Mechanically Changing the A/F Ratio
 

I had a brainstorm the other day (okay, more like a braincloud) and was wondering what y'all thought. A big performance and fuel economy problem is the air fuel ratio...our cars run pig rich. I did some runs with mine and it was in the 10.5:1 range at WOT. So, to get more power, we need more air or less fuel. Canzoomer has come up with an electronic solution, but the RX-8 ECU seems to get in the way too often.

So, here's what I'm thinking. The cruising A/F ratio is determined by the 02 sensor which likes to keep things at stoichiometric. The WOT ratio is determined by the MAF with fuel and timing curves determined by the amount of air coming in (along with a few other things). If we were to allow a certain amount of air to bypass the MAF, shouldn't this allow improved A/F ratios at WOT while not affecting the cruising A/F ratio?

I was thinking of a tube from the air box around the MAF and back into the intake. This tube could have a bolt that threads in and out to limit or increase the amount of air being allowed to pass (or some sort of more elegant adjustable valve).

Would this work? Am I off base? Any input is appreciated.

Thanks,
~ Matt

Keep your day job :-)

Hey, my day job is as an engineer....

So, what would be wrong (or right) about my idea? In my old car (Volvo 1800E) the idle is regulated by a similar arrangement with air flowing around the throttle plate being regulated by a screw.

Matt

Why dont you try it out ?, it cant hurt , :) as they say what doesnt kill you can only make you stronger :D

B...free
michael

Just want to hear someone who is more knowledgeable about these sort of things to chime in (rotarygod, richard paul, etc.) before messing with my car. :)

So, in open loop is it still using the oxygen sensor to look at how rich or lean it is? I thought that it used the MAF readings and then applied a fuel curve based on that.

That would be interesting, but I don't know if it would work. Basically, all the CZ is doing is fooling the PCM about the amount of air flow it is getting, right? See how I changed the MAF on this graph using the CZ?, and how the AFR leaned itself out because it thought it had less air coming in? The MAF on the grpah is gm/sec divided by 3 to fit on the primary scale to the left.

Yeah, I was thinking that rather than electronically fooling the PCM, we could mechanically fool it.

Hmmm...this is from Racing Beat's site:

"The inside dimensions of the MAF tube have not been changed as compared to the stock unit, however, we have found that the stock unit is not actually round, making it difficult to determine the exact inside diameter of the stock unit! It appears that this may be due to manufacturing tolerances; we had to average the stock diameter. We did not intentionally try to make our tube a different diameter; we tried to determine the size as close as possible to the original unit. Changing the dimensions would effect emissions, power, and apparent mass airflow. The mass airflow output from the sensor is based on a given diameter. If you change this diameter, you change the meaning of the output. Although there might be performance benefits from increasing the tubing diameter, corresponding changes to the ECU programming would be required to account for this change."
http://www.racingbeat.com/FRmazda4.htm

Looks like they thought about this a bit. Or at least, about something similar.

Try richening up your mixture a little bit in the 6000-8000 rpm range. Shoot for around 12.8-13:1 You might pick up some power. You're a little too lean from what I can see.

Yep, I am lean up there. I have the CZ pulled right now, and will start all over after the reflash.

Hey Rotarygod...what do you think of my above suggestion of mechanically (rather than electronically) changing the A/F ratio by allowing air to bypass the MAF?

Thanks.

~ Matt

I agree with Abbid. The O2 sensor will compensate for the MAF changes and probably idle rough and the CEL may come on.

The stock ecu does some very weird and very backwards stuff. It might work or it might not but I doubt you'll hurt anything trying. Just make sure you use a very small orifice at first. It might actually get richer though so watch out!

Okay, so based on the responses above, this could work in theory, but this may not work in practice because of the ECU. However, no one knows for sure (except for the programmers at Mazda) so it'd be something that could only be found by trying.

A friend of mine has an OBD-II scanner which outputs all the details that I'd need (A/F ratio, rpm, etc.) so one of these days I'll look into how to do it. I'm thinking that I'll look at the A/F ratio at WOT, come up with how much more air on a cross sectional area basis is needed, make that the max bypass, and include a valve that I can slowly open up. I suppose a second intake tube might be needed to make this all happen.

I don't think I'll be doing this right away (not enough free time) but if someone else wants to try it or has an opinion on why this is or isn't a good idea, let me know! :)

I've thought about something that would drop voltage to the fuel pump, causing the engine to go leaner....but never fooled around with it. If someone trys this, and is successfull, I get a free unit!! CJ :D

When I had a 97 Mustang I used to lean or rich the A/F with an adjustable fuel regulator.

When I had a 02 Mustang, I used to lean the A/F electronically by putting a variable resistor in series with the MAF sensor, the higher the value of the resistor would drop the voltage from the MAF sensor, making the computer "believe" there was less air and therefore injecting less fuel. I used to monitor the STFT and LTFT.

For the RX-8, it would be really easy to add such variable resistor, it cost about a buck, and it surely has the capacity to drop the voltage from the MAF. I would like to try this, but I don't have any tool to measure the A/F ratio of my car, and I'm afraid of going too lean.

Wheter it is done mechanically or electronically (with a resistor), there won't be a fine adjustment for every rpm, like done with the canzoomer. But it would be cheaper.

if this method works you should be able to make it more advanced, like you could make the response of the circuit dependent on the value of the sensor reading (non linear) while this would not be rpm dependent it would be close at WOT

or you could use a microcontroller that can read the rpm from the computer (probably the hardest part) and has an input for the sensor voltage and then use a look up table to output your MAF readings

man I wish I was still in school I would do something like this for one of the project courses, any 4th years EE's looking for a term project?

it seems like this could work on almost any car as well as long as the method of reading the RPM (OBD-II) is standardized.


Anyone know how the car computer calculates the AFR? Becasue when you in closed loop mode the computer would likley counter the adjustments you make, but in open loop mode it would take effect.