injector timing question

Matt82 · 7 Sep 2008 at 23:15

its a 95 mk3 gti, it has motronic digi3.2 management. i can plug it up to a laptop and check certain parameters.

i did this the other day as mpg and power didnt seem right

basically when totally off the throttle, but in gear, i swear in the past it used to show 0ms for injector timing. today it was showing 1.53ms

anyone know if it should go down to 0ms, or am i remembering it badly?

i found a vac leak earlier, its not fixed the injector issue though

at idle injector timing is 3.06ms. lambda is .7v

off throttle, in gear, revs up at 3krpm or so, the lowest i see is 1.53ms injector timing and lambda goes to almost 0v

flat out, the most i see is 16ms injector timing lambda still shows .85v though. is that possible??

so lambda says the engine is ok, maybe a touch rich at idle, but the injectors are now showing what i want to see

Dogbreath · 7 Sep 2008 at 23:19

If the RPM is high enough on the over-run then it should cut fuel (i.e. 0ms injection duration), but below that threshold it'll start injecting again. IF this isn't happening it's most likely a throttle position sensor problem.

Do you see the lambda voltage cycling up and down under light throttle cruise conditions (with engine warm)?

Simon · 7 Sep 2008 at 23:22

it should go to 0ms if fuel cut on overun is turned on.

Matt82 · 8 Sep 2008 at 06:34

throttle position sensor works. 0* when closed and it maxs at 74* (weird figure, but they all do)

there is a "closed throttle switch". mine is connected and functioning. you can test it on a dmm, also if you hold the switch closed and open the throttle by hand, the revs cycle up and down as the ignition is cut

even if it was a closed throttle issue, why am i only getting 15ms on the injectors when flat out (ie 6krpm in 2nd or 3rd)?

Matt82 · 8 Sep 2008 at 06:39

as for lambda, im getting about 0.7v (idle, which im sure it didnt used to be that rich)

on the over run it goes down to almost 0.0v. flat out it goes to .8v (how can that be possible with the injectors barely doing anything ?

Simon · 8 Sep 2008 at 09:33

Injectors 'doing anything' is proportional to airflow.

Are you using a narrowband sensor? Get a wideband on there

Dogbreath · 8 Sep 2008 at 09:38

Matt82 said:
even if it was a closed throttle issue, why am i only getting 15ms on the injectors when flat out (ie 6krpm in 2nd or 3rd)?

At 6000 RPM, one cylinder fires every 5 milliseconds!

Matt82 said:
as for lambda, im getting about 0.7v (idle, which im sure it didnt used to be that rich)

on the over run it goes down to almost 0.0v. flat out it goes to .8v (how can that be possible with the injectors barely doing anything ?

As Simon has mentioned, the absolute voltage on a narrowband sensor is pretty useless information. The important thing is that the lambda voltage should be cylcing up and down every 1-2 seconds under light throttle is it's working properly. At *** the ECU typically goes into open loop, and the lambda signal isn't used.

Matt82 · 8 Sep 2008 at 09:56

ive not checked to see whether the lambda cycles. ill be able to check that today. tbh i dont remember seeing it cycle up and down.

if it isnt, is this an issue with the probe not detecting the cycling or the ecu not instructing the injectors to cycle?

as for fuel vs airflow, mine operates on a map sensor. i think there could be a chance the hose hose melted/collapsed (new silicon item last year but it does pass above the exhaust so its possible)

so 15ms is a plausible timing figure whilst flat out? its 3ms at idle, im sure it goes up to 15ms quite sharpish (by about 4krpm) and then just stays there

Simon · 8 Sep 2008 at 10:30

At 6000rpm you only have 5ms for injection, after that the next cycle starts.

Even if you were hosing fuel in all of the time at 100% duty you would only have 10ms.

Dogbreath · 8 Sep 2008 at 10:33

Matt82 said:
ive not checked to see whether the lambda cycles. ill be able to check that today. tbh i dont remember seeing it cycle up and down.

if it isnt, is this an issue with the probe not detecting the cycling or the ecu not instructing the injectors to cycle?

A dead oxygen sensor is the usual cause of this.

Matt82 said:
as for fuel vs airflow, mine operates on a map sensor. i think there could be a chance the hose hose melted/collapsed (new silicon item last year but it does pass above the exhaust so its possible)

so 15ms is a plausible timing figure whilst flat out? its 3ms at idle, im sure it goes up to 15ms quite sharpish (by about 4krpm) and then just stays there

Yes, that sounds plausible. You are probably hitting peak torque at around 4k which is where the engine requires the most fuel (per firing cycle).

The parameter that's more interesting is duty cycle, i.e. the ratio of on time to off time. If this ever approaches 100% then the injectors aren't capable of delivering enough fuel for the engines requirements. Typically on a standard production car it should go to no more than 85%.

Matt82 · 8 Sep 2008 at 18:21

i dont get to see duty cycle. im going to try tonight to see if i see the lambda cycling, if not, ill have a look into changing it (might even check its still connected up first!