Note to all those reading this thread - the drag co-efficient of AB's car is considerably lower than that the rest of us experience as there is no associated friction from the air over AB's car.

This can give considerable performance gain and makes it a bugger to open the door as your hands do not stay on the door handles.

 

You may have said that in jest but yesterday whenI went to open the door, my hand slipped!!!!

I was planning on doing some more driving today - wow - yes 3 days on a trot but now I am a bit worried until Brian sheds a bit more light on the subject of A/F ratios.

 

I knew it, I knew it all along

If the weather stays nice are you going to join us at the meet tomorrow night in Surrey? I appreciate the concersn you raised about the A/F ratio.

IIRC and await further input from Brian, but 14.1:1 used to be deemed ideal for an IC engine. But my info is well out of date

 

Stochiometric rate is the rate name I believe (it just came to me) but again I could be wrong.

 

Eh???

 

AB should be the opposite, from Standing he has more advantage because it is AWD and from rolling the AWD advantage almost disappear


Juan

 

Yeah that's what I thought too! But he can't get his car off the line to save his life!

Anyway, Juan, can you shed some light on this situation pls:

Mugen CAI
J's manifold
Amused dual
Stock ECU

Am I running the car dangerously regarding the a/f ratios?

 

You should run ok, this are basic bolt ons and the ECU can compensate the extra air.

 

This is a reply from Brian Marshall which he has asked me to post as he is having difficulty in getting onto the site.


Quoting myself here:

I've not said that the car's about the blow up. Chill out a bit peeps.
Emphasis on the word 'chance'.

Right I'm not going to go into the nth level of detail here. There's still
going to be quite a bit.

The standard ECU is mapped using speed density. As are the Mugen, Amuse, and
J's ones - which are just remapped standard Honda ECU's share the same
approach.

Speed density works like this. It has the rpm of the engine. That's the
speed. It has the intake manifold vacuum from the map sensor - that's the
load. Those provide the lookup for the fuel and ignition. Happy with that?

So for a given rpm, and load, the ECU adds a given amount of fuel based of
what Honda wanted.

Then you modify the car. To get more power at a given rpm you need more
torque at a given rpm. To get more torque at a given rpm, you need a bigger,
or more efficient 'bang'. People who say it's all about airflow are only
seeing a very small number of the variables. Anyway, you're either getting a
bigger bang, or increasing the efficiency of the bang.

So you bolt on a new exhaust manifold. Wow! Pulls a lot better now. So
where's the pull come from?

In normal cirumstances, the fresh intake charge mixes with some of the burnt
charge from last time. Can be around 20% burnt at low throttle, with a lot
less 7% or less on a good engine at full throttle. Obviously, a burnt
mixture won't have the oxygen in it to support combustion, so you're not
going to get power from it. So, one of the ways a better exhaust manifold
creates gains is by reducing the percentage of burnt charge. (Trickier than
it sounds as it's a balancing act - do too much and you end up pulling the
intake charge downt he exhaust which isn't good for emissions etc...).

So, it's likely that one of the thinks the J's manifold has done is reduce
the burnt charge, hence more fresh... which means more fuel.

The other way is by getting a bigger charge in. Remember how the density
side of things is coming from the map sensor? That's measuring the pressure
in the intake manifold. The higher the pressure (or lower the vacuum) the
more air that's going in. So... traditionally it's viewed that an N/A
engine can only see a vacuum. At full throttle the map sensor will see a
small amount of vacuum - so close to atmospheric pressure. However, there
are tuned N/A engines that actually see a small amount of positive pressure
now. That's where all the wibbling you see about airflow alone fall down.
Above ambient pressure only comes from reinforcement from pulse tuning -
intake and exhaust. I'm not going to go into any details on that, only say
that the better exhaust manifolds out there can contribute to more pressure
in the intake.

So, it's possible that the J's manifold is also increasing the charge. (Less
likely as the design while probably the best easily available, isn't cutting
edge by some margin).

Okay, back to the ECU. The rpm the engine is running at is the same, so
that's not a factor. So the ECU has to rely on changes to the map sensor
signal to make any changes to the fuel and ignition timing.

Now, most of the time, the ECU will just see less vacuum at a particular rpm
than it did before. If there's room to go in the fuel map, it'll lookup
there and often you'll have a usable value. Not ideal, as Honda didn't map
the ECU for what you've bolted on. But normally okay.

But what if the map sensor isn't 'seeing' everything? The map sensor itself
has a low level filter built into it. You get some fluctuations from the 4
cylinders sharing the same plenum, so the filter damps a bit of that out.
Problem is, if you see pulses above a certain frequency, the map sensor
can't 'see' them because of the filter. The other thing is, the ECU only
samples the map sensor once to do the lookup then squirts that amount of
fuel. So if you're gaining a lot from a pressure pulse after the map sensor
has been sampled for that injection event, it can't see the pulse even
ignoring the filtering. When you're mapping the car, you're generally not
too bothered what the precise map sensor reading is. All that matters is
that it's repeatable - so when you have that rpm, with that load, you set
the fueling and timing and it's going to be right every time. In theory with
speed density, you're measuring enough to calculate exactly what's going in.
In practice, as I hope I've just shown, things are less than perfect.

So, in theory with a speed density mapped ECU, if you make a small change to
the mechanical setup, the map sensor signal will be different, the lookup
different, and things will be okay. (Not perfect, but okay).

However, the further you go from what the original setup was, the more
chance that things will be less than ideal.

One thing to mention before I go any further. UK ECU's do not appear to have
the same map as US cars. I'm basing that observation on comparing published
results from US cars, to results I've gathered myself.

Right, onto AB's car. The top end power comments don't both me one bit. The
standard Honda ECU runs very rich there. I seriously doubt the setup runs
lean enough to be worrying at all (indeed, it's likely to want to run leaner
there for peak power).

Then the comments on the mid-range - pre-vtec, 4000 to 6000rpm. That's a
different kettle of fish. The standard ECU runs pretty lean there. It's very
much mapped for a combination of economy and emissions. So if you make a
significant change there.. well, you want to have a look at things. (Hence
the suggestion that seems to be interpreted as 'oh my god my engine is about
to blow up').

The biggest gains on the F20C, without doing any internal work, come from
exhaust manifold changes. One of the UK cars I worked on, running with the
stock ECU, changed to a 'good' exhaust manifold. That ran lean enough to
give some nice hot EGT's. Not enough to cause immediate risk of damage. But
hotter than was ideal from engine wear, cat life, valve stem, seal life,
clutch system seals, and a few other factors. Suffice to say it wasn't run
like that for long!

So, given I've seen that on a UK ECU before, and AB's comments on the
midrange... I have a degree of concern. I don't think AB's engine is about
to blow up. I also would not run that setup without checking that things are
within a reasonable margin. That may be. But my experience to date on UK
cars with similar mods is that if it is making that much more pre-vtec, then
it will be well outside what's sensible to run long term.

The reason to change the cooling setup before running the Mugen ECU is
unrelated to this. Running the engine at the right temperature is important
for power and safety. The standard ECU has many layers of protection. The
Mugen ECU doesn't - it removes nearly all of them. So to give yourself
enough safety margin to run the Mugen ECU on the road with the variance of
UK fuel, you do the cooling mods. (Worthwhile on any F20C - the standard ECU
has the safety margin, but still runs better when running consistently at
the right temperature.).

The reason not to run the Mugen ECU with the J's manifold is simple. The
Mugen ECU is mapped to remove the margins to gain power for a car with
either the standard exhaust manifold, or the Mugen one (which is very
similar to the standard one in design). Pair it up with a very different
exhaust manifold, and the mapping will be a long way from the ideal. So a
much greater risk than the J's manifold with the standard ECU. That goes for
any of the fixed map replacment ECU's - it's pretty much match them to the
same brand bits, or you're better off going for a mapped solution. (Which
doesn't necessarily cost much more....).

Anyway, I'm going to stop posting these long technical posts. I keep getting
the feeling that they are being at most part understood. A little
understanding of some of these things is at least as dangerous as none! If
anyone wants to talk in detail on engine or suspension stuff for the S2000,
pop along to a Honda Revolutions event I'm at and have a chat. I'm at the
Honda Rev's Croft trackday a week on Saturday, and -might- be at JAE on the
Sunday. Aside from that, I get along to a lot of the events (and now the
northeast regional meet), so it's not too hard to track me down. If that's
too hard... well I'd probably be wasting my time and yours if that's the
case.

-Brian.

 

Thanks for the write up Brian

Got to do some thinking now...