Emerald: stupid question #1

[charlie_pank]

My (returning) overfuelling problem has spurred me into ordering an Emerald. - When I phoned rover and spoke to them I had to negotiate with the manager to pursuade them to look at a seven as opposed to a tin-top, and then they still couldn't do it for 2 weeks.

 

I'm convinced that my overfuelling is caused by a faulty sensor, but without replacing all of them, or stealing the right software & cables from rover to hack into my ECU, I can't tell which one it is. I don't like stuff I can't fix myself, in fact I don't own many things that I can't at least have a go at fixing myself, so I've decided to bite the bullet and get an emerald, so I can see exactly what information the ecu is getting.

 

Reading through the manuals, to just aquaint myself with the system I was wondering if there are some general rules of thumb about altering your mapping. I understand that running too lean will result in detonation & high temperatures, and that running rich cokes up everything and runs cooler, but what different results do you get from changing the spark timing?Also, are there some characteristic signs that make you say aha! this is sparking too early, or this is underfuelling at this engine speed etc...?

 

I apologise in advance for what will most likely be many stupid questions along these lines, probably ending up with an ad in the wanted section for a new engine, after destroying the last one through messing about with the map!

 

 

C

 

Charlie'n'Kermit

The plan is: Find out what time to be at Caterham for baywatch.

S5EVN

[scooby dooby doo]

sounds like you need the haynes book DW wrote about engine mapping etc.

my crude memories of this are:

 

lean = too hot = bad

rich by upto 20% = high fuel consumption / emissions

VERY rich = coking etc

 

spark = more advance = more power until you get too much and it goes wrong.

 

its WELL worth getting them to map it and then you can tinker a bit afterwards. especially with things like cold start etc that need it to be winter to do for the lower temperatures.

 

HOOPY

R706KGU Hoopylight R

[JAG]

Hi Charlie,

 

I don't have an Emerald ECU but do have an injected car and I have found much useful info on the DVA Power website here

 

There is some basic info regarding fuel and ignition settings and the whole 'mapping' process. I found it very useful

 

 

 

Justin

 

Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.

[Peter Carmichael]

You can get away with a very derivative ignition map. The injection map needs to be *right *. You can guesstimate some fixed pooints on it if you know how much power the engine is expected to make and at what rpm it makes peak torque. By the time you have set up an idle, it is a reasonably easy to get a workable map by interpolating between the known points and working out any drivability issues. Lean spots will appear as stutters in the acceleration.

 

Basically for the first few passes you fix all problems by throwing more fuel in. You can get something very workable from this basis. My car is running on a map that have not been near a rolling road. I have been very hot on version management, so I have a full record of over 30 attempts to get it where it is now.

[V7 SLR]

Glad it's not just me... I've version controlled mine too. I've tried to indicate the state of tune of the engine in the comments field but there's not enough room. Obviously there's a limit of some description. So, I've started to change them to reflect only what's changed since the last map. Seems to work as there's quite a few people running versions of my maps out there.

 

 

Only about half of mine are from the RR so anyone using mine, be very afraid.

 

Worcs L7 club joint AO.//Membership No. 4379//Azure Blue SLR No. 0077//Se7ens List Tours

 

[Bill Shurvinton]

I would recommend if you like to know what is going on inside your car, get a Wideband 02 sensor. It gives a very good insight in what is going on in your car. The standard one is good for emissions with a cat, sod all else. You can get up and running for around £250 and the price 'may' come down unless VW throws a wobbly.

 

As to the numbers. Best power is usually quoted as 13.5:1 AFR, however this is not quite true. Anything richer than 14:1 will ensure complete combustion of all the oxygen in the charge, and additional fuel is entirely for cooling. Depending on load you could need to go as rich as 12.5:1

 

However load is a funny thing. On the rollers you are held at maximum load (full throttle and static speed). On the road/track you are unlikely to reach this position, except perhaps at the ring, and even then not for long. Usually in a se7en full welly will cause acceleration. This means that you can run a se7en a little leaner than a tin top. Load has nothing to do with 'load site' on the ECU map unless you also take vehicle speed into account.

 

Another tricky is that, with an alpha-n system, throttle angle bears a strange relationship to output power. At lower RPMs, full throttle may be at only 20% opening with TBs, and you may not reach the point where over 75% is needed until over 6000RPM.

 

This is why I prefer MAP based systems, as these give a better representation of what is actually going on. Either system can be mapped to give the same power and driving manners, just one is easier to understand.

 

As you back off the throttle from max power at any RPM point you can start to lean out quite a lot. I expect a K will cruise happily at around 16:1 if you bung a bit more advance in ( If you lean 1 AFR you can advance spark by about 5degrees).

 

Personally I tend to look at RRs as just to get the bragging sheet. For everything else I find a drive around with a friend bashing the laptop far more satisfying. This is however 'my' view and YMMV, yadda yadda, won't make you more attractive to women etc.

 

Bill

[Peter Carmichael]

Load has nothing to do with 'load site' on the ECU map unless you also take vehicle speed into account.

 

I'm interested in this statement... and I may get toasted by superior knowledge... and believe me I am just thinking this through as I type...

 

Under big throttle openings in a seven, the car accelerates...for sure. But is that significant from the charge's point of view? The charge is in a closed cylinder with a piston face below it. The only way it knows about different road speeds of the car is if the characteristics of the piston dropping away are different because of crank acceleration.

 

Let's take blipping the throttle from idle to 6000 rpm in a second as an extreme example of crank acceleration. I'll have to assume a linear characteristic of crank acceleration. I'll also make it simple with a 1200rpm idle. 1200rpm is 20 revs per second. Each charge is only on a firing stroke for half a revolution every two revolutions, so there are 40 firing strokes per second. At 6000 rpm there are 200 firing strokes per second. Because of my assumption of linear acceleration (linear torque) I can bung the numbers into a spreadsheet and work out the top dead centre points by approximation. I know the greatest possible increase in crank speed will take place in the firing stroke that takes the longest time. i.e. the first one.

 

I've done this and it gives me numbers like...

 

At the start of the first firing stroke the crank is at 1200 rpm. At the end it is at 1315rpm when the piston accelerates and decelerates anyway in the course of the stroke. This is 9.6% increase. Can the charge really tell? Maybe the acceleration is more concentrated because the firing stroke does its best by 120 degrees ATDC and maybe was ignited 10 degrees BTDC,

 

Now look further up the speed range to 5000 rpm. I've got an example firing stroke in the spreadsheet where the acceleration is from 5006rpm to 5035rpm. This is a 0.6% increase over the course of the stroke.

 

So my example is for free revving. If you have the car in gear, the acceleration is obviously much less, so how can the *load* on the engine vary with the road speed. How can the charge in the cylinder tell what the load is?

[Bill Shurvinton]

fsking thing just lost my answer. Bother. Will retype in the morning

[Bill Shurvinton]

Right, lets have another go.

 

This is a tricky thing to get your head around. I'm still only 90% happy with the concepts. However,.. Startining with your revving from idle. In the case of your £££ cheese engine all you are overcoming is the minimal inertia left in the engine/flywheel system, which requires very little power to accelerate it. You could lean the mixture off a silly amount and it would still rev cleanly and quickly. However on the road it might cough, bog, spit and misbehave as the load has increased.

 

Your sums bring up an interesting issue though. Best spark timing is usually set to ensure that max cylinder pressure occurs 16 degrees ATDC where the mechanical advantage is highest (unless you have offset wrist pins). When accelerating hard in a low gear, if you fire the spark as set up on the dyno to give this, then by the time the pressure builds up in the burn, the piston has moved beyond the optimal point, so you are actually losing power. So you need more advance in lower gears. (ignoring here that you are usually traction limited in these gears, so may want to retard). So, to do an optimal tuning job, you need both gear and speed input into the ECU.

 

Anyway back to load. remembering that best power comes at 13.5:1 ish. If you are accelerating, the time at any RPM point in which heat soak can build up, requiring extra cooling is minimal. If you goose the throttle at 4000 in top, you'll leap forward like a leapy thing. By 8000, this thrust will be starting to diminish, so the heat in the cylinder head will be starting to build up. You cool this by adding more fuel, or water.

 

If you put the same engine in a Rover 45, added 3 kids and a caravan and repeated, you would not accelerate very fast at 4000, as the load is higher, so you would need more fuel to prevent the heat build up. not intuitive, but that's how it is.

 

Now if the tune for the shed jerking 45 was put back in your rocket, you would not notice a lack of power, you would just use more fuel. The tune would not be optimal. however, it depends what you are optimising for. As and when ION sensing ignitions become readily available, where you can time each spark for optimal performance it will be interesting to see how much you really can reduce fuelling in a se7en before performance is measurably impacted.

 

At the extreme you have the mitsubishi GDi engine, which above 3k modulates power entirely with AFR, leaving the throttle wide open. This is because it can burn a 40:1 mixture if it wants. V.trick, shame its fitted to boring cars like the charisma.

 

I don't expect you to fully buy this BTW, but if you pull out any good text on engines you will see the chart that shows power vs BSFC (or AFR). Then talk to an aviator on how they tune for level flight by opening the throttle and leaning the mixture. Bits of it will slowly start to make sense.

 

Bill

[DohNut]

Thanks Bill

 

My first stumbling block (of many) is the 40:1 AirFuelRatio

I am happy that you can take the throttle position as a "request for power" and then add enough fuel to produce that power but will this not result in overheating 🤔

Although I have to admit that I never understood the mechanism of overheating due to lean mixture.

 

Nick

[Peter Carmichael]

Ok, I'm buying into a lot of this, but still have a couple of queries...

If you are accelerating, the time at any RPM point in which heat soak can build up, requiring extra cooling is minimal.

Why is heat soak aware of rpm point? If you dump a load of heat into the cylinder walls at 4000rpm with an accelerating engine, is it really any different as far as the engine is concerned that the next cycle is at 4040rpm and still dumping heat into the cylinder walls. If you average it out, wouldn't reject heat closely follow fuel consumption; whether the heat soaks or is assisted in dissipating by extra fuel evaporation is, I agree controlled by air:fuel ratio richer than stoich.

 

Another quibble:

Starting with your revving from idle. In the case of your £££ cheese engine all you are overcoming is the minimal inertia left in the engine/flywheel system, which requires very little power to accelerate it.

Except in my assertion the piston forces remain the same and the lack of inertia just results in very high angular acceleration of the crank. The torque/power at the flywheel drops, but this is down to accelerating internal components. My assertion is that the difference in position of the piston throughout the firing stroke is minimal and hence fuelling and timing changes to achieve maximum output from the firing stroke do not change. The charge does not know whether the engine is running at constant velocity or accelerating.

 

Edited by - Peter Carmichael on 11 Jul 2003 12:13:59

[Bill Shurvinton]

To understand the lean engine eating you need a copy of Glassman's work on combustion, a bottle of whisky and a cold compress.

 

A number of factors are at play, the main being that combustion temperature goes up with a leaner mixture UP TO A point. Peak temperatures actually occur at stoich. leaner than this and they actually go down. So the statements about running lean should be related to best power mixture. Once you go below lean of stoich you are safe again.

 

This all assumes the swirl characteristics of the engine will allow you to run lean of stoich. If there is not a good mixture distribution, then you can get uneven burn and local hotspots and other nasty things

 

Bill

[Peter Carmichael]

I've got the whisky 😬

[Bill Shurvinton]

Peter,

 

like I said it's tricky to explain why dynamic conditions in an engine are different from static but they are. I'll try and trawl through some of my weightier tomes and see if there is something there.

 

But for instance look at pulse tuning in the inlet. If you are accelerating, then the pulse tuning will not be at the same level as it would be if you were static, changing the volumetric efficiency of the engine. Likewise in the exhaust. Not to mention cylinder head temperatures. Your charge model doesn't quite take all this into account.

 

On the dyno you tune for Max BMEP without knock, which is fine, but only the starting point for the 'perfect tune'. If you look at the tuning maps that the OEMs use you will note that, if the same engine is used in 2 applications with different diff ratios, then they will use different fuelling and ignition curves.

 

However the onus is on me to go away and try and find some substantiative evidence, so I will.

 

Bill

 

 

[Paul Ranson]

Seriously good thread under an unlikely heading.

 

So why do we need acceleration fuelling?

 

Paul

[Bill Shurvinton]

Accel shot is primarily to overcome the fuel dropping out of suspension when there is a sudden drop in manifold pressure.

[Peter Carmichael]

My understanding of acceleration fuelling is that it is mostly to do with fuel paths into the cylinder.

 

Again, I am ready to get toasted...

 

A lot of fuel gets into the cylinder from the inlet tract wall. Lots evaporates off the hot valve. Some atomises in the shear of the high speed airflow. In manufacturer's research centres, carefully timed sequential injection and big research budgets might get that sequential thing working quite well. In the realm of the motorsport engines that I deal with, where I don't care if a bit of extra fuel gets sloshed in, but I do want to chuck in a lot of fuel into a big charge it is a bit different.

 

At steady state, by definition, the amount of fuel arriving on the inlet tract wall equals the amount evaporating off the inlet tract wall and entering the charge. The inlet tract is hottest nearest the valve and gets colder the further out you go towards the injector position. There is the most chance of the inlet tract evaporating all its fuel, until it is dry, nearest the valve. Indeed it is no different to the "evaporate off the back of the valve" model except the inlet tract is much cooler because it can also dump heat into the water jacket. The further up the inlet tract you go, the less likely it is that the inlet tract evaporates dry. There will be a boundary between the permanently wet area and the area that evaporates dry. As the charge whistles past it will drag fuel down from the colder regions of the inlet, rewetting the walls. The new injection charge will rewet the walls. Remember, this is just steady state I am describing.

 

When you crack open the throttle to a new operating position, you inject more fuel. More fuel splashes onto the inlet tract walls. It follows that for reestablishing a steady state, something needs to change so that more fuel evaporates back off the walls and into the charge. The evaporation from the tract walls is a function of surface area, meaning the wetted surface area. In the course of a few cycles, the boundary between the permanently wetted area will extend a bit further down the inlet tract and this will be achieved. Until the larger wetted area of the wall is established, there is a shortfall in fuelling. I can't remember the knowledgable person who told me this, but I found it fascinating to learn that the quantity of liquid fuel wetting the inlet tract is typically 10 times the injected quantity.

 

Above a certain point in the rev range, the other fuel paths become more efficient, lessening the need for acceleration fuelling because the wetted wall fuel path does not dominate.

 

There are certain corollaries...

 

... in the reverse situation of a closing throttle transient, less fuel is needed - deceleration fuelling. A plenum based induction reduces the rate of effect of sudden reductions in throttle opening, making it easier for production engines to manage emissions.

 

... as the fuel demands of the engine increase, the permanently wetted area of inlet tract will grow in the direction of the valve. When the fuel demands are such that the wetted area reaches the valve itself, further fuelling will result in a poorly prepared charge entering the cylinder with a consequential reduction in power and increase in emissions. The inlet design has reached its fuelling limit. You can design against this by moving the injection point further away from the valves, to provide more inlet tract wall area, but it is important to remember that this is cooler wall area so less efficient. Eventually you have to start looking at siting injectors out at the bellmouths of the trumpets and for this to work you need a dual injector setup (eight injectors for a four cylinder engine).

 

Like I say, this is my understanding. I am ready to be shot down in flames...

[charlie_pank]

oi! what do you mean "unlikely heading"! 😳 - (I still have an overfuelling problem and no emerald has appeared on my doorstep BTW! ☹️)

 

Charlie'n'Kermit

The plan is: Fit an emerald so I can find out why Kermit's overfuelling

S5EVN

 

Edited by - charlie_pank on 11 Jul 2003 15:15:59

[Paul Ranson]

Sorry Charlie...

 

I hope you find that the air or water temperature sensors are reading nonsense. Good luck.

 

The seriously interesting stuff remains seriously interesting.

 

Paul

[Bill Shurvinton]

I hate these batsrad message boards that require a logon each time. Another message lost.

 

Peter, your theory doesn't stack with anything I know (which ain't much)on intake effects. With port injection you target the injector to the inlet valves where it evaporates before the valve opens. Evaporation is not a good thing though as it takes up the space where air should be. Atomisation is the key to power.

 

For atomisation you want the injectors higher up the inlet and you time the shot for an open valve to minimise evaporation effects. You have now crossed from a 'dry' manifold into a 'wet' manifold. You also have injector on times larger than the closed valve time, so you have a number of phases occuring, including the little puff of exhaust gas that pops out when the valve opens, vapourises everything and then disappears down the exhaust, taking your nice charge with it.

 

The accel shot deals entirely with fuel falling out of suspension, and its duration IS related to the time taken for a steady state to re-establish and excess fuel to evaporate. If you do not have a heated inlet and high up injectors this will take longer than with a heated manifold.

 

In an old fashioned single point injection system there will be a lot of wall wetting and evaporation may dominate, but in a weber style system this should not be so unless you have the wrong injectors. In extremis you 'laser' the fuel through the open valve and there is no contact at all.

 

But in the end what is important is the difference in power between batch and sequential (which is not difficult to set up with a good ECU). Based on dyno plots I have seen, its a whopping....3%

 

[charlie_pank]

Sorry Bill, didn't intend that to imply that your conversation wasn't interesting. I was just being indignant!

 

Charlie'n'Kermit

The plan is: Fit an emerald so I can find out why Kermit's overfuelling

S5EVN

[Bill Shurvinton]

Opps yes,

 

Kermit. Shouldn't take long to sort that one actually. Have you tried disconnecting the battery just to reset things?

[Bill Shurvinton]

Long soak in the bath last night thinking this one through and I think I came up with an acceptable explaination.

 

Imagine the same engine in 2 cars as discussed before. Also imagine that fuelling is set to best power. One car is light and can accelerate at this point, the other is at full 'load' and therefore static speed.

 

A number of effects act to cool the cylinder head such as the water jacket, flow out the exhaust etc AND vitally, the new fresh incoming air/fuel mixture. Remember I said earlier that evaporation of the charge before it gets into the chamber is bad? If it is atomised, then you have the high latent heat of vapourisation to help cool things. Sure it cools the inlet valve, but its the exhaust that is struggling [digression] This is why liquid phase propane injection is so good as it cools the inlet charge to -30C[/digression]

 

If we also assume that the best power mix is not quite enough to keep an equilibrium then the heavy car will be heating up the exhaust valve and cylinder head, and will either overheat or start to detonate. However in the case of the lighter car, as it is accelerating, the mass flow of air and fuel will be increasing and this extra cooling will help keep the temperatures stable.

 

So it can be seen that the heavy car will need fuel (or water) above the best power mix to provide extra cooling compared to the lighter car, which will not need this until much higher up the rev range, if at all. If you are to believe GM, then the extra cooling is not needed until you have spent around 10 seconds at high load. Mind you they also have knock sensors monitoring things.

 

How you handle all this very much depends upon the flexibility of your ECU and how much you care about wasting a bit of fuel. And don't worry if this doesn't make any sense. Speedy Steve bought an ECU 2nd hand for his Vx with a map already on which was 'a bit rich for safety'. It was so rich it wouldn't fire above 7000. Leaned off to a more sensible 12.5:1 (still a bit rich in my book for a NA 7) it flies.

 

Here endeth the drivel

[Peter Carmichael]

Still don't buy the: "as it is accelerating, the mass flow of air and fuel will be increasing and this extra cooling"

 

If the mixyure is the same and more of it is being burnt (more mass flow), the reject heat is higher.

 

Shurely... This must have something to do with expected duty cycle. Agree it has loads to do with heat soak. I still think a light car on track can have high duty cycle, on account of never wittingly using less than the power available around a circuit (just to rather greater effect than in a heavier car - i.e. more laps/distance, but no more time so heat soak conditions the same)

[charlie_pank]

Yes, I have tried resetting it, by taking out the FIA key. I replaced the lambda probe as well, which fixed it for about 100 miles, before the new one coked up too - this has made me think that it's one of the other sensors, but don't know which...

 

Charlie'n'Kermit

The plan is: Fit an emerald so I can find out why Kermit's overfuelling

S5EVN