Fuel Injected fun...

P

PeterZRH

Well-Known Member
OK so a 30kg parcel arrived containing the complete fuel injection system (loom, manifolds, sensors, plenums, injectors, ECU) from a 1997 3.9 Discovery based around the ubiquitous 14CUX. Seems unbelievable this lot was barely £150 (if it works - the ECU I know does for certain).

Anyway my plan is to go this route very nicely detailed here: http://www.britishv8.org/articles/rover ... re-efi.htm and run it is closed-loop mode for efficiency and the fact it effectively auto-tunes.

I have some question that hopefully I can answer before without gaining it through bitter experience.

1. Does the plenum fit under an S1 bonnet without any height reduction? It looks like 0.8" can be had without much pain. What is the penalty for shortening the effective length of the inlets? (if any - have a feeling it might lower torque ).
2. The fuel pressure isn't very high in this system so the Jaguar electric pump is a fairly painless solution. However we have our reserve tanks which are switched with a valve. Is the very simplest solution of simply bringing both tanks together (doing away with the reserve but keeping the overall capacity) and the pick-up and return lines at a single point directly under the tank workable or do we really need a dedicated return into the tank?
3. What is the thread size for the lambdas? I understand they should be as close to the manifold in the downpipes as possible but how deeply should the be positioned into the exhaust stream?
4. Recommendations for airbox/filter (standard LR can?). Want quiet. I don't like the breathy sound from K&N and the likes.
5. I'm doing this mainly for efficiency and my own curiosity as I?m an electrical engineer and this might be a stage towards megasquirt/megajolt as the basic hardware is fairly sound for that application also. What sort of output/torque on a 3.5 engine with 10:5:1 engine should I expect? Would the compression and the superior EFi system crack the 200bhp barrier given the Vitesse was 190 or 205 with the twin plenum? Would a 3-5mpg improvement be realistic? I read that the later injectors were good for an extra 1-2mpg themselves...

Anyone else done this? If so any tips/pitfalls?
 
I can't answer all your questions as I have only got as far as trial fitting a later efi on a car for fun but the manifold will easily fit under a P6 bonnet. As far as an air filter goes, most people just use a pod filter mounted behind the lh headlights. I don't think the cylindrical Landy one will fit at it's length but it may be possible to shorten it to fit.
There is just the one fuel tank in a P6, it just has pickup pipes at different heights. Easiest solution is to use the reserve line as your main feed and the main line as the return however it is often recommended to have a surge tank of some sort fitted for EFI to avoid fuel starvation when cornering. It would be relatively easy to make a small 1-2 litre tank and fit this directly below the fuel tank, gravity fed, and take your pump supply from this.
I would expect a small improvement in fuel economy but if my Range Rover is anything to go by, don't count on much. Though I've heard tales of 30mpg with a manual 'box in a P6B efi.
 
Thanks KiwiRover. I'm hoping for this and a ZF 4 speed as well - maybe the 1st 40mpg on a run P6! But I need better garage facilities to entertain this.

I'm surprised so few people have tried this before especially considering the simplicity and low cost compared to special manifolds/4 barrel carbs and the result will be inherently superior except in significantly higher performance applications.

Found a very fine write uphere also for a Land Rover: http://www.conehead.org/
 
Actually I did think of megasquirt because firstly I am an electrical engineer by training and know my way around microcontrollers pretty well. Second I work with someone who's corvette runs it - nice to have this knowledge at hand. However the more I read, the more I thought that after the initial setting up it'd get pretty old fairly quickly changing these if I changed anything else on the car.

My sentiments are similar to Cairan Brady's (http://www.conehead.org/Projects/LandRo ... i-2.15.pdf)

Another off the shelf DIY system is known as megasquirt. This consists of an ECU
designed along the lines of open-source with enough instructions / help to build a
fully working injection system using additional off the shelf sensors, injectors etc.
Megasquirt has a significant data logging capability which when coupled with a
laptop provides a tremendous degree of flexibility. It is also very popular and
6
clearly is a viable option for this engine. However, it works by referring inputs to
a fixed map of fuel requirements. That means that any change to the engine
necessitates the rebuilding of fuel maps in order to achieve proper fueling – an
aspect that appears only to be a useful asset when selling the product. To the end
user it is a potential liability.
There are advantages and disadvantages to these systems – but it is worth
factoring into the decision land rover research leading to a flexible, self adaptive
system known as the C family and which culminated in the 14CUX. This system
has two major advantages. Firstly it is self adaptive because it measures air flow
into the engine using a mass air flow sensor – using that to determine the
required amount of fuel. The air flow sensor employs a hot wire anemometer to
sense air intake – and is consequently known as a hot wire system. Within certain
limits, engine changes including displacement changes from 3.9 to 3.5 litre do not
significantly alter the 14CUX’s ability to correctly fuel the engine. Even aggressive
profile cams are well within the range available to a 14CUX – except when the
overlap becomes greater than about 12 degrees. The second advantage is that it
is a readily available given it was used extensively from 1990 to 1995 on range
rovers – many of which are now being retired and broken.
Click to expand...

So basically do I want it to fuel the car well or become an additional hobby in itself...

The best solution for the for pump and anti-surge is to combine the two into a single unit and have a fully enclosed pump in its own litre tank. Something like: http://store.034motorsport.com/fully-en ... -tank.html (I think this one runs somewhat too high a pressure). It would even keep the noise down. The price isn't even that outrageous when you think of buying, adding and plumbing both in seperately.
 
That is a nice bit of kit. It doesn't actually say what sort of pressure it runs as it doesn't include the pump for that price but my understanding is that the pressure is controlled by the regulator on the fuel rail anyway. So even if the pump is capable of 90psi, if your regulator is set to 45psi then that's what you'll get at the injectors.
That's basically what I need for my ei project but I'm torn between just buying something modern like that to see if I can make it go or trying to replicate the original setup with a Lucas PI pump and filter and a home made surge tank.
 
I've always been a great believer in 14CUX on V8's. It gets knocked by some who seem to think it's a pretty naff system that's past its best and unreliable, but personal experience tells me it's a lot better than that.

I converted several of my old V8 powered trucks to efi, including dumping a 4 barrel Edelbrock off the 3.9 in my old 90, and the nasty horrible Strombergs that were on my 101.

They were both high compression engines (9.35:1) which makes a difference to how effective it is, but both had a minimum 50% jump in fuel economy (7-8 mpg to 15-16 mpg for the 101 and similar for the 90) whilst they were just so much easier to live with, both on petrol and on gas.

Reliability wise they had a few issues, but none of it was ever down to the 14CUX hardware itself. Most of the issues with the 101 were down to the fact the original wiring was 35 years old amd well hacked about, whilst the 90 had more issues with the in tank pump.

Add in a Mark Adams chip for the ECU and you've got a pretty much ideal system for running a V8. Why bother with me gas quiet if you just want so.e thing that will run the engine.

Worth noting that efi engines have the inlet ports slightly reshaped to suit the injectors and allow them to spray properly. Easy sorted with a die grinder or a dremel if it needs doing.
 
Following with interest and reading Ciarans extensive paper and other supplied info, must say he makes it seem very viable as a conversion to a P6B.

Graeme
 
It surely is, and once you understand the basics of the system it's so fundamentally simple you wonder why people persist with carburettors, I mean when was the last time you had to balance an efi system?

To be honest, coupled up to a ZF 4HP22 (with off the shelf parts available to make it work with the 14CUX) you're left with a car that will comfortably hold its own in terms of comfort, performance, usability and reliability against machinery half it's age.

I don't doubt a more modern quad cam V8 and some wider tyres would bring it even further up to date (there's a rather nice P5B on Retrorides that's had a Toyota 1ZU-FE grafted in with some modern wheels that proves the concept) but it depends how far you want to go really.

I still can't decide if a more modern diesel -six cylinder bimmer perhaps?- or a worked over V8 would be the best thing in a P6, although I know exactly how I'd want the car to look.
 
I think the V8 is the heart of the car TBH and after running EFi and a decent gearbox with an overdrive gear I think that's about as far as I'd like to go. I'm not really feeling the need for any more BHP. I think probably dampers, springs and roll bar and thats it. After that it's too far into the "well why didn't you buy a newer and better car anyway" territory.

As for the reputation, well LR had quality issues but I think its probably more down to the crowd who bought second hand Discoverys and RRs. Although some were enthusiasts and looked after their vehicles, it's fair to say a 10 year old model wasn't really an attractive prospect to your average motorist. The kind of person who could realistically afford to run one could also most likely afford a new or newer one. I took my car for service and ahead of me once there was a P38 with 4 bald tyres, the owner cheaping-out on basic safety items. I hate to think in the wet that trying to stop behind me, 4wd won't help then.

Interesting you say about the compression... Got to wonder what it'll do on a 10.5:1....
 
Cant help but wonder why in his conversion he just didn't transplant a complete 3.9 14CUX engine into his engine bay, it would have been a lot less process (and less reading for me :LOL: ).
One advantage to me in reading his paper is that it has broadened my understanding of injection systems and engine processes such as breathing, emissions, and control.

Graeme
 
Lambda sensors are usually M18x1.5, weld on bosses are readily available on ebay. I think somewhere between 18" to 24" is the distance from the heads, just need to confirm with another source but somewhere at the bottom of the downpipes near the back of the engine would be my guess.

I doubt you'd be seeing 190-200bhp just from dropping the hotwire on, they were rated at about 160bhp in the Discovery 3.5efi and about 180bhp with the 3.9.

When I was researching retro-fitting the hotwire system I found the wiring diagram colour codes are wrong for the later wiring looms, and read it's recommended having all the extras like air-con and heated screen pick-up wires run through resistors to tell the ecu they aren't working or not fitted.

Getting it running was easy, just find the main power wires, hook them up and turn it over. I had the advantage of having the right type of distributer (I have an SD1) to work with the efi, but years ago I was running the older efi from a Range Rover from a Lumenition optronic kit on a converted points dizzy.
 
Worth noting that 14CUX systems won't work properly with your cheap 'generic' lambda sensors, and does need to be fitted up with the correct ones for the system.

I forget the exact reasoning, but it's to do with the working range of the sensor being different enough as it would make incorrect fuelling adjustments with generic sensors.
That came directly from Mark Adams (14CUX Tuning God) so I'd believe it to be genuine.

As for power outputs, fair enough a fuel injected 3.5 was only 160bhp in a Discovery, but you need to remember that's on a 9.35:1 engine with a different cam and setup for more bottom end torque than outright power.

Cam profile, valve and ignition timing, compression ratio and flywheel weight are just some of the variables that make a big difference to engine outputs.
A Land Rover 3.5 is actually fairly easily coaxed up to 200bhp with a 4.2 camshaft timed correctly and a little massaging of the ports.
 
No it won't work AT ALL - they are the "titania" type. Probably it'll fault and run open-loop if fitted. Guess what? These are the more expensive type and I haven't got the sodding things.... Was hoping to bring the project in around £400 which is what a new 4 barrel carb would cost on a good day (and probably the same for a manifold). They actually read almost opposite to the other type. This is covered in the excellent write-up I linked. They are better than the other type.

I won't get 200bhp as I'm going with the 3.9cam, possibly SD1 heads (which I have ready for a rebuild). As I've got an auto, the top end maximum isn't going to be very important anyway. I'm guessing 15% more torque where is counts with the 3.9 cam plus EFi which along with better consumption is my main goal. Might be enough to give my auto P6 about the same sprint times as the manual - probably more a function of simply being "in tune" all the time than anything else - if you look at the press reviews back in the day the measured times vary wildly - a "hot" auto might crack 10 seconds 0-60 others as high as 11.5.

My questions were more directed because it's a 10.5:1 engine and no engine of this CR ever ran fuel injection from the factory. In carb form it's either 184 or 164 bhp depending on what you read (is the higher figure some ridiculous SAE measure?).

CR is very important to these engines after all Land rovers came with as little as 90bhp and to match the power outputs of the original 10.5:1 engine it took new cylinder heads, fuel injection and the best part of 25 years. CR is THAT important, but I do worry as there is no knock sensor in this set-up so I guess I'm still needing 99/100 octane. (modern systems can run these CRs wit regular fuel)

I don't think 200bhp is that unrealistic given the Vitesse is 190bhp. The hotwire injection is better AND the CR is higher so it makes sense. But my instinct was I'd need a racier cam to unleash it but that'll work against the rest rest of the car - probably also not really what you want in your Land rovers too.

I'm not interested in a high performance machine with snarling exhausts, just something that feels super-smooth, unburstable, has no annoying flat spots/hesitation and eager even with the party-pooper BW35.
 
If we assume top end bhp to be a measure of efficiency of the engine then the figures do have a use. I'm with you on cr being crucial, I want to up the compression on my SD1 to something around 10:1 to help get a better more efficient burn, I'd estimate it would take the power output from around 160 to 170-175, anymore would be a bonus. For mine I will be converting to coilpack ignition to squeeze a little more out of it, and have no worries on what type of fuel I use as it will be mapped to suit my specific installation.

I don't know the specifics of the Vitesse system, apart from it having a completely redesigned plenum chamber yet still ran through the same airflow meter and being 9.75:1cr.

Will you be running the 3.9ecu, that may give you an edge depending on how it copes running lambdas closed loop?
 
The ECU is from a 3.9. I'm not aware they are any different from the 3.5, the greater volume simply moving to another point on the same map was what I believed. Could be wrong but reading up on the Tornado chips it would seem the map is very conservative and generic. I doubt it changed in a significant manner.

My understanding is the closed loop comes into play under 3000rpm. Although it might be the case that overall touring efficiency might not increase that much I expect more improvement around town and a more consistent average. Less of this dipping down into single figures. This seems borne-out with the land rover experiences here which won't involve much above this point. It would also mean an overdrive gear would be be a double win by lowering the revs into this band. But I'm supposing it helps most in dynamic situations to prevent overfuelling (and simultaneously improving response).
 
I bought a specific 3.5 ecu for mine, though the general internet consensus was not conclusive as to how well it would work. Also, I bought a complete 3.9 running engine for my conversion, but left the manifolds with the seller (Landy tubular style) with the lambda's still attached, maybe if I see a cheap pair at an autojumble I'll snap them up and weld some bosses into my downpipes and see if that makes any difference.

I did notice an instant change in exhaust note when fitting the hotwire, it went to a smoother lighter tone. I do have a wideband lambda in my P6 which I could swap into the SD1 to monitor the exhaust should I want to, keep an eye on how well it's fueling as that was a great help in setting up my P6 engine.
 
PeterZRH said:
In carb form it's either 184 or 164 bhp depending on what you read (is the higher figure some ridiculous SAE measure?).
Click to expand...
Bold is my change for clarity...
Brake horsepower (bhp) is the measure of an engine's horsepower before the loss in power caused by the gearbox and drive train. In Europe the DIN standard tested the engine fitted with all ancillaries and exhaust system as used in the car. The American SAE system tests without alternator, water pump, and other auxiliary components such as power steering pump, muffled exhaust system, etc. so the figures are higher than the European figures for the same engine.
Click to expand...
http://en.wikipedia.org/wiki/Horsepower
 
Maps are different between the 3.5 and 3.9 ECU's.

The 3.5 ones were specific to 3.5 engines, whilst the later 3.9 ones contained maps suitable for 3.9 and 4.2 engines, these were chosen with the 'tune resistor' which tells the engine which map to run, and if it should be looking for the lambda sensors or not.

You're quite right in thinking the stock maps are conservative, they're setup more in terms of emissions than anything else, so even a light work over will give improvements in outputs amd economy.

It's worth having a bit of a chat with Mark Adams, he knows the system immensely well, and is happy to spend his time talking to you about it and what could be done for you with no sales pressure.
Add.in that his 14CUX maps are only about £300 plus VAT it's actually pretty cheap in the grand scheme of things considering what it can do for the engine.
 
You must log in or register to reply here.
Back
Top