Non-starting Griff - how to test injector signal

[FFMan]

Making some stuttering progress with the (1994 5L) Griff that won't start.

Its got sparks, and its got fuel in the rail, so I suspect the ecu is not firing the injectors at all. Plugs are dry.

How can I see if the ecu is firing the injectors, will a moving coil voltmeter work or am I going to have to get the oscilliscope out to look for pulses ?

What tells the ecu the engine is turning, is there a crank sensor or is this driven from the dizzy ?

thanks

[cptsideways]

You need an LED bulb it'll flash with the signal, or a small wattage bulb though it won't flash just start pulsing/glowing a bit if your lucky.

[GreenV8S]

With practice you can hear the injectors firing via a stethoscope (or even just a screwdriver blade once you have got used to it), they make a characteristic tapping sound.

You can get test lamps called NOID lights which are basically LEDs that connect across the injector. You can make these up very easily yourself using an ordinary 12V LED, if you wire this up to a male and female minitimer plug you can simply plug it in between the loom and the injector to see what's happening or for a quick and dirty simply connect it in any way you like. Make sure you get the polarity right. The injector has permanent 12v supply on one side and the other side is earthed via the ECU to trigger it. You need to ensure that the +ve side of the LED goes to the +ve terminal on the injector. While you're there you can also check that you do actually have 12V getting to the injector +ve. If not you might be chasing a fuse/relay/wiring problem.

In principle you could use an analogue meter or ordinary incandescent bulb instead of the LED, but in practice those will not react fast enough to show the very short pulses that occur while starting/idling.

[FFMan]

thanks for the reply.

i did try an analogue avo meter set firstly to 12v setting, and then 1v setting but not a flicker. I will check for 12v feed first - maybe an issue there ?

I am ok with electronics and have LEDS etc to hand so will try this also - i think i'm homing in on it slowly.

will keep you posted

[GreenV8S]

Something else worth checking:

see whether the fuel pump (a) primes for a couple of seconds when you switch the ignition on and (b) runs while the engine is cranking. If it does both of these then it confirms the ECU is alive and getting a credible signal from the dizzy.

[FFMan]

at last - she runs !

Had a decent 12v on the injector but no trigger was showing using an LED and plugs were dry. Opened the bible to consult - how does the ecu know when to fire the injectors - no crank sensor, so must be dizzy. Did a continuity check on -ve output from ign amp to ecu and nothing, so traced the wire through and found a corroded connection. Cut and crimp new and hoopla - fires first turn !

Hurrah. So the big question is, was the occasional misfire I started searching for in the first place related or not - time will tell.

thanks for all help. Good job these tvrs are simple underneath it all.

[GreenV8S]

Do you still have the 8k dropping resister in the circuit?

[FFMan]

oh is that what that wire is about ? Better put it back in circuit then before I fry the ecu

thanks

Alex

[GreenV8S]

It's normally a little device with a male spade terminal on both sides that is in series with the wire somewhere in the region of the back of the near side rocker cover. Apparently on some looms it's strapped to the side of the loom by the ECU (not to be confused with the tune resister). I don't think it's catastrophic to run without it, in fact that's one "get you home" bodge if the resister breaks down, but long term I wonder whether it's doing any good for the ECU durability to run like that.

[FFMan]

I'll test the resistance of it and its around the 8k i'll put it back in place and see how we go. It was the bullet connector on one of that that was corroded badly.

thanks

[Uncle Fester]

Would it be feasible to leave an injector pulse LED connected on a permanent basis?

What I 'm thinking is a small diagnostic unit fitted somewhere. Basically a box with a few LED's.

One LED connected to a pressure sender on the fuel rail.
Another LED to show injector firing, etc.

Any suggestions for other simple lights to confirm a system is/isn't working.

I'm not trying to pinpoint faults, just let people eliminate most areas and focus on the problem system.

A quick glance would tell you which system is the problem.
and save ages trying to sort it on the roadside.

Now if the whole thing was in a neat, reasonably priced, easily fitted kit, it would probably say Mod-Wise on the label.

Fester

[spend]

I thought the resistor was normally cable tied to the induction bend between plenum & AFM on Griff/Chims.

Noid lights are handy little things for testing injector wiring BTW, can be had cheaply

Dave

[griffter]

spend said:

I thought the resistor was normally cable tied to the induction bend between plenum & AFM on Griff/Chims.

Noid lights are handy little things for testing injector wiring BTW, can be had cheaply

Dave

On my precat it's buried in the loom above the transmission tunnel.

[griffter]

Uncle Fester said:

Would it be feasible to leave an injector pulse LED connected on a permanent basis?

Fester

All you need is to pop 12v LEDs across the injector pulse signal pins of the ecu. You can access these whilst the plug is still connected to the ecu by removing the back of the plug then reconnecting it. You can then run with LEDs connected and see the injector pulse width vary as you drive. I still don't see the need to fabricate a complicated 'plug in' kit using 8x male/female junior timers across the injectors then run the whole lot into the car - UNLESS you want to monitor idividual injector signals. But remember they're bank fired, so a problem with one injector is a problem with the wiring to that injector (or possibly the injector itself), not a problem with the ecu or injector signal per se.
The ideal solution IMHO would be an extra ecu connector male/female between the loom and ecu, from which you could monitor all ecu inputs/outputs with LED/meter monitoring as required. You could even log data... But that would require buying and butchering a loom (for the plug), an ecu (for the socket) and a lot of splicing and soldering! It's easier to do tests one by one - although I still like the idea.

Edited by griffter on Friday 12th October 08:38

[GreenV8S]

All these diagnostics are best done as close to the engine as possible. NOID lights at the injectors prove that the wiring is OK (up to but not including that last connection at the injector plug). Connecting at the ECU end shows that the ECU is alive and kicking, but doesn't tell you whether that signal is reaching the injectors.

[blitzracing]

I reckon the hand held osciloscopes got to be the way forward. You can use if for injector pulses, and lambda output, and (possibly) the switching side of the coil- but depending on your plug lead types there can be 300 volt surges, that might be too much. I blew up a standard oscilocope when measuring primary voltages with copper HT leads. Anyway here it is for about £80.

http://www.maplin.co.uk/Module.aspx?ModuleNo=32611...

Mark

[GreenV8S]

blitzracing said:

I reckon the hand held osciloscopes got to be the way forward.

I've never tried one of those, are they actually any good in practice?

[shpub]

blitzracing said:

I reckon the hand held osciloscopes got to be the way forward. You can use if for injector pulses, and lambda output, and (possibly) the switching side of the coil- but depending on your plug lead types there can be 300 volt surges, that might be too much. I blew up a standard oscilocope when measuring primary voltages with copper HT leads. Anyway here it is for about £80.

http://www.maplin.co.uk/Module.aspx?ModuleNo=32611...

Mark

I've used these before and it indicated a faulty crank sensor. Used my 20 year old trusty scope and the signal was perfect. The problem is that the displays can leave much to be desired because they are pixelated as opposed to analogue.

[shpub]

griffter said:

lots of LED stuff

A word of caution here as LEDs can alter the signal seen by the injector. LEDs work by breaking down at a cedrtain voltage and this effectively means that when the LED is on it holds the injector voltage at the breakdown voltage. Normally the current limiting resistor can help isolate this effect but many injection drivers are quite intelligent and can see this as a possible fault. In some cases I have seen, the injector pulses actually change as a result. The LED acts as a voltage clamp.

The safest way of monitoring the injector is with a scope or a buffer driving the LED. The 12V led is a simple way of doing a basic check but I am not sure I would want it permanently wired in.

Also bear in mind that with some setups, the injectors are banked so that checking at the injector is not that good an indication as all it does is confirm that the ECU is firing the signal and that it is reaching the plug and not that the injector is firing in response.

The DTA system I use on the 520SE does have a diagnostic test where it can fire the injectors and ignition coils without turning the engine over. This is extremely useful and saved me no end of time when sorting out problems.

[GreenV8S]

shpub said:

this effectively means that when the LED is on it holds the injector voltage at the breakdown voltage.

Bearing in mind that the LED has to be in series with a current limiting resister, I find it hard to believe that the LED will have any impact sufficient to be measurable at the ECU, even if the ECU was smart enough to try to diagnose injector faults (which the 14CUX isn't). It certainly won't hold the injector voltage down. NOID lights are standard test tools, and I've never had of heard of any problems caused by them.