Car breaking

Yeah, it *sounds* easy ... ;-)

Would the injectors go into the inlet manifold? I'm assuming I'd also need a high pressure fuel pump to feed the injectors and probably a re-circulating fuel system? Crank position sensor (probably the easiest bit) and what else?

Cheers, T i m

Yep. Indirect/Port injection is by far and away the most common way of doing it, and the simplest, because it lets the valve timing sort out the fuelling timing. Direct injection is an order of magnitude more work.

Yep. Not hard. Most '80s stuff has an inline pump.

Yeh, toothed wheel that either fits to the flywheel or a crank pulley, usually 36 teeth, one missing (36-1) IIRC.

Airflow (usually part of the throttle body) Coolant temp Air temp That's about it, unless it's changed markedly since I was looking at it all a few years back.

Have a look at some of the MS websites - it's all very well explained.

You need much more precision on the crank/cam position if you're doing spark as well as fuel, but I was never planning to.

And still better than single port FI I'm guessing?

Understood. I was initially thinking diesel when you said it was easy and I was imagining some fancy spark plug / injector conversion unit. ;-)

Ok.

Understood.

Is that what they call the MAF (sensor)? I have one here off a Discovery. ;-)

Not the same thing as what currently feeds the temperature gauge?

Goes in the air intake?

What about lambda sensor(s) Adrian? Isn't that part of the feedback loop re potential over-fueling or just used if you are running a cat to stop it getting polluted?

Ok, I will do.

Well, I think I've been suffering a bit of timing scatter because the dizzy bearings are a bit worn and the opto-chopper on the Luminition doesn't apply any side loads to possibly steady the shaft like the points did.

Do they do a kitcar-friendly coil pack (or does a production one off something else make a good donor) as I'm guessing that is what I would need if going MS/ECU driven spark?

Do you still have vacuum retard using the MS/ECU or is the ignition timing under load (rather than just advance with revs) dealt with via other means (MAF)? Or do we still have a inlet manifold vacuum sensor as another input to the MS/ECU?

Cheers, T i m

Definitely. Throttle-body injection is just a slightly more precise carburettor, in many ways.

Diesel injection is REALLY hard...

There's several ways of doing it - but yes.

Yes and no. Yes, in that it sends the same information. No, in that it'll probably not be the same sender.

Yep.

You don't have to have a lambda, you can run open-loop. But a wideband lambda is really, really useful for setting it all up, and running a normal narrowband lambda closedloop can help a lot.

A less shagged distributor might not be a bad plan...

I've not looked at the ignition side at all. Dave?

Yup, along with a less shagged rocker shaft and entire engine probably. ;-)

I have 'spare' 1300, 1300GT, 1600 Kent and 2L Pinto lumps, any of which could go in the kitcar and I'm thinking of waiting till it's 40 years old in two years (and TAX free) and put the 1600 in (as it's an easier job than the Pinto as all the bits are on the same side).

Yeah ... 'Dave'? ;-)

Cheers, T i m

I'm more of an RV8 person. Is there an injection manifold and plenum made which will fit that engine from a later version? Port injection using batch is by far and away the easiest to tune for a road car. Multiple throttle bodies may look cool and can work very well but are obviously more complex.

The easy way would be to check on a suitable Ford forum. It will have been done by many already - so no point in re-inventing the wheel.

It can make sense to do things in stages. Start with MS fuel only using the original stand alone ignition system. Then when that is reasonably sorted, add ignition control. If you make allowances for what you're aiming at in the wiring etc, not actually making much more work. It's what I did with my SD1.

You do need a high pressure fuel rail. Either recirculating with the normal vacuum referenced regulator, or MS controlling the pump and pressure using PWM.

A 36-1 on the crank is ideal for both. But obviously wasted spark for ignition. If you want sequential ignition (and maybe sequential injection) you also need a cam position sensor. Easy way to obtain this on an older engine is with a modified dizzy.

MS generally computes the airflow by using a MAP sensor for load, which is built in. Just needs a simple vacuum connection from the inlet. But getting a good smooth average engine vacuum feed can be more tricky without a plenum of some sort.

But there are a choice of algorithms that MS provides in software anyway. It's just that speed density which uses MAP is the easiest one to tune for a road car.

You can use a MAF. Not really investigated that, as the AFM on the original Rover installation was a PITA when worn. And a fortune for a new one.

Understood. (I learned similar as a lad but the other way round with one of those double piston fish tank air pumps going to multiple outlets. So, instead of multiple 'T' valves in series I soldered some brass tubes into a small tin can and used that as a little plenum / pressure balancer).

Ok.

And that would be part of such a project, trying to use stuff that was readily available, easily adapted, reasonably priced and worked well.

Cheers, T i m

Makes sense. ;-)

The latter sounds easier as I wouldn't have to modify so much of the fuel line?

Not that I'm really asking specifically (atm) but JOOI, can you think of a good donor vehicle for a 'generic' fuel pump (that you could get cheap s/h) or do they sell an aftermarket / generic one cheap? Same question with injectors ... are they something you could retro-fit to any inlet manifold (within reason) or are they very specific?

(With my lack of any RW experience of this area), is there such a thing as a generic 'bolt though' injector (like an aftermarket windscreen washer nipple) as whilst I understand there is need for a good seal / path between the pump and the injector, once the injector outlet is generally in the area of inlet port and as long as there are no air leaks in general, should that be enough? Do they actually point 'into' the inlet or can / do they inject into the manifold in general and the fuel just evaporates and is drawn in during the induction stroke?

Cheers, T i m

The only 'V' I've ever had was a Ford V6 in my Bedford CF camper. ;-)

Not that I know of but as I have mentioned elsewhere, this isn't really my field.

'Using batch'?

Ah, ok, using a manifold with a single / common TB.

Ok.

For my kitcar (Rickman Ranger) they generally seem to fit a Zetec engine and so most of this will already be covered. However, that's probably too much 'heavy engineering' for me compared with just doing the bits we are talking about here so far.

Another thought, what would you imagine I'd likely to see re increase in fuel economy on something with the aerodynamics of a garden shed? I believe injected engines are more efficient than a carb / choked engine from cold? eg, if you did lots of short cold-starting trips an injected engine might show a more measurable return on mpg than a carb / choke (especially a manual choke).

Cheers, T i m

Batch injection. MS - at least, the version I was playing with - has two batches of injectors.

So the ECU fires batch 1 - pots 1/3/5 - at once, and batch 2 - pots 2/4/6 at once.

How long is that piece of string?

They're really just slightly cleverer auto-choke. That's what the coolant temp sensor's for - it says "The engine's cold, use a bit more fuel" - rather than "the engine's cold, use a bit less air".

You can certainly use a MAF with some versions of MS. The MS3 - top of the range - can do just about everything. At a cost, of course.

I'm a lover of the MS2. You can pick up a good used one for about 250 quid on Ebay. Will do pretty well everything you'd need on a fairly simple 2 valve 4 cylinder engine.

I do have an MS3 waiting to fit to the SD1. Going to play with sequential injection. Snag is not it - but the mods I need to make to use different injectors and fuel rail.

The problem is that the fuel circulating keeps the pump itself cool Go to a PWM controlled single rail, and you need a 'special' pump. Ie, more expensive. But I've no direct experience of this.

Pretty well any inline injection pump will be OK. The one on the SD1 which is gravity fed from the tank is shared with many other cars - right up to the Jag V12. Smallest being a 2 litre. A pattern replacement costs about

Injectors come in a bewildering choice. They generally just push into the manifold with an O ring or whatever for a seal - as there's not much of a pressure differential to seal.

As I said, though, find a forum which deals with your engine. You'll get all the full answers there - I can only give general ones, unless an RV8.

Batch injection is where 2 or more are fired together. Normally alternating. And you can set the number of squirts per cycle. Most common would be 2. The injection timing isn't as critical as spark timing because fuel is only drawn in on the inlet stroke. You can go to sequential where the firing of each injector is controlled, but this doesn't make as much difference as you might guess. Although is the norm these days.

An MS2 has two injector drivers. Each of which can drive up to 6 injectors all at once. And MS3 with the extra board has 8 drivers.

You certainly should get a MPG bonus if correctly tuned. But it must be said, tuning for the very best economy take a lot longer than just getting it running really well. My SD1 is very similar MPG wise than with the original Lucas injection. The big difference is how much more grunt it has at low to medium revs. Like you'd fitted a bigger engine. Maximum power is about the same.

With the SD1s, generally the injection ones do give better MPG than the carbs ones, as well as more oomph.

Ok ...

So, it will be 'injecting' when 2 of the cylinders won't be on induction?

Sure, but I didn't know if there was a generally accepted approximation, you know, for any given vehicle with a carb and injection version the injection was typically (say) 5% better mpg?

... that in turn draws more fuel (/ air ratio)?

Doesn't the manual choke open the choke jet, lift the tickover and operate the choke butterfly? (I can get confused between cars and bikes if they do stuff different at times).

Cheers, T i m

Yup, You can also alter the number of squirts per engine cycle. 2 is perhaps the norm. 4 is better - but you may start to get problems with injector minimum opening time at idle.

If you really want posh, you go for sequential. Which needs a cam position sensor to tell MS where the inlet stroke is happening.

Don't forget mapped ignition timing. With a mechanical advance, you are restricted on the shape of the curve by what's possible with bob weights and springs.

With mapped, you could make it go from 5 degrees of advance under say 300 rpm for cranking to 15 degrees at 350 rpm. Or whatever. The sky's the limit. Most can cope with a lot more advance under light load than the makers used.

Ah, swings and roundabouts etc. ;-)

Ok, good to have even a ballpark figure (to help comprehend the bigger picture).

Ah, cool.

Which for my general attempts to get an idea about it all is actually fine. I mean, the fact that you even consider it possible means I'm not going (thinking) up a dead end, it's just a matter of at what cost, how difficult and for what advantage(s) (if any).

Understood, thanks for the jump start though. ;-)

Cheers, T i m