AFM (+DME) modifications revisited... OR tuning a 944 for free!

[Kaos]

Here's my opinion on the subject. I didn't read all of the posts, but here's some input anyway.

Your car feels faster when it is cold out not because the mixture is richer causing more power, but because the inlet air is cooler. Once the engine is at it's operating temperature, underhood temperature (and hence intake air temperature) is much higher. This of course makes the air much less dense, and gives you less power. Just touch your intake manifold after driving your car for a while. It is damn hot, and even though cold air may be coming in from the fender, alot of heat is still being transferred to the air before it enters the cylinder.

The coolant sensor does add considerably more fuel when the engine is cold, but this is not for power. The reason your engine needs more fuel when it is cold is due to the nature of fuel. Ever heard of summer vs. winter gasoline? This is a measure of the volatility of the fuel at a given temperature. Since it is hotter out in the summer, you need to run a less volatile fuel. In the winter, you need a more volatile fuel. So, the main reason your temperature sensor needs to add more fuel is due to the volatility of the fuel. When the engine and fuel is cold (say 60F) more liquid fuel needs to be injected to get the same amount of vapour fuel into the engine.

Also, you need to realize that peak torque occurs at roughly 0.90 Lambda for most engines (each engine will vary slightly due to combustion chamber design). Lambda is a way of describing air/fuel ratio - where 14.57:1 is stoichiometry, Lambda = 1.0. So, at Lambda = 0.90 you are 10% richer than stoich - so an afr of 13.113. If you go any richer than this your torque will actually be reduced.

When a manufacturer calibrates an engine, they will typically make it slightly richer than RBT (richness for best torque) for safety. This ensures that a wide range of engine tolerances are accounted for, so that some naturally lean engines don't blow up.

Making your engine richer than it is right now is not going to give you any more power.

The Barn-Door style airmeter is a big restriction, but drilling holes in the flap is not going to give you any more power.


Kaos.

[mike944]

Ahmet,
If the DME knowingly runs the engine richer when cold, it has to ignore the O2 sensor. The O2 sensor is only accurate for a very small range (about +5% and -15% from stoichiometric) I don't know how much it's trying to richen it, but if it's more than 15%, it'll have to ignore the O2 sensor.
http://www.pelicanparts.com/techarticles/mult_air_fuel_monitor/FIG1.JPG


It's all a moot point anyway, because after some more digging on the internet, i found that the Bosch K-Jetronic (our DME), like almost all computers disables the O2 sensor at full throttle (which is what we are really talking about here anyway) At full throttle the computer is programmed to run the engine at max power ratio, 12.6 (maybe it's 13.113, whatever)

http://www.pelicanparts.com/techarticles/mult_air_fuel_monitor/FIG2.JPG

therefore if you F_ with the temp sensor, under full throttle, you'll be running it even richer than optimum, which will hurt power.

Mike

[This message has been edited by mike944 (edited 04-24-2001).]

[TheStig]

Ahmet I think I heard of this idea before...except for the holes in the flapper weren't drilled out. A 944 owner used a 928 Flapper to get a lil more of a bang out of that air intake. I am curious to see what would happen if I did the drillng. Here comes another project for me..

------------------
The only good ricebox is the one I'm about to cook..

[TheStig]

If there is a cone filter on the 944 that will have the holes drilled out, chances are there will be a lot more air in! I tightened the spring tension in order to get the car to pass smog at the sacrifice of idling like a 924. When i was all done, I loosened it up again. Ahmet I'm definitely gonna try that drilling idea.. I need to obtain another AFM though so i can have a smog legal one and a smog illegal one. NOtto mention a future big bore throttle body will make the air intake more monstrous. Than after that, its off to Extrude-Hone to get my intake manifold bored =) MUHAHAHA

[TheStig]

Ahmet, now that i think about it, is there a possibility of fouling spark plugs and back firing if holes are drilled into the little flap? If not properly altered, the AFM can and will be damaged. So, the wholse shouldn't be that big if you decide to drill them in. 1/8 inch holes would be a good idea to start with. But how many would need to be drilled? Keep in mind if there are too many holes drilled or the holes are too big, the flap may not even open at all. This idea won't be easy..

------------------
The only good ricebox is the one I'm about to cook..

[GTL]

Hello Ahmet, I am glad you are still pursuing this issue. I also thought about the drilling holes idea and have some issues to think about further:

1. If you lower the air flow restriction by drilling holes on the flap plus lowering the spring tension and calibrating so that for exactly the same input airflow you get the same rotation of the pot, you will run the engine leaner. To picture this assume that with the stock arrangement the engine was requesting a negative air pressure P1 at certain RPM. But because of the AFM flap restriction it was actually getting an air pressure P2 (which is lower than P1 because of the force required to move the flap). If you eliminate the restriction then the engine will be getting an air pressure P1 which is higher than the original P2. Think of P as being more mass of oxygen delivered. So you must compensate by making the mixture a bit richer. This make sense, since more power means more air AND more fuel. By decreasing the restriction you are only giving more air. So you must deliver more fuel when calibrating the holes openings and spring tension. The flap must open a bit more for a certain input air flow to compensate for this. You have a mix sensor so you should be able to calibrate for Lambda 0.9 accurately.

2. Probably it will be better to drill many small holes than only a single one. This is to minimize complicated turbulence and make the air flow more diffuse. Ideally a fine grid should be much better. Also, when the flap starts to open, the holes cross section in the direction of the flow changes, so we may have some restriction non-linearities to worry about as well.

3. Drilling holes in the flap is like increasing the idle air mix and CO content. There is a screw on the AFM to calibrate the idle mix. Probably you should close the screw all the way after drilling the holes. But if the holes are bigger than the bypass, then you may be creating overly lean idle mixes. Unless you also compensate for this somehow. I do not know how at this point. Also closing the throttle idle by-pass a bit also (?)

Well, I hope this helps. Overall I think this effrot is worth pursuing. As you know, I also have a spare AFM for experimentation. But with the drilling holes thing I will prefer to see how all comes with you first. Please let keep us informed. This is exciting stuff. I will contribute as much as I can.

Cheers

[1.2gees]

Lots of things to discuss here, glad more people read this topic though

OK, as I've noted earlier, PART of the reason your car goes faster when the engine (not the outside air) is colder depends MOSTLY on the fact that more fuel is delivered, AND altered timing.

Now, if you foul the computer into thinking that the engine is cold, supplying more fuel it'll likely NOT ignore the oxy sensor reading... Why do I say that? Cause the DME WANTS to run the engine richer, knowingly. I don't think it would run as rich as you guys are suggesting though, keep in mind this system WAS designed to be used with a Catalytic converter, they heat up if the mixture is too rich...

Now to address several issues, after further research, and a bit of head scratching, I think the best route may be to back off the spring tension, and drill holes (lots of em ), leaving a functional oxy sensor in place. Also, setting the DME for advanced timing, (through it's own adjustment), as well as more fuel (if you want) could be the way I go with my car, at least for now. The oxy sensor's reading though narrow, is more than enough to aid the motronic unit to the 'perfect' mixture, especially with the temp sensor hooked up like it's supposed to.

OK, now I've got two things to add, first of I still appreciate any and all comments, but many of the issues raised have been discussed before, and adressed.

Secondly, I'm thinking about a 550 spyder, with upwards of 140hp (1300lbs, with a regular battery, an optima will knock nearly 75lbs from that, and I'm sure that's a noticable difference)... And the idea of having more than 1.1hp for every 10lbs is, well, nice... If I do this, I be able to afford a 944turbo, or a 968, meaning I'd hang onto my 944 NA for a whole lot longer than I initially thought.

One more thing before I get too far off subject, I started my 944 today I think it's been a little over a week. It started in about a second, I'm assuming that means MY temp sensor works, good point to start from!
Ahmet

------------------
It's all the driver...

[lm6y]

It's nice to see we have a lot of thinkers here. It looks like the only way to make more power, is a forced air induction system, and a computer to deliver more fuel. Wait a minute... that's a 951!

[Kaos]

Ahmet,

There are 5 main variables that determine when to run the engine in open loop mode (when it ignores the O2 sensor input). These are: Engine speed, WOT position switch, coolant temperature, mass airflow, and time after start.

It is possible that the engine will start to run closed loop (using the O2 sensor) based on the amount of time elapsed since the last start. In this case fooling the coolant sensor will not really help. But, the control system may just use coolant temperature alone as a condition to go closed loop, in which case you will be in luck.

Most engines run open loop when at WOT.

If the engine is running in closed loop, and your mass airmeter is not reading the correct amount of air (lets say it is reading less air than actual), then fueling will still be at stoich due to the O2 control, but more spark advance will be applied.

The only reason aftermarket chips make any more horsepower than stock is because they run on 93 octane fuel. This allows the aftermarket tuners to add some spark advance without the engine knocking. They probably only change the amount of fuel being added minimally, as it is probably pretty close to optimal from the factory. The only time you really need to change the amount of fuel added is if your car is highly modified. Otherwise, the benefits are minimal (or nonexistent).

Kaos

[TheStig]

Ya'll early 944 owners know about Autothority? They sell conversion boards for about $375 on an exchange basis of your older DME. These boards are updated and chipped to advance the timing, extend the rev limiter and to alter air and fuel mixtures to up the torque curve. A little pricey, but its interesting....I think I'm gonna check into that. NOt to mention try out the drilling thing.

------------------
Heel'n' toe your woes away..

[justin]

Well Ahmet, still working on this? Whats new?

------------------
Drive it like you stole it........Justin

[1.2gees]

No, I haven't had time to work on it, sorry...

I'm making up for all the work I missed at school (when I was rambling about Porsches), so I can graduate, I went out on a trip out to bum f#ck on friday, got back sunday night, and just got done cleaning the garage...

HOWEVER: My parents are leaving for 5 weeks on may 21st, and coincidentally, that's when school gets out, guess what I'll be doing
(Besides the parties).

Oh, btw I just met a 944 owner that lives 4 houses down! I'll make a new post about that!

PS: Watch for updates, trust me there will be some.
Ahmet

[justin]

Where waiting Ahmet......don't keep us waiting long.....

------------------
Drive it like you stole it........Justin

[Kaos]

Here's the website regarding early 944 processors.


http://members.tripod.com/prsch944/dme.htm

[This message has been edited by Kaos (edited 05-04-2001).]

[This message has been edited by Kaos (edited 05-04-2001).]

[EGSMachine]

Or, you could put a small valve in line with the Fuel Pressure Regulator. the less vaccum on the regulator, the higher the pressure, up to 55PSI. at least, that's what i determined from the Haynes manual.

[EGSMachine]

i am concerned about the turbulance that drilling hols in the flap may cause. may be more detrimental than the flap.

[GTL]

The other side of the L shaped flap acts like a damper, thus minimizing the much higher engine induced turbulences. No problem drilling holes.

I am almost done with my bench test experiments. I cut a big square opening on the flap and carefully epoxied a fine grid mesh. With masking tape I am optimizing the optimum 'hole' area. What I have found so far (using a high flow industrial quality vacuum cleaner and a motorized throttle valve) is that at a certain tape opening and 9 notches down in the spring, the flow vs. flap opening curve is identical to stock on the lower third of the curve but starts going higher above that. At about 3/4 of the way (the vacuum cleaner does not have more power) the curver is above by an amount equivalent of moving down 2 notches with the stock set-up. So it will ran a bit rich in that area. So far I am very encoraged about the tests. 9 notches down is like dropping in half the air flow restriction.