Offroad-Express

As the title implies this is about the later 'electronic' injector pumps that are commonly fitted to TD27's in Nissan Mistrals, Pathfinders and Terranos later than 1995 or so.

I use the 'electronic' term somewhat lightly since in fact they turn out to be electro-mechanical and from what I can see there's no electronics in the pump assembly whatsoever.

It turns out that these later pumps are, arguably, not as reliable as the purely mechanical pumps fitted to earlier engines - their longevity is renowned whereas the electronic pumps appear to fail at between 100-200,000km according to various sources.

The failure mode is commonly reported as surging or intermittent throttle operation and occasional cutting out amongst other things.

If you have the misfortune to have such a problem the typically quoted repair work involves a complete pump overhaul with replacement of the 'electronic' unit for a total cost today in excess of $2000.

At that sort of money in some cases it can be nearly half the value of the vehicle and may lead to it being written off, or the wallet being somewhat lighter than one would like that week.

This problem occurred to me through a R50 Terrano I purchased with a cooked motor. Since it wasn't working at the time of purchase I didn't know there was an issue until I put another motor in and fitted up the original electronic pump. In my case it was almost impossible to accelerate away from stop smoothly, often the vehicle would take off as if one had stamped hard on the accelerator and when slowing for intersections the motor would cut out completely on occasions.

Now I hadn't paid a lot for the machine and I was damned if I was going to spend in excess of $2k on a recon'd pump, and if I did I wanted to do the work myself.

In discussion with a local specialist he commented that they always replaced the electric module in the pumps as they were generally the problem. IIRC the cost of the unit itself was circa $1100 or so, a significant component of the overall rebuild price. I got the impression that a few people backed out at that and he would be interested in an alternative solution...

So in my case I decided I would install a mechanical injector pump (since the TD27 used one up until the mid-90's) and I had one on hand. This was not without its own issues as I have recounted elsewhere and will perhaps expand on sometime for those that are interested but this thread is intended to shed some light on the Covec pump 'cos I'm not convinced that replacing the electric module is needed (at least in this case).

I preface the rest of this by saying that I've not seen a Covec manual and have no idea how the pump is intended to work, I've surmised quite a lot and present what I've discovered in the form of notes that may be of use to others if they're inclined to take a look at their electronic pump....

Covec injector pump operation.

This pump appears to be an electro-mechanical pump with the main lower body of the pump assembly being entirely mechanical and the upper squarish cover assembly being the 'electronic' section.

The upper section splits off from the lower mechanical unit fairly easily with 4 or 5 screws. I found I had to use an impact driver to loosen them sufficiently to split the assemblies but once I'd done that they came apart fairly readily.

The upper section consists of three coils sited within a tiered assembly contained within a cast aluminium body. It would be possible to remove the electrics from the body easily enough but the wires that lead out to the plug assembly are spot welded to the electric assembly and would need to be broken off. This suggests repair is perhaps not intended and the unit may be considered a throwaway.

Travelling through the assembly out through the bottom of the casting is a shaft that has a peg located at the lower end. This peg fits into a sliding ring or sleeve of the mechanical section of the pump and clearly provides the required mechanism for the increased pump stroke and ultimately variation of the motor speed to occur.

At the top of the electric assembly there is a semi-circular half-moon twin laminated iron core that contains two shunts or chokes around each twin half-moon (I'll sort some pictures out so this makes sense!). One of these chokes is fixed and from the look of it is adjused by 'tweaking' when the pump is set up, the other is rotated around the half-moon by action of the shaft running through the whole unit.

At the inner end/core of the two half-moons there is a small coil that is fitted around each core. These coils have a specific inductance that in the case of the core with the moveable choke is variable. In measuring the inductance I found this coil to have an intial inductance of around 3.5mH with a change in the range of 0.6mH when the choke was moved to its extremities.

The shaft containing the moveable choke and the peg at the bottom is rotated by, I guess, a moving iron action due to application of power to the third quite large coil that is sandwiched within the tiers of the assembly. At this stage I've not dismantled the unit further (I've no need) so the exact mechanism isn't certain but that's how it moves.

Once connected to the vehicle it is operated through a PWM signal from the ECU which on idle seems to be just a few hundred hz (judging by the sound, I didn't acutally measure it).

So to re-cap the PWM signal is varied, the shaft moves, the peg and the choke move, the sleeve is shifted and the motor speed varies and the inductance of the small coil(s) varies according to the shaft position...

...contd

As the unit stands it is quite sturdy and operation of the pump along with information as to the precised location of the throttle shaft/sleeve is achieved without the use of any moving contacts (such as in a potentiometer) that could cause problems over time.

Incidentaly I think the precise location of the shaft is achieved due to the fixed choke that is initially set - the inductance on this half-moon is relatively fixed whereas the other is variable but the two can be compared (without going into to it too deeply!) to give a reasonably precise location of the moveable choke along the other half-moon laminated core - and therefore the shaft and sleeve location.

So we know where the shaft is and ultimately how much fuel the pump should be metering out from this information. Coupled with various other inputs from the motor the ECU makes use of this data as required and can independantly vary the PWM signal to the actuator coil, moving the shaft, as needed.

In the case of my electric assy I really couldn't see anything at fault with it, the only criticism might have been slight binding of the shaft although that was being very critical. On the whole it appeared to me to be something that either would work or it wouldn't - an 'in between' or intermittent mode of operation seems unlikely. The unit generally seems fairly robust and when connected up to the ECU moved around as one might expect via operation of the throttle position sensor (TPS) which is essentially just a dual-ganged potentiometer, or as determined from the ECU by various other inputs when the motor is running (bearing in mind it now has a purely mechanical pump).

That said a jerky movement of the shaft is occasionally noted which might cause the problem originally exhibited, albeit the reason for this is currently unknown. My present thought is that in fact, despite closely inspecting and measuring it, the TPS may be faulty (another not uncommon problem) or the shaft may indeed be binding enough to cause erratic movement and overcorrection by the ECU.

If it's the latter I should think the solution to be fairly trivial since it appears the shaft could be easily removed and cleaned/lubricated or otherwise replaced if needed. Whether this would be enough to fix the original problem is unknown but if I had known at the time what I now know I think I'd have tried that - hence writing this lengthy missive to assist others in case they're faced with the same issue.

It is possible also that the feedback information provided by the small coils around the laminated half-moon core may be problematic in some way.

If I came across a vehicle with this problem in the future I should think that a simple PWM controller made from a single 555 timer and potentiometer could be directly applied to the injector pump actuator coil in order to vary the throttle shaft without any other input. This should in my view work well enough to be able to determine if the pump itself is operating ok (ie. separating off the real electronics in the form of the ECU, TPS and other sensors) or not. If it worked fine then the problem is either the feedback mechanism, TPS, ECU or other sensor (or indeed the wiring harness) but if not then the pump would need to be looked at more closely. This may still just be something like a sticky shaft but that's indeterminate at this stage.

So that about covers my musings to date. It's possible someone else will have the manual, or be very familiar with these things and be able to shoot my theory down in flames but in the meantime from what I can tell this is more info than is available anywhere else on the 'net. Hopefully it's of some use to someone in the future!