I know we can't seem to crack the code on our Hitachi ECUs. But what if we could influence one of the sensors so that we get the ECU to look at another part of the mapping?

My Assumptions: Our cars are tuned to run a little rich, so we're not able to optimize the fuel curve to increase power or timing.

My Hypothesis: (This is for the fuel curve only and won't impact timing.) What if you created a small velocity stack to install in front of the MAF sensor? This will gather more air and move air more quickly over the wire, therefore the computer will think more air is flowing and will adjust the fuel (and possibly timing.) This is a SWAG, so I don't know if it will work or if I have it backwards. (Maybe less air will help lean-out the amount of fuel in each injector pulse?)

I don't know... it's a wild-haired thought that hit me in the middle of a work-day (when I SHOULD be working! :-P )

Bob K.

<center><img src="http://www.revolutiontuning.com/prod_images_large/safc2.jpg"></center><p>
Allows + or - 20% fueling, over stock ECU mapping!

Showed "positive" results on my 90 V6! ;O)

I had one out of the box and strewn across the counter at a store I was in yesterday checking out it's hardware and instructions. Love what it says it does but hate it's looks. Anime in a box! It may come to that. But I thought it also had a feature for timing adjustment too. Or is that the difference between AFC1/2? They had both and I was thoroughly confused by the time I left. Trying to get it all back in the box without bulging it contributed as much to that as the wiring schematics/installation instructions. I killed a WHOLE LOTTA cells in my youth.

http://www.apexi-usa.com/productdocumentation/electronics_safc_us_update.pdf

http://www.apexi-usa.com/productdocumentation/electronics_safc_wire.pdf

http://www.apexi-usa.com/productdocumentation/S-AFC_Diagrams.pdf<ul><li><a href="http://www.apexi-usa.com/productdocumentation/electronics_safc_ins.pdf">http://www.apexi-usa.com/productdocumentation/electronics_safc_ins.pdf</a</li></ul>

There are several of these type devices out there, on the open market.<ul><li><a href="http://www.huntleyracing.com/massairflow.htm">http://www.huntleyracing.com/massairflow.htm</a</li></ul>

<center><img src="http://www.valley.ne.jp/~sirogane/2438a.jpg"></center><p><ul><li><a href="http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&amp;item=2411435939&amp;category=33 557">http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&amp;item=2411435939&amp;category=33 557</a</li></ul>

<center><img src="http://ebay0.ipixmedia.com/abc/M28/_EBAY_5f6d5720b818d00f5d36b6fdb94a2ef4/i-1.JPG"></center><p><ul><li><a href="http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&amp;item=2412023561&amp;category=33 596">http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&amp;item=2412023561&amp;category=33 596</a</li></ul>

There are 2 versions. I like the version one better than the newest one.

The MAF wires and calibrated module are sensor only. That is to say they can only sense rather than control. The control is handled by the ECU which receives 0-5v signals from the MAF based on airflow past the sensor wires. The ECU then controls the fuel based on it's pre-programmed fuel maps.

At cold idle the MAF's output voltages run in the 1.65v range. Once warm it backs down to 1.35v. At no-load WOT blips it'll output 3.3-3.65v. But at load it'll output up to a maximum of 5v when climbing a hill under WOT.

Here's where the plan goes to hell in a banana boat: At 4 bar fuel pressure settings with our stock injectors the ECU cannot make the car run any richer than it did at about 4.7v output from the MAF sensor. In other words at 4.7v under load the car is at full rich compliments of Hitachi (ECU &amp; injectors). Making the MAF output 5v by climbing a hill doesn't make it any richer than at 4.7v. It relies on it's pre-programmed safeguards (knock sensors) at that point to protect the engine if it were not rich enough to fuel the situation, ie; it backs off timing so that the fuel it IS delivering is within safe parameters.

Thus more air thru the sensor can likely richen it up at (sub 4.7v outputs) low and off-idle RPM's where it's least needed but not at higher RPM's where it's most needed.

Thats why fuel pressure increases along with a nice set of balanced/blueprinted injectors in the 187-197cc/min range will likely provide stellar results on a non-invasively modded car, ie; intake/exhaust mods only. The ECU is dumber than a stump when it comes to reasoning. All it knows is what it's been taught. If you trick it by upping the fuel pressure with a good set of injectors it plods along, happy as a clam outputting all it's old values as far as injector pulse-width, duration and flow rates etc, thinking it's stock. But now those flow rates, pulse-widths, durations are a bit greater, wider and a bit longer which means more fuel. Sure the O2 sensor will try to compensate while in Lambda or closed loop and if someone has done their homework in fuel psi settings and injector sizing it'll stay remarkably close to stock/stoic at all but WOT applications thanks to the O2 sensors. Likely as good as it was before provided someone did their homework on injector sizing. At WOT the computer will still be preaching the same sermon, but this time to an audience consisting of higher fuel psi and slightly larger injectors. It will be running richer to make better use of it's new-found air/fuel through all the recently-touted induction improvements. And all the inherent safeguards that came with the car remain in place. The car still has knock sensors. You make more power thru more fuel/air and the ECU is happy to oblige.

Now if anyone knows a way to similarly "lie" to the timing maps I'm all ears.

requires a laptop for programming though, but at least you can stash it away under the dash somewhere.

Has an RPM vs. MAF input voltage table, and you can adjust the MAF output in 0.025V increments.

The RPM points are 500 - 8000 RPM in 500 RPM increments, the MAF input points are -0.8V to 5.2V in 0.1V increments.

My only gripe is it is not programmable real-time. You can watch and see where in the table to make the changes, but you have to shut off the car to program the changes.

It costs $235 from Split Second.<ul><li><a href="http://www.splitsec.com/products/psc1/PSC1003.htm">http://www.splitsec.com/products/psc1/PSC1003.htm</a></li></ul>

I'm not sure about the difference in capabilities among the Split Second and A'pexi, but I got some great e-mail support from Split Second when I was looking into this. They at least attempted to answer my questions.

and you can adjust the flow +/- 50% at each data point. You can set two different values, one for part throttle, and one for WOT.

The split second has 16 RPM points, and can adjust the output +/- 2.5V.
It doesn't have a TPS input, but theoretically, you can still adjust separately, since WOT will flow more air at a specific RPM than part throttle will.

I've talked with them on the phone a few times. They've been able to answer all my questions, and have called me back when they said they would.

originally a 1.8T (now 2.0)

Wasn't that hard to install, the toughest part was the RPM signal. I ended up tapping into the cam sensor (and set the RPM trigger for 1 cyl)

Giving up a little sensitivity for faster and on-the-fly programming. And I ordered their ARM-1 air/fuel gauge as well but I'll likely mount it only semi-permanently or with an on/off switch until the controller is fully dialed in to my liking then stow it or turn it on only occasionally. The one in my UrS4 is enough to know it's too busy/distracting at night even when dimmed so I mounted an on/off switch in the pod beside it.

If I like this and get verifiable and reasonable HP/torque increases I'm going to hit em up on pricing for 6-12 units, make it an optional part of the induction kit.

Please document the install (wire colors &amp; locations). Unless you are going to sell a wiring harness, in which case I would gladly pay for it.

I've been wanting to do this for a long time, but can't decipher the Bentley wiring diagrams and have been very hesitant to do it on my own.

Scott

As I understand it the only leads used for signals are the ones coming off the MAF plug. Same for ground. Besides that theres only a 12v ignition source, dash-light dimmer. Appears to be a VERY simple and straightforward installation. I can handle that. It's the routing wires into the cabin and lying on the door threshhold hump and looking up under the dash that breaks my back... literally.

I bored a HUGE turbo inlet today for a friend who's running the Pikes Peak Hill Climb next month then built him a velocity stack restrictor sleeve for SCCA's 40mm restriction rule for his class. He asked "how much do I owe ya?" I said "nothing... I have something I want you to wire for me weekend after next." He said; Deal!!

He'll wire it in completely. I'll take notes of what goes where so I can supply a specific 2.8 12v "How-To" and list the parts/tools needed.

But it will be nice if someone else does it first!

The ARC2 was the unit that I favored, but I was going to get the A'pexi since that's what MikTip had and I figured he could give me some advice on using it. Now that you are getting the ARC2, I'm pretty excited. Time to get my car back together so I can take it apart again!

A little too late for me to switch back!

Maybe you can adapt a short piece of silicone hose between your MAF inlet and air filter... anything to lengthen the tract between filter and MAF. Likely even a half inch would help but 2.75"-3" is what I've found to be optimum. That's the reasoning behind the velocity stack MAF inlets I'm making for the stock airbox. Gives the sensor wires a much-improved quality of air to sample.

I still can't figure out why the people that make the MAF adapter plates for cone filters don't make their flange tubes 2.75"-3" long... this aint rocket science!

There's not too much room to extend the cone due to the black-plastic drain that get's in the way. I already had to trim some of it to make room for the new cone. It would probably take a complete removal to extend the cone any further.

I'm not too worried about it though. I probably won't do anything unless I see a major problem.

Thanks,
Scott