectors or their circuit have anything to do with it. BTW, what color is the smoke? (helps tremendously with diagnosis)

Get a real noid light, don't use just a normal ~12V LED, since once you hit breakover voltage for the LED to light the thing will have nearly zero resistance and will burn up. Assuming it doesn't have some kind of impedance added of course. If none is listed on the package...

Just get a noid light.

there was/is a unversal noid light available if you don't want to lose the individual lights that come in the kit above. pretty cool idea imo. you may be able to rent a noid setup from the local idiotzone.

JamesMcD wrote: I also checked resistance on all injectors leads, from the connector to the chassis. They all read as being the same value.

What I said here, above, is incorrect. I think I was breathing too much brake cleaner or something.

I appreciate everyone's replies so far. Really I do. Before I go forward with the noid light I have one more thing I want to explore.

Here are the correct resistance values, these make more sense:

Now here's where I think there may be a problem. Looking at the wiring diagram, you can see that the (+) side of all 4 injectors goes to the EGI main relay, as mentioned above:

However, when I hooked up the battery and tested for voltage at the 4 injector plugs, this is what I got:

For some reason, the primaries (which the car idles on) are basically not seeing any voltage on the (+) side.

Whatever is causing this, perhaps is the reason that the resistance readings I got on the (+) side are different between the primaries and secondaries.

So I'm now thinking that my primary injector positive wires are grounding to the chassis somewhere instead of going to the main EGI relay like they are supposed to, and that is why I am seeing continuity through them but basically no voltage.

Does this make sense?

Going further. Will an injector work if plugged in backwards? If so, then it seems that the problem I just described would cause the primary injectors to pulse in a pattern that is opposite from what the ECU intends. In other words, since the voltage side is now grounded, then they will pulse when the ECU IS NOT grounding the ground side. So it's backwards, spraying when it shouldn't be, and not spraying when it should be.

What doesn't make sense to me about this theory, is that, if the (+) injector leads are shorting to the chassis, I would think that their resistance values should be closer to the (-) ones in the chart.

Anyone following this?

thanks. James

Okay. Injectors are not polarity sensitive, they'll work either way.

See the black triangle next to 11 in the diagram? Triangles indicate the direction of current flow, meaning there's got to be a diode somewhere in the circuit (probably inside the relay). It's still possible with a diode in the circuit for a DVOM to show continuity to ground. BTDT on my Nissan pickup, nearly ripped my hair out. So that's a blind alley.

Again on a 2nd gen, all 4 injectors have battery voltage with the ignition switched on. From the wiring diagram, so does your car; notice all 4 injectors share the same black/yellow wire, just like a 2nd gen. If the primary injector wiring was shorted to ground, it should be popping fuses like mad and should kill ALL the injectors, not just the primaries. So if you have NO voltage, that indicates an open- NOT a 'short'- in the harness, like a broken wire, failed splice, etc. But again if the primary injectors aren't working it won't run at idle. Remember, the secondary injectors are just along for the ride until the engine RPM and load say for them to kick in. Don't do what my dumb ass did, as in assume the injectors in the intake are the primaries (red faced) because (at least on a 2nd gen) they aren't. The primaries are the ones that mount in the center iron.

When checking for voltage on the ground side of the primary injectors, does the DVOM show a '-' sign in front of the number? If so, then it's reading a voltage 'leak' from the ECM circuit that grounds the injector. Nothing to be concerned about.

Yes, I am oriented properly regarding which are the primaries.

How does the car run yet have no voltage at the primaries on the (+) side? The only answer that can be is that the ECU is not giving the injectors voltage on the (+) side until the car is cranked over. But why would it do that? I don't know.

I guess I have to go forward with the noid lights and see what they do.

Since all 4 injectors share a common '+', then the ECU has nothing to do with it on the battery voltage side, it controls only the ground side. But if the engine runs with no voltage on the primary injectors... ya got me there, pilgrim.

Yeah you're right, that the ECU has nothing to do with it. I forgot.

When I purchased my engine harness used, off the RX7 forum, the guy told me that he had repaired a few sections and spliced parts of another harness into this one to make one good one... I didn't want to contemplate this, but my new theory is that he switched around the primaries and secondaries in the wiring. If so, perhaps my car is running off of the secondaries at idle, and that is why it smokes...too much fuel.

Does that sound plausible?

I guess I can use the noid light to check this. If the primaries don't flash when cranking over, but the secondaries do, then there's my issue.

If that's my problem, I wonder how the car would run if I swapped the positions of the primaries and secondaries.

Thanks for entertaining all these questions.

If it turns out the wiring for primary & secondary is mixed around I'm just going to extend the primary wiring to reach the secondaries and plug the secondary wiring into the primaries, and then everything will be where it should be and working properly.

I wish I was working on something else but I have to finish this stupid thing so I don't feel like a failure.