Is anyone running a OBD2 60* in speed density?

[robertisaar]

also, that device doesn't appear to support VPW, which is the protocal most GM OBD2 stuff used until the mid 2000s. CAN took over around that era.

[3400-95-Modified]

also, that device doesn't appear to support VPW, which is the protocal most GM OBD2 stuff used until the mid 2000s. CAN took over around that era.

yeah I was just noticing that when I was reading the specs... Bummer.

Found that J1850 is supported by the hardware but not enabled yet because it's in development... But I haven't seen much action on that website so who knows.

[Bridgetown]

not that i'm saying it's the end-all solution, but in tinytuner....

97 lumina (09355440) Fuel->MAF->MAF Failure Airmass Calc Mode

3 options are shown, including speed-density and alpha-n.

at that point, i imagine you could just either set the MAF min/max fail frequencies to something that will cause the alternative fueling mode to be entered, or just simply disconnect the MAF sensor and it should jump to that.

For the itb's from what im reading, and mabey you guys can offer opinions, is that alpha-n is the better way to calculate airflow. Would you guys agree?

[3400-95-Modified]

It appears that Alpha-N is a VERY crude setup that completely ignores and eliminates fuel trimming... So you would have to be monitoring your ECU all the time and make sure your tables are spot on...

I had to look this up to figure out what it was.

I don't know if this directly applies to our ECU's though...

Gruppe,

I know many folks on this board understand quite well what Alpha-N is, but this is for those who don't understand it. I remember how that felt, so I will offer my version of an explanation.

An engine management computer, regardless of how sophisticated or crude, really only controls two parameters. The ignition timing and the fuel delivery. The computer (DME or some aftermarket brand) uses an array of input sensors to try and determine what state the engine is in at any point in time. Is it loafing along on the freeway at low rpm, or at max throttle climbing a hill at high rpm? Is it hot outside so that the air is thin? Or is it butt cold and the air is real dense?

The list goes on, but the goal of the computer at all times is to sample its input sensors and then determine the appropriate output signals for ignition timing and fuel delivery (usually this determines how long the fuel injectors fire, which is called the pulse width, but it can also entail "when" to fire the injectors if the injection is sequential).

The two most important things that the computer wants to know in order to determine spark and fuel are: 1) engine load, and 2) engine rpm.
Engine rpm is easily determined from a crank sensor. But engine load is a little more tricky. The best parameter to use in determining engine load is the MASS of air entering the engine at any time. If you can measure the mass of air directly then you are well off. Current BMW's do just this with what is called a MAF sensor, or mass air flow sensor. This is usually some type of heated wire or film, which is cooled by the air flowing over it. In order to keep the wire at the same temperature additional electrical current must be provided, and the measure of this extra current gives a fairly direct indication of the mass air flow.

Our E30 M3's used a similar system, although it measures the VOLUME of air flowing into the engine. The sensor is called an AFM, which stands for air flow meter. The volumetric flow is determined by how far a flapper (or barn door) is pushed aside by the incoming air. The flapper is connected to a potentiometer (variable resistor, POT, wiper, there are many names). But for the computer to know how much fuel to mix with the air it needs to know how many molecules of air are coming in, and that can only be determined by knowing the MASS of the air, not just the volume. So, we need to combine the AFM air volume signal with an air TEMPERATURE measurement Then the air mass can be determined. So we are ok.

Note that in both cases the computer has some hard data on how much air is entering the engine, so if you improve engine breathing (header cams etc...), the basic code will still sort of work as the computer is aware of the additional air. It's not perfect, but it works ok most of the time for changes that are not too drastically different than the baseline.

Alpha-N is different. Here there is no direct measurement of either the mass of air, nor the volume of air entering the engine. That way any possible obstruction from either a MAF sensor or AFM sensor in the intake path is removed. The air can flow right into the engine unobstructed. So how does the computer know what to do? It still has rpm info from the crank, but it can't measure the mass of air coming into the engine for fuel mixture determination?

The answer, in a crude sense, is that you "train" the computer what to do in a given situation (on the dyno). For every possible combination of throttle position and rpm, the tuner determines the appropriate ignition timing and fuel delivery to yield max power but not incur detonation. The computer just remembers all this (in the form of maps which are stored on a chip), and when it sees a certain combination of rpm and throttle position in the field, it just says "what did they tell me to do in this situation?" and does that.

So now throttle position and RPM are the two dominant input parameters to the engine computer. This is where the Alpha-N name derives from. Alpha for the angle of the throttle plates and N for RPM.

Now if you have an Alpha-N system, and add a hotter set of cams that allow more air into the engine, but you do not reprogram the Alpha-N computer, then the computer has no knowledge of the extra air now entering the engine and it just gives the fuel and spark that it was told was appropriate for the milder set of cams. One can see how this could be a problem. Your mixture especially, but also your ignition timing will be way off. If your mixture goes lean you can potentially damage the engine. That is why folks say that an Alpha-N system needs to be retuned everytime you make a change to the engine. The computer is "flying blind" to some extent, although it does at least have an air temp sensor and a barro sensor to try to compensate for density changes in the air due to temperature and elevation.

I have oversimplified some subjects to make it easier to understand, and as usual there is probably more that I did not say than that I did say about Alpha-N, but hopefully that gets some folks a little closer to understanding it.

FWIW,
Gustave

[robertisaar]

with GM OBD1 applications, usually when a sensor fails that would cause it to revert to a backup fueling mode(but not limp-home mode), the fuel trimming reaction speed is actually multiplied somewhere between a value of 2 and 4. so it takes between 1/2 and 1/4 of the time it normally would to get the fuel trims correct for a given situation.

also, that post is what happens when you think the MAF is flawless and the greatest piece of engineering on a car. it's not. they're flawed, they're easily fooled, some are VERY prone to failure.

[3400-95-Modified]

also, that post is what happens when you think the MAF is flawless and the greatest piece of engineering on a car. it's not. they're flawed, they're easily fooled, some are VERY prone to failure.

I was really posting that for more of the information on the Alpha-N setup.

But I do have to say speed density is not flawless either, every setup has its flaws, all sensors can fail... I agree he has to work with either speed density or Apha-N but I was just advising what I found about it and how it doesn't appear to use trimming like Speed density would and it only runs at exactly what's tuned in the ECU. Just seems like a lot of attention is needed to ensure it's right and you have nothing to verify that against, then again it's kinda like reverting to a carburetor...

[robertisaar]

BMW would be Bosch Motronic.... which i have about 0 experience with, so i have no real comment on it.

it's similar to speed-density in that you have a 3D fuel map rather than 2D for MAF, but all 3 need 3D main spark tables. the actual difference in how much effort is required, i wouldn't think there would be that much more. just getting a good initial calibration setup is where i can see lots of time being spent.

[Superdave]

I love OBD1... lol


It ran really good in SD mode, I tried AlphaN but SD was always better. If you can't easily do SD on OBD2 then i'd have to agree that swapping back to OBD1 or aftermarket like a MS would be your best bet. If you go OBD1 i'll just give you a copy of the tune that was on it. I had a good sized cam and headers with 33lb/hr injectors so you'd have to adjust a few things.

[3400-95-Modified]

We've pretty much determined if you unplug the MAF OBDII will revert to full SD and not a limp mode... He was just asking if he should change that option to Alpha-N or not, and with your input of it being better is SD may tell him which is best route.

[Bridgetown]

I agree (based on my research anyway) that speed-density, is the more accurate and more adaptable air estimate, however, alpha-n does lend itself better to an itb application, solely based on the fact that the map signal is not as accurate with itb's. If I can get the itb's to provide a solid map signal, then I would prefer to use sd. Sd "should" require less tuning.

I have a wide band, and a data logging app on my tablet via Bluetooth. I'm going to hookup the itb's to my engine, sans maf, and see what happens. With the backup fuel mode set to speed density, I'll be able to get an idea how obd2 is going to cope with them. I'm hoping the initial data will give me an idea whether or not to stock with obd2 or degrade back to obd1.

That's not going to start until fall though, I have a road trip with the car at the end of the month.

Of course, a crazy little though has been creeping into my head lately. I do have a Rotrex supercharger that I was planning on hooking up to this engine, before Dave's itb's came up for sale. I could just build plenum and run itb's and the supercharger. Then I could keep the maf, and tune for boost. Would the boosted itb's be better then a boosted 3500 uim, I have no idea. I know that back in the days of turbos being allowed in F1, boosted itb's were all the rage.

Your thoughts?

[3400-95-Modified]

Sell the Rotrex to me problem solved...

[Superdave]

well, the stacks are Oring'd at the top so if you can get them to seal well to an air box... lol

I cut out big cork gaskets when i had my air box on top, you'd need something better if boosted for sure. Maybe a billet piece.

The MAP signal was jumpy at idle but that's normal with a big cam. It really wasn't much worse than the plenum. under load it was fine but the problems come in about 1/3-1/2 throttle when the MAP is reading 99Kpa. It's the same issues that you'd have when running too large of a TB. It just takes fine tuning to clear up.

[Bridgetown]

Sell the Rotrex to me problem solved...

Well, if your serious, send me a pm

[Bridgetown]

I just received some bad news today, looks like I'm going to be selling off some stuff, and these itb's will be the first thing to go. If anyone is interested here's a link to the sale thread:
http://60degreev6.com/forum/showthre...ck-up-for-sale!!!

This sucks...

[Superdave]

That sucks man, i was hoping to see them poke out the back of a Fiero!


I'll spread the word.