I have a fault on my MY94 UK Impreza that I have been trying to track down for a number of months, of which neither Barratts or Power Engineering have been able to solve (although both said they may be able to fix it if allowed to spend a few days working on it, Not at their Labour charges !!!).

Anyway after substituting Coil Packs, Plugs, Boost Control Soleneoid, Cam Sensor, Crank Sensor, Lambda Sensor and changing the Plug Gaps I have had no joy in fixing it.

The Fault mainly occurrs on hard acceleration from the line in 1st, but does occasionally happen in 2nd, and manifests itself as a large hesistaion as boost comes in (although not a severe as overboost).

However, I have recently read that the ECU changes from Open Loop to Closed Loop (or Vica Versa) at 4,700rpm.
Subsequently I have found that this is the exact point where the hesitation occurs, and suspect that it only occurs in 1st and 2nd due to the faster revs increase in these gears.

Therefore I ask the ECU Guru's on here, what the differences are in terms of sensors used during the closed loop and open loop operations, and also ask if from this analysis if a particular sensor or ECU fault can be determined.....

Thankyou....and lets hope I am finally getting to the bottom of this.

[Edited by Scott.T - 1/2/2002 5:33:49 PM]

[Edited by Scott.T - 1/2/2002 8:45:22 PM]

 

I am no ECU expert but can regurgitate things others have said quite convincingly.

The 4700rpm or thereabouts is what I think Bob Rawle said about the ECU going open loop
OR
when over a few PSI boost
Whichever is first.

So if you are flooring it you will be on boost and open loop. From your lambda display you can tell when you have gone open loop by the change from rocking around stoichiometric mixture to progressively higher voltages/richness as the boost and revs increase. Observe what it does on your own car.

Also when the hesitation occurs a key question to ask would be what the boost is doing.

If your problem occurs when on closed loop on a car of this age the number one suspect would be the lambda sensor - esp in older imports as it gets hotter as it is apparently mounted nearer the cylinders.

 

Sorry John, but the Lambda sensor was changed by PE (so have added it to my list above). This gave me trouble on MOT due to the Lambda Heater element had gone O/C, but made no difference to the hesitation problem.

It doesn't matter whether I'm running 12psi of 15psi the fault is still there. You can even provoke it with the Boost Control Solenoid disconnect (Only 7psi), but it is harder to reproduce as the revs obviously don't rise as fast.

 

The "change over" is related to boost and rpm, not sure about the 4700 rpm but that does indicate that the problem is map related, have you tried another maf sensor ? Or just remove and clean yours, the way to do it is to remove it from the airbox and uses liberal quantities of brake cleaner (Halfords finest is ok) blasted at the hot wire, you can see it if you look down onto the middle of the plastic cylinder, if you feel brave a very gentle stroke of the wire with cotton buds can also aid the cleaning process. Aside from that you have attacked all the other usual culprits. The only other thing would be to suspect the ecu.

A thought, does this happen at full throttle only i.e. if you depress the throttle to say 60% or say 80% is the effect the same?

Other fault conditions could be related to the map sensor, throttle position sensor or even the ecu grounding as if this is flakey then interferance can affect ecu function. Also its worth considering the ignition control module which is stuffed down behind the pressure sources switching solenoid and boost control solenoid, this module fires the coil packs and could be playing up.

Bob

 

Scott, a knackered MAF does certainly not always show up on the select monitor.

Could you swap one from another car for a few minutes to see if you can still reproduce the hesitation ?

 

Unfortunatly all my local collegues with Scoobies have newer models, which have a different MAF.

 

Will try cleaning the MAF with a cotton bud tommorrow and also will check out the condition and connections on the ignition module.

The TPS seems to measure a correctly with a smooth voltage transition, but then this was checked by Barratts anyway.

 

Hi Bob, thanks for the reply.

The problem does exist at part throttle although it has to be fairly wide open in order to reproduce the fault.

I have cleaned all the grounding points under the bonnet, but no difference was found. Whereabouts does the ECU ground ?

I have had a go at cleaning the MAF, but only by spraying brake fluid at it, as did not have the nerve to touch it with anything.

On an MY94, is the ignition module that fires the coil packs screwed to the drivers side inner wing, about half way down ?

Power Engineering could not see anything on the select monitor (but made me feel sick holding it whilst blasting around Uxbridge), but I suspect the sample rate of the select monitor may not of been fast enough to detect anything.

[Edited by Scott.T - 1/2/2002 9:52:52 PM]