Race Engine Forum UK

Are you considering modifying a fuel injected engine?

Fuel injected engines with Bosch LE, L or derivatives, OE fully mapped systems, are all 'dumb' engines. If you try tune them by means of increasing airflow by porting or cam change etc as one would with a carburetted unit they will not self-compensate for the higher airflow into the head and will in the main - not give more power without further important steps by you. Don't ever try to modify a Bosch L/Le flowmeter, even Bosch say 'don't do it', and they should know. Don't imagine for a minute that it's as simple as 'adjusting' spring tension etc etc! I have never heard of anyone successfully doing it. If you think you can get the fuelling you want on your engine by fitting an airflow sensor from an engine of higher output and bigger cubic capacity all I can say is 'good luck finding the right one!'

To adjust the fuelling (ie: more gasoline to accompany the increased airflow on an engine with comp cams or flowed head etc) things like bigger injectors (you must do the maths on flowrates and duty cycles before contemplating that and frankly it's a job for an expert) and /or higher fuel injection pressure can be needed. Doing the setup yourself by altering the standard system to see how it works after head mods needs a lots of tools plus adjustable fuel pressure regulator & wide-band Lambda system and fuel pressure gauges.

And more often than not fully mapped versions (any make of car) need more than just that and they do not have not usually have a mappable ECU which can be accessed (via laptop) to adjust fuelling and ignition to give optimum torque throughout the whole engine load/speed range, and a complete new aftermarket management system (eg: Weber, DTA, Omex) is needed, which may mean also a new wiring harness and different sensors. The trim function of an OE fully-mapped fuel injection systems will rarely be able to cope in standard form with the fuelling demand of a well-tuned motor.

Thinking it's a cheap option and buying an 'off the shelf' 'chip' will not yield the power gains you paid for. Unless it is certified in writing that the chip has been optimised on the dyno for your engine with its hardware and modifications. Which it will never be. Not if the head work and/or cam change yields a big increase (8% plus) in airflow.

Far better to have two days work on a good dyno with a very good operator to set up a modified fuel injection engine, with due attention to the accurate setting of the fuel/ignition load sites at 250 or max 500 rpm steps at throttle angles from idle to full open at say, 5 deg increments (especially if the injectors are changed). Most systems come with a 'base' map, for base, read BASIC - nowhere near optimised. I also emphasise if I have not said it often enough elsewhere on this site that the biggest handicap to more power can often be the OE inlet manifold itself, a problem often only solved on normally aspirated engines by going to independent throttle bodies.

I routinely turn away a lot of head work because of this (in as nice a way as possible, for it is abundantly clear the callers have no knowledge of this at all and they heard if from me for the first time). Most callers are mystified - or annoyed - by my outrageous suggestion that 'doing' the head work will not in itself yield any bhp gain - or worse - may damage the engine because it can go chronically lean.

Unless you a regular top ten placer at 10/10 in National level racing/rallying I would advise anyone who wants to modify their motor with FI to:

EITHER
A. Be open-minded and sanguine about what you are embarking on and wake up to the fact NOW that you are going to get a potentially large bill for the costs not only of the injection hardware/sofware but also the time and expertise of an accredited expert - like a Weber FI agent - for setting up. That total bill may well be totally justifiable by those firms but can easily double or even treble the cost of the engine mods.
OR
B. Use race carburettors


Even after setting up on FI the truth is that you may never find out how much better your engine is than it would be on, say, carbs unless you can refer to a comparative dyno test of another the same spec.
Speaking as someone who has always been very forthcoming with tuning and setting-up advice, I have little time for the apparent 'glamour' that seems to go with fuel injection. Be practical, I do this for a living. You can be sure that the glamour soon fades if/when you discover you really don't know what you're doing and the true costs of components, setting-up, advice & time are added up. Many projects just stall because of this. Don't write off carbs as 'old fashioned'. They are, considering what they cost, still the best solution for the clubman driver. Modifying FI engines and embarking on fuel injection just for the thrill of it is definitely not a good idea unless you are already au-fait with everything I have said here or are pretty expert already with race engine prep, fuel/ignition parameters, mapping software/laptop and electronics.

GC

Hi Guy,

Thanks for adding this post as it was something (naively) that I had not considdered when starting my new engine for my Peugeot 205 1.9gti.

From what I read on other forums I assumed that more power and torque could be gained from adding a big valve head from a respected engine builder that has done alot of work on the 1.9 8V.

I suppose the limiting factor for me (using std injection) is how close to fully open the air meter is on the standard engine when at full throttle. If it is close to the stops then the additional air flow will not result in any additional fuel.
If I were to up the fuel pressure to account for the limitation of the air meter then my fueling is going to be rich at lower RPM?

Have you done any head work on the 1.9 8V (or 1.6) and achieved a good gains with all other parts left as standard?

If you have not, then can anyone else (perhaps Peter in Australia) comment on this?

thank you,
Ben

I think - Bosch LU Jetronic/Motronic, you could have some work to do there. Suggest you get hold of Charles Probst's 'Bosch Fuel Injection & Management' ISBN 0 8376 0300 5 -

and have a good read.

GC

My car has LE2 Jetronic.
Flap style meter to measure air flow and no O2 sensor.
thank you,
Ben

benlilly wrote: I suppose the limiting factor for me (using std injection) is how close to fully open the air meter is on the standard engine when at full throttle. If it is close to the stops then the additional air flow will not result in any additional fuel.

Not really, but you are on the right track. The LE2 AFM output is not linear with airflow and does dont measure all the air entering the engine. Rule of thumb has the air flow meter at full deflection around the peak torque area.

At full power the ECU has no idea how much air is entering the engine. Its has been tuned blind so to speak.

benlilly wrote: If I were to up the fuel pressure to account for the limitation of the air meter then my fueling is going to be rich at lower RPM?

Yes. but if there is no O2 sensor, its likley to have a CO% screw which can be adjusted to bring the idle mixture back down. The tuning will not be perfect but maybe good enough to get away with. You need to check anything like this on a dyno with an AFR meter or wide band O2 sensor.

Some of the biggest gains in after market fuel injection is made with the ignition, not the fuel. The Peugeot 8V on LE2 Jetronic is a classic case of an injected car that still has a mechanical distributor, providing a linear advance curve. It has far too much advance in the 2000-3000 range, resulting in pinging if CR is changed.

The ideal advance curve for one of these engines (and probably many others) is far from linear and can only be obtained by mapping the ignition properly. Many dyno operators either don't have the knowledge or the time to do it thoroughly. Once a reasonable air/fuel mapping has taken place, it's then time to do the ignition. This is done by advancing/retarding the ignition at a particular load/RPM point, whilst watching the torque readout on the dyno. The ideal advance point is just before the Nm's start to tail off. Note, this is not the point of detonation or pinging. It might be 5 degrees before that!

As the Nm's increase, the AFR will start to lean out, as more fuel is used up to make the torque. Conversely, if timing is retarded, the AFR will richen. So after the ignition is optimised, the AFR's will all need checking again.

Now we've just spent quite a few hours on the dyno to get this right. A typical fee is around AUS$500. If someone tells you they can tune the advance by driving it, find another tuner!

I've achieved excellent results with the small valve 205 heads- 39.5mm/33mm. The key in my mind is seat/throat work, not bigger valves, unless you're going to race it. Correctly cut 3 angle seats with simple blending/porting yeilds impressive results. You're far better off investing in a good mappable fuel/ignition ECU, then bigger valves. Of course, people who sell/install big valves will say otherwise. I remove the AFM, replacing it with a 65mm OD straight plastic pipe. After market ECU's measure engine via manifold vacuum (MAP) or throttle position (TPS). I prefer the former with smaller cams, as you get a far greater tuneable range at each RPM range. MAP sensors are also much quicker at reacting to changes in engine load.

I have a graph of mapped 8V ignition curve somewhere. I'll try and dig it up over the weekend.

I've spent quite a bit of time playing with LE2, changing resistors etc., to alter the fueling, but I've stopped doing it now. When combined with the mechaical distributor, it's a hopeless system that should be thrown in the bin, unless you're looking for "originality is everything".

Hi Guys,

Right, got hold of Probst's book and it isn't all good news. As Steve mentioned, the air flap is fully open at around 4000rpm and after that the fueling is governed by stored values referenced to rpm.

So if the engine is breathing much easier, it is likely to go lean above 4000rpm.

Probst recommends upping the fuel pressure to sort the higher rpm fueling and then increasing flap spring tension to bring the lower rpm fueling back down.

Anyone tried this?

Anyway, I'll be saving for an ECU so I can get it all singing.

Peter, any chance of that ignition curve?

thank you,

Ben

Hi Ben

I have done part of what you write about on a Bosch L Jetronic Fiat 124 Spider 2 liter. I'd gas flowed, used std cams and ran 10/1 CR.

I expected the power we got, rolling road result obtained later, 135bhp at 6100rpm, about 118 standard or so, but knew that we would not get there without adjusting the fuelling.

I installed a fuel pressure gauge in the fuel injection rail feed hose and read the idle fuel pressure. The I changed the pressure regulator for a Webcon rising rate adjustable unit and set the idle line pressure 3 psi higher, about +7%. I would not have been happy going more than 5 psi higher in any event, just an instinct really. I proved out the fuelling with an on-board CO analysis to achieve 3.4-4% CO under full load at speeds up to maximum (3rd, 4th gear full throttle, using a long hill to keep the load on without going too fast) by road testing. This is the Spider on p232 GC book, it says no mods. Well, beyond reseting the injection.

As for going further than that, with flap valve mods, don't know about that, and I would under no circumstances have changed the cams to billet competition ones on that setup.

GC

GC

Hi Guy,

So the Webcon pressure regulator increases fuel pressure with increasing engine speed?

Ben

It is load sensitive - that's why it has a vacuum pipe from the inlet manifold, to raise pressure according to demand. Re my adjustment, if recalibrate it for higher fuel pressure at idle it will be higher throughout the range, accepting that there is a limit to the pump delivery pressure, of course.

GC

benlilly wrote: So if the engine is breathing much easier, it is likely to go lean above 4000rpm.

Probst recommends upping the fuel pressure to sort the higher rpm fueling and then increasing flap spring tension to bring the lower rpm fueling back down.

Anyone tried this?

Yes I have tried and done this many times. It works but its crude and not as good as an aftermarket system.

The above is logical but its all based on assumptions. I've lernt that you just cannot assume these things. You need hard data. Get it on a dyno with AFR meter and find out what it is doing first, then after aswell. All cars are tuned different. Some need more or less modifications to get the mixture correct.

benlilly wrote: Peter, any chance of that ignition curve?

1500 15
2000 21
2500 26
3000 27
3500 27
4000 28
4500 30
5000 30
5500 31
6000 31
6500 31
7000 31

Modest peak advance there Peter, sign of a very good burn on that motor I guess?

GC

Thanks Peter, thats great.

What spec motor is this curve from?

Have you experimented with changing the spring rates in the 205 distributor to get closer to ideal ignition advance for a modified motor?

Do you have a stock advance curve to compare to? I imagine the stock curve is slower to advance.

Ben

that was 11.4:1, 30 thou. squish height, Grp. A cam

I haven't bothered swapping springs. The programmable ECU is too easy. I'll have look for the std. curve.

Remember they're full load settings. You'd need to add light load advance to those curves.