Injector wiring......
Discussion
I've got an ongoing misfiring problem with my '97 Griff. I'm beginning to suspect the injector wiring. Tie-wrap melt test suggests that all cylinders are firing but it doesn't sound like it.
Are the eight injector electrical plugs permanently connected into the engine wiring harness?
Or, is there a multiplug connection somewhere, if so, where?
If I remove the plenum to access the injectors how can I test that they are servicable?
Any help will be appreciated.
Are the eight injector electrical plugs permanently connected into the engine wiring harness?
Or, is there a multiplug connection somewhere, if so, where?
If I remove the plenum to access the injectors how can I test that they are servicable?
Any help will be appreciated.
The injectors are wired in parallel in blocks of 4 on each side of the v8, so each bank gets all 4 injectors fired at the same time and the same duration per 4 stroke cycle. loom wise, you have the mini timer connectors on the top of each injector, but no other connections until you reach the main plug on the ECU. In terms of diagnostics- invest in "noid" lights you can plug into the top of the injector. These light up as the injector fires so you can see the switching signal.
+12v to all Injectors via the ECU relay contacts. The +12v supply fed from ECU pin 15 to contacts.
Injector pulses to LH bank ECU pin 11 and RH bank from ECU pin 13 (or vice versa)
AS BR says get some Noid lights saves a lot of messing about
This disagrees with Steve heaths ECU diagram in the Bible but was how I traced it. Hopefully correct!
Injector pulses to LH bank ECU pin 11 and RH bank from ECU pin 13 (or vice versa)
AS BR says get some Noid lights saves a lot of messing about
This disagrees with Steve heaths ECU diagram in the Bible but was how I traced it. Hopefully correct!
Edited by Loubaruch on Monday 23 November 21:16
blitzracing said:
The injectors are wired in parallel in blocks of 4 on each side of the v8, so each bank gets all 4 injectors fired at the same time and the same duration per 4 stroke cycle.
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
davep said:
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
Never checked for sure - but it does not make sense as you only have one intake opening per two crank revolutions- OK- so your required amount of fuel hangs around the inlet tract until the inlet valve opens on whatever cylinder, but if you inject twice is it half the required amount? But then you could inject half the required amount and then the inlet valve opens, and the next half after it closes, which is no use to anyone. Im happy to believe that it fires alternate banks, once per 4 stroke cycle, or effectively you get one injector cycle per revolution- just on opposite banks.blitzracing said:
.... Im happy to believe that it fires alternate banks, once per 4 stroke cycle, or effectively you get one injector cycle per revolution- just on opposite banks.
You can't have it both ways if I've understood that correctly.AFAIK the firmware is coded for two odd bank and two even bank firings for each 4 stroke cycle, with one odd, one even per crank revolution. The injector bank firings are a direct function of a simple toggle on the ignition spark pulse train and the resultant alternate fuel, no-fuel interrupts. There are three firing patterns: no injector firing, double rate injector firing and normal firing (two per crank revolution). The code flow doesn't deviate from this. To achieve a once per 4 stroke cycle pattern the code would have to do a spark pulse train divide by 4 or fuel, no-fuel interrupts divide by 2, and I haven't seen any evidence of this.
Can you set up a 'scope to check the injector firing to ignition pulse timing relationship? It would help resolve this issue that's been nagging me for ages.
Apologies to the OP for hijacking the thread.
Thanks for the replies.
I shall check the ECU pins and 'clicking' at the injectors next.
I have done a very thorough check of the ignition circuit replacing all parts with either new or known
good items. Have, as suggested, tried with and without extenders.
What makes me suspect the injectors is that the ECUmate shows a lambda fault on one bank. The lambdas have been replaced but the fault code remains. Therefore I suspect that the lambda is being affected by un-burnt fuel. The engine has a regular 'beat' at idle and shunts badly up to about 3000rpm. Also, very occasionally it drops to what seems like half power as if a whole bank of injectors has dopped out. Turning the ignition on and off reverts the engine to full power but is still uneven as before.
Plugs look similar to each other and tie-wrap test melts on each exhaust outlet equally.
I shall check the ECU pins and 'clicking' at the injectors next.
I have done a very thorough check of the ignition circuit replacing all parts with either new or known
good items. Have, as suggested, tried with and without extenders.
What makes me suspect the injectors is that the ECUmate shows a lambda fault on one bank. The lambdas have been replaced but the fault code remains. Therefore I suspect that the lambda is being affected by un-burnt fuel. The engine has a regular 'beat' at idle and shunts badly up to about 3000rpm. Also, very occasionally it drops to what seems like half power as if a whole bank of injectors has dopped out. Turning the ignition on and off reverts the engine to full power but is still uneven as before.
Plugs look similar to each other and tie-wrap test melts on each exhaust outlet equally.
1) Check the plug colours. If you have a stuck or open injectors it will be very obvious buy it being pure white or very black.
2) Put a test meter on the black and white output wires of the lambda probe and looks for the physical voltages. O volts- very lean or the lambda has lost its 12 volt supply (red wire). 1.2 volts or more very rich.
3) If it shows lean, then try pulling the vac pipe off the fuel regulator (block it so air wont get into the plenum). This will boost the fuel pressure so you should get a rich signal from the lambda.
What does the ECUmate show for lambda trim? its should be cycling constantly around a mid point of 128 from memory- this will show you you at least have a lambda signal reaching the ECU.
ECU mate wont show the long term trim- but if it did you would see it keep moving until it hit 100%- and this is what triggers a lambda error after several minutes. You will get this effect if the ECU does not see any lambda voltages at all.
2) Put a test meter on the black and white output wires of the lambda probe and looks for the physical voltages. O volts- very lean or the lambda has lost its 12 volt supply (red wire). 1.2 volts or more very rich.
3) If it shows lean, then try pulling the vac pipe off the fuel regulator (block it so air wont get into the plenum). This will boost the fuel pressure so you should get a rich signal from the lambda.
What does the ECUmate show for lambda trim? its should be cycling constantly around a mid point of 128 from memory- this will show you you at least have a lambda signal reaching the ECU.
ECU mate wont show the long term trim- but if it did you would see it keep moving until it hit 100%- and this is what triggers a lambda error after several minutes. You will get this effect if the ECU does not see any lambda voltages at all.
davep said:
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
I always believed it to be this. And yes fuel gets injected onto a closed valve and just waits there until the valve opens. Its why when going to fully sequential injection you need bigger injectors to get the required amount of fuel in, in one go.carsy said:
davep said:
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
I always believed it to be this. And yes fuel gets injected onto a closed valve and just waits there until the valve opens. Its why when going to fully sequential injection you need bigger injectors to get the required amount of fuel in, in one go.I think the notion that fuel just sits static behind a closed inlet valve doesn't take into account that each inlet tract (including the chamber behind the valve) is dynamically pulsing an air/fuel charge up and down at a frequency determined by the leading and trailing edges of the opening inlet valve, as its designed to do. Even at idle each inlet tract is pulsing back and forth at about nine times a second. So with bank firing, fuel is introduced via four odd or even injectors and it's the tract air/fuel mixture pulsing that balances the charge for all four cylinders. That's just my theory and it could be complete b
ks.Edited by davep on Tuesday 24th November 21:33
Hedgehopper, i`m sure i read a post fairly recently about someone with a similar problem that turned out to be a lamba fault. I`m sure whoever it was traced the lamba wiring back and found a joint in the loom ( think in engine bay ) which had gone bad. The joint cleaned up and all was good. Maybe worth tracing said joint.
davep said:
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
Dave, What are you like? (Thats a Compliment). I bet you’ve been waiting for my pennies worth, so sorry if a little off topic.Going back to basics, for decades Dave and I have always believed the 14CUX is a digital copy of the 4CU (flapper) system. You can see from the 4CU training manual below the injectors open twice per engine cycle meaning twice per camshaft revolution (not crank) and therefore the injectors open twice per valve opening.
Furthermore earlier this year Dave and I thrashed this out in the code with Dan, I won’t go into detail here, but we did arrive at the same conclusion as the 4CU.
davep said:
I'm going to put an oscilloscope on the ECU's ignition pulse input and injector firing outputs to verify this.
Nevertheless it would be good if you can prove the theory with an oscilloscope.I’m now more curious to know if aftermarket non sequential systems like Emerald also fire twice per valve opening as they also don’t know what cylinder is firing, its set randomly at start-up.
Where is ChimpOnGas when you need him
Edited by stevesprint on Wednesday 25th November 21:58
stevesprint said:
davep said:
Isn't it each injector bank gets fired twice per 4 stroke cycle, and once per crankshaft revolution?
Dave, What are you like? (Thats a Compliment). I bet you’ve been waiting for my pennies worth, so sorry if a little off topic.Going back to basics, for decades Dave and I have always believed the 14CUX is a digital copy of the 4CU (flapper) system. You can see from the 4CU training manual below the injectors open twice per engine cycle meaning twice per camshaft revolution (not crank) and therefore the injectors open twice per valve opening.
Furthermore earlier this year Dave and I thrashed this out in the code with Dan, I won’t go into detail here, but we did arrive at the same conclusion as the 4CU.
davep said:
I'm going to put an oscilloscope on the ECU's ignition pulse input and injector firing outputs to verify this.
Nevertheless it would be good if you can prove the theory with an oscilloscope.I think the line where it states: "this means that one or the other banks is operated on every second ignition spark." should be double underlined in red.
Also, the 14CUX ECU injector firing code has no idea that things like a camshaft and valves exist, it doesn't need to. The clever people that also established the inlet tract length, firing order, crankshaft, camshaft and valve timing did though.
Once again sorry to Hedgehopper for hijacking the thread.
Edited by davep on Thursday 26th November 12:20
Loubaruch said:
Alex,
If the plugs will reach try swapping the feeds to the injector banks and see if the ECUmate still reads the same.
If the fault moves you then know that the injectors are OK but the feed/wiring is faulty
That will cause ECU mayhem- the lambda for one side will now be controlling the fuel on the other side so it wont get the correct feedback- dont go there.If the plugs will reach try swapping the feeds to the injector banks and see if the ECUmate still reads the same.
If the fault moves you then know that the injectors are OK but the feed/wiring is faulty
Gassing Station | Griffith | Top of Page | What's New | My Stuff