Ign Table options: TPS/RPM vs MAP/RPM
Turbo Mini Forums

Hi,
Bill ... the 101 sounds good, I like learning things!
Cheers,
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary

Getting worried here, difn't expect anyone other than Stevie and Axel to know who I was.

The tools are around to log MAF, but it is a slight faff, as they are not linear. Using innovate gear, you can log MAF and then use logworks to linearise the logs for display on your PC. If your ECU has a spare analogue input (and many do) you can use excel to do the same.

Bill, you are world famous!

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Part 1 (Really need to finish some mods for Axel tonight).

What is engine load? You hear it mentioned a lot, but what is it. Well the confusing part is that most of the time it not the actual load (I'll explain this another time). As far as ECUs are concerned. load actually means 'how much air has the car just ingested' and it needs this to know how much fuel to add. Very simple to say, not quite so simple to do.

Ignoring flapper meters, which are nasty and old hat, we have 3 methods currently available to us which give differing amounts of information to work with.

Firstly there is alpha-N. This tells us what the throttle angle is, nothing else. You cannot infer how much air is flowing from this information. However it has advantages for very highly tuned engines.

Secondly there is speed density. This measures the manifold pressure, which combined with the ideal gas law (PV=nRT) allows you to infer nR (number of moles if you are interested)

And finally there is hotwire/MAF. This measures the mass of air flowing past the sensor and on first inspection would seem to be the perfect method to use and for many cases it is, which is why pretty much every new car today uses one. But it has limitations. Not least it gets confused if air flows out as well as in.

I can't go through all the pros and cons tonight, so I will start with speed density, as it is the method most often used with boosted motors and aftermarket ECUs and it is my favourite.

Reread the ideal gas law from above. P=Pressure, V=Volume, nR=amount of luvverly 02 molecules and T is temperature (in kelvin). Nice simple law, and linear to boot. For example you can use it to work out how much more air will get in if you fit a cold air induction and drop inlet temps by 10 degrees or fit a better intercooler. say inlets drop from 30C to 20 C then you get 303/293 or 3.5% more molecules in.

All seems great. The problem comes that the engine is tuned system and so in some cases won't fill the cylinder fully and in other cases will manage to overfill by inertial supercharging. So you also need an measure to tell the ECU what the volumetric efficiency of the engine is at different RPM. For the record 100% VE equated to the american goal of 1HP/CI for a hot V8 (although these days stock corvettes do better than that). The good news is that, for a particular RPM PV=nRT.VE does hold, so you don't needmuch information for the ECU to get on with the job. This is especially useful when you are boosting, as the VE doesn't change between off boost and on boost to any great degree (exhaust back pressure does change it, but if the effect is major then you have a big problem you need to fix).

Open up the tuning screen for an ECU and you will see a table. It will be usually between 12by12 up to 32 by 32. For an SD system if you run at the same AFR everywhere you only need an Nby1 table (This is important and why many people get very confused with setting up this method at first). The rest of the entries on the y axis (MAP) are really to allow you to change the mixture from lean at cruise to rich at full throttle (or under boost) and some overrun settings for when you lift off. Don't worry if that doesn't make sense at first as it is non-intuitive.

To find VE, you look at the torque curve. SD ECUs care nothing for power, they work in the torque domain. VE and torque are, for our purposes the same, so peak VE occurs at peak torque. If you happen to have a torque curve for a setup like your own, you can use this to generate a map that should be 85% there.

Need to get back to soldering. Let me know if this first installment helps or hinders, and please ask questions. For all I know you lot actually understand this better than I, so feedback is helpful so I pitch things at the right level.

Bill

On here bill you will find a huge variation in understanding of anything engine related!

There will always be those who understand and those who don't - I guess I am saying if you start simple and work up you will please everyone :)

Alex

"I always assume people know more than me about a subject - that way we all learn"

AlexF

Bill I think that was pitched just right for eps. 1 ... Some will be thinking getting on with it ..... others will be wondering what a small sub-terrainian creature has to do with fuel injection ...

As Alex suggests we have everything from those still to learn which end of a combination spanner works best as an emergency hammer right up to those who are professional automotive sector engineers liek myself.

Esp.2 soon I hope .....

Every day is a school day ...........

How fast and how expensive ...... the same question...

Just time for part 2. - when SD goes wrong and why that isn't a problem for turbo motors.

As I mentioned in part 1, engines are a tuned system. The more radical the cam, the more tuned it becomes. Eventually you get to the point where, at idle or low RPMs the reversion is such that MAP is bouncing around all over the place as gasses trundle up and down the inlet. This confuses the ECU no end. You can damp the MAP signal with a fuel filter and mig jet but very leery engines just don't get on well with SD until they are doing 2500-3000 RPM.

Now this is not a major problem with turbos, as 98% of people run a fairly mild cam with a turbo. With a turbo cam you generally get a nice stable signal to work with most of the time

So what are the other limitations of SD? Well, whilst it copes with altitude changes that we see in europe quite happily, if you drive from California to colorado without stopping it will get a bit messed up unless you have barometric correction (if anyone is interested this is because the lower atmospheric pressure improves exhaust scavenging).

It's other limitation is that it cannot autoadjust for anything that fundamentally changes the torque curve of the NA engine. If your airfilter gets clogged and then you clean it, it will adjust for that with no problems. If you up the boost and want to maintain the same AFR, no problems. But if you port the head or change the cam, then you need to retune.

It is still my favoured mapping system. Like everything, when you understand its strengths and weaknesses it all makes sense.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Part 3 .......... ?

LOL

good stuff Bill.

The question will no doubt be what about one of the major systems with compensation via one of the others ....

Every day is a school day ...........

How fast and how expensive ...... the same question...

Patience. I'll see what I can do.

Every day is a school day ...........

How fast and how expensive ...... the same question...

OK Part 3a or why MAF is not always such a good thing.

On paper MAF looks to be perfect. Just bolt it on the front and it measures mass of air going in, which is what you need (nR from the ideal gas law). And in many cases it is very very good. However as ever there is no free lunch and it is these gotchas that you need to be aware of.

First off, bar Motec and Megasquirt very few aftermarket units support MAF at the moment. Key reason for this is the target market, which is generally loopy NA engines with IRTBs where a MAF just wouldn't work well.

Secondly, even though you have air mass, you still need to measure Inlet air temp (IAT).

Thirdly, if you are boosted you generally want to measure boost pressure so the ECU can control the wastegate solenoid or back off if things go out of control. So you need more inputs than less.

Now to the MAF itself. They are wonderful things but you need to know 3 key things about them. Firstly they are totally non-linear, so give great resolution at high flow rates, but the transfer function is nearly flat at idle. The higher the peak flow, the more this is a problem. So pick a MAF just big enough for your needs, as too big is bad.

Secondly they need a nice laminar airflow over the sensor and they can't necessarily tell which way the air is flowing. Modern units are better at dealing with reversion, but OEMs spend a fortune characterising the MAF to the inlet track to make sure that what the MAF reports is what the engine is injesting. This is also one of the 2 reasons that changing the air filter can mess things up on a MAF car. The second is that oiled filters like K&N will really mess with that hot wire.

Thirdly, if air escapes anywhere then the MAF doesn't know this. This is why tubo cars need a recirculating dump valve as the air has already been metered, so can't be vented to atmosphere.

If this all sounds a bit negative remember the fact that turbo cams are generally mild but with high lift and MAF makes a lot of sense in terms of a relatively easy to tune system that will automatically adjust for most of the changes you will make to an engine. Find a car with 150HP or so stock that turns up in the scrappies regularly and you should have a good source. They do fail/cake up over time so a cheap spare should be kept in the garage.

I haven't personally looked at the range of MAFs available in UK scrappies. Mainly as I have a car that doesn't suit them.

Before I get onto alpha-n and where it can be really useful, a few notes on ECU boost control. If you are going to put EFI on a turbo get, spec an ECU that can do this. After mappable ignition it is the second best thing out there.

Why? Because it puts you back in control of the boost. You can run a bigger turbo and get a nice smooth build up of boost (rather than the nothing...nothing...nothing...hedge you can end up with). Of course if you want loopy spool up you can have that as well. They can be a bit of a swine to setup, but are worth the effort.

Im a fan of aftermarket electronic boost controllers myself.

they aid spool, and offer excellent boost control. Its surprising teh difference they can make over simple bleed valves etc.
If you buy sensible ones, they are also very easy to adjust and setup.
For me, thats a huge bonus over an ecu controlling boost.

Simple and fast. Ecu's can be slow and tedious to set up to get the same results, if indeed most ecu's are capable.

9.85 @ 145mph
202mph standing mile
speed didn't kill me, but taxation probably will

That is often more that the ECU implementation is suboptimal than the controllers are particularly good (although some of them are super bling).

Thast not a word I thought i'd see you writing bill :p

Alex

AlexF

Which word?

that ? lol

9.85 @ 145mph
202mph standing mile
speed didn't kill me, but taxation probably will

bling ....... :)

Every day is a school day ...........

How fast and how expensive ...... the same question...

How else do you describe a boost controller that comes with a remote control and colour screen?

chavtastic?

Japanise?

:p

But your dead right LOL!

AlexF

Yup. Actually some of them are incredibly advanced and can do per gear boost control and all the other good stuff you need for the perfect setup. Suspect that 90% of the purchasers have no idea how to use them properly mind.

Personally, bar perhaps having a wet/dry setting I think the right foot should command boost and good ECUs will allow for a table of boost vs TPS to setup.

The emerald does boost control and in gear stuff but TD said it wasnt all that good?

I havent played with it yet as i havent got round to get a N75 valve to use.