3.9 14CUX EFi on a 3.5?

P

PeterZRH

Well-Known Member
I have a RR EFi system from a 3.9 Range Rover and I intend to fit it to a 3.5 10.5:1 motor.

Now from my understanding it is a MAF i.e. air flow based system which means it can adapt a fair bit to changes in tuning and setup.

So my question is are the 3.9 fuel maps OK with the move to a smaller engine or will I need to re-chip? From reading up, I've been led to understand that the 3.9 fueling is actually quite lean in the mid-range, so it isn't impossible that a smaller motor might be a better match if anything.... Or is this wishful thinking?

Opinions please gentlemen......
 
Logic would dictate that a smaller capacity engine would use less volume (mass) of air, thus requiring proportionally less fuel.
Any tuning effect would be determined by the circuitry and how much fuel it adds for a given amount of air and all the other monitored parameters.
There are re-mapped 14CUX controllers out there, so if you are struggling, it is possible to remedy.
 
Well that's the issue I'm trying to get m head around. In the case of a MAP system, then the fueling is a function of the pressure, the engine displacement and engine speed. A MAF approach would seem to account for everything except engine speed. This is bourne out by the only external connector to the EFi being to the distributor (ignoring the fast idle switch obtained from the road speed).

Now the "map" apparently is effectively a 2D map with load and engine speed. Now how is the "load" derived with MAF as the volume of air is presumably also related to RPM but the load a reflection of manifold vacuum i.e. the "P" bit of "MAP". So I guess "load" effectively comes from the throttle position sensor?

I'd love a more expert description.
 
I have had a quick google and there are a few posts from people on v8forum.co.uk and lr4x4.com where people have apparently used the 3.9 ECU on a 3.5 with success. To me it sounds like it should work on paper.

This site has quite a lot of interesting info on the 14cux system:

http://www.g33.co.uk/fuel_injection.htm
 
Yes, I read that one. I like this system because like SU carbs, they are cunning in their simplicity. Seems really bomb-proof to me. I can only consider the bad rap this system got is quality related or perhaps fear of something different from those familiar with carbs. Seems like fit and forget to me. No tuning. No rejetting and rolling road with upgrades. Better economy. More power and best of all far better drivability and throttle response than any workable carb system in road conditions.

What's not to like?
 
PeterZRH said:
A MAF approach would seem to account for everything except engine speed.
Click to expand...
Engine speed would be irrelevant because all an increase in speed does is to increase the mass of air entering the engine, leaving the map to sort out the fuelling ratio and the distributor to sort out advance etc.
 
I fitted a Hotwire efi system to my carb'd Rover SD1 with very few issues and nothing serious. Obviously the engine is a lot lower compression than yours, and I sourced a genuine 3.5 ecu for it. Not long after that I upgraded to Megasquirt knowing that the Hotwire hardware was mostly functioning well enough, however the MS tuning highlighted the injectors not giving equal fuel flow. Personally I'd chuck the 3.9 ecu on and have a wideband Lambda sensor and gauge plugged in to monitor the fueling to start with. I'm not sure if the early 3.5 ecu's understand the tune resistors for the 3.9 sized engines and different markets/specifications?
 
Hi Simon. I have the later 14CUX (AMR3242) and I have the means to reprogram the maps. It's a question of finding an appropriate starting place.
There is an 11% capacity (smaller) difference BUT there is also a similar CR difference (higher), so I'm hoping it's not massively out of whack.

I don't have enough understanding as to which factors are important in this equation.
 
The way I see it is you could spend forever asking for peoples opinions on whether it'll work or not, or just do it and make it work if it doesn't. The combination of MAF, rpm, throttle position, lambda if you have provision, etc, I'd expect it to fuel the engine to an acceptable standard. They work fine on an engine with a knackered cam which must in reality hurt the volumetric efficiency making the engine in essence a smaller capacity. If you can get a wideband lambda kit that'll tell you roughly where you are in terms of fueling almost straight away, no guessing games, just factual figures. As you're looking at re-programming anyway you'll need the additional monitoring of a wideband lambda unless you hand it over to someone else to tune it on a rolling road?

I should've tried the 3.9 ecu on my engine when I had the chance, just got over-enthusiastic and went all the way with the Megasquirt!
 
Basically mines an economy installation, so I wont be MS any time soon. I can reprogram by virtue of being an electronic engineer, so I have all the gear. It is really easy to image the chips.

As it happens the Mk1 Discovery had a 3.5 running on the 14CUX but this was 9.35:1. Would I be better starting with this or the 3.9 for a similar compression or lower? I guess higher compression in term of volume of fuel and air is effectively making the engine "larger". I know that's not the whole story but it doesn't need to be spot on for now.

The name of the game at the moment is finding a place to start and not burning my valves for being lean or fouling my plugs.

Even with the standard narrow band lambdas I think I can get a fair look at how hard the correction to the map is using "Rovergauge" software. If I can get a map that isn't working too hard up to 3400rpm then I'm guessing it isn't going to be too far off.
 
Last edited:
I'm not saying you should go to MS, in fact, there is little advantage in spending the time and money on fitting MS over a good Hotwire setup from my experience.

As you have an engine which specs do not exactly match any of the factory ecu's I'd say again, first thing to do is buy and fit a wideband lambda gauge as it will tell you how far out you are. If you can find a 3.5 ecu then that would probably be safer, though I still think the 3.9 ecu should cope happily as the 11% difference in capacity should be well within the tolerances of what the ecu can handle. I can dig out my 3.5 ecu and get the part number off the side for you later?
 
I'd agree with Sowen, fit a wideband and then you can see exactly what is going on.

I doubt that the fuelling will be out enough to cause you any damage- I've had a couple of situations over the years where due to one reason or another I've been getting very lean mixtures and never damaged anything yet *touchwood* the key is to be able to see when/if this happens and just back right off the throttle. I'll post the AFR map from my MS when I get a min to give you a guide on what seems to be ok- I've not melted anything yet!
 
OK, when I get the standard lambda bosses put in the top of the downpipes, I should get another boss in the Y-piece so I can tap in there?
 
Yes a third Lambda boss welded into the Y piece will give an average of all cylinders for a seperate wideband gauge. I managed to run my V8 very lean when the lambda probe failed whilst I was doing motorway running, no issues since I fitted a new lambda probe. They aren't as reliable as the narrowband sensors with a much shorter life, though mine had been fitted for extended periods on two previous projects so had been well and truly abused!

If you plan on doing any tuning yourself then without one you may as well put your finger in the wind ;), they aren't perfect but a lot better than nothing.
 
Good advice thanks. Extra insurance. I see your map is extended and doesn't use the standard RPM points.

I had all the necessary hardware except the serial/usb cable with the special connector in the cupboard! That's how DIY this is....

I believe you can actually to a certain extent tune the car with the standard setup. Roverguage can log the position on the map together with the lambda correction data, so you can do it up to 3400rpm at least....
 
You must log in or register to reply here.
Back
Top