learn me ecm tuning

Dusterbd13 wrote: I'll be working with a map system. So if im understanding this, map should be a pretty stable value across the board, directly related to manifold vacuum. Higher vacuum, less fuel. Lower vacuum, more fuel. And as long as I can find a tune that is roughly the same cam and displacement, it will be close enough to run. Then tune map and fuel for air/fuel ratio desired at idle, cruise, and wfo in that order. Am I understanding this correctly? How do I figure out the map calibration if I can't find a similar setup? Say im starting with a stock tuned port vette tune, an engine that made maybe 200 horse at the wheels. Im going to change it to feed a 425 horse 370 small block. Obviously more volume and fuel per firing cycle, and obviously different vacuum values everywhere. At that point, do I just leave the map table alone and work on fuel?

So that more relates to more and less relates to less- change your thinking to pressure instead of vacuum. Less manifold pressure = less fuel, more=more. The math is a little easier to deal with. Besides, the pV=mRT formula uses absolute pressure and temperature, not relative. Manifold vacuum and Fahrenheit are both relative scales.

As for the base engine set up.. you have a 6.0 v8 engine that you think you can make 425hp. Which is just like a 3.0l I4 engine that theoretically can make 212 hp. I could go deep into that, but if you can find an engine that has a similar specific output, with the similar cam timing and number of valves- you can have a good starting map correction table.

That's one major elegance of the MS system. The voleff correction table is dimensionless. All it cares about is basic breathing- valve timing, number of valves, and rough valve area/displacement. If you could find a decent 2V 2.0l 4 cyl that has similar cam timing, take it's table and start there.

The rest of the engine dimensions, you tell the system what they are- and it should do the mass calculation all for you.

That also goes for the fuel injection- send those out, have them get flowed, and stick to that data.

Unless you are getting some really odd voleff numbers (say a lot of it over 1.0), it's easier to stick with one part you are reasonably sure about- and I'll suggest leaving the injectors alone. Change the air tables/functions/calculations.

Going back to the air estimation- that also can be modeled- which is why I asked about a free simulator that can give a decent estimate of flow when one puts in the data. So you can run some simulations and use that data to fill in the tables.

Do you have access to the code GM runs?

I do have the stock vette calibration, as well as the tuning software. I can send them both to you If that will help.

And with pressure: higher vacuum means less pressure, correct?

Also, can you break down the formula to what the letters mean? That might help. Math was the reason high school took five years.

Is your engine already wired up to the GM ECU?

I would really recommend moving to a MS based system. There is a whole community for support. (Unless that exists for your ECU already) I would bet you can find a similar tune on their forum today.

Yes. Already setup the harness, ecu, sensors, etc. For the gm ecm. Already paid for. I have thought about megasquirt, but there are notuners aaround here for it, and the extreme diy aspect of it makes me nervous for some reason.

Ah, ok. Megasquirt does use most GM sensors by default, just saying.

Dusterbd13 wrote: I do have the stock vette calibration, as well as the tuning software. I can send them both to you If that will help. And with pressure: higher vacuum means less pressure, correct? Also, can you break down the formula to what the letters mean? That might help. Math was the reason high school took five years.

the basic ideal gas law is

pV = mRT

p = the absolute pressure of the gas. So at one atm, it's 29.8 inHG, or 101 kPa, or 14.7 psi.
V = volume of the gas. Cubic inches, liters, cubic centimeters- whatever units you want to work in. m = mass of the gas- kg or lb. R = gas constant. Just a number- and it's value depends on what units you are working in. T = gas absolute gas temp- either in K or R. so at room temp- roughly 20C or 68F, it's 293K for metric or 528R. (0deg C = 273K and 0F = 460R)

So using that- you measure pressure and temp, and know the volume of the cylinder, you can calculate the mass of air in it. It does make one assumption, that the air pressure is same everywhere when the valve is open- but that's a good assumption.

And once you know the mass of air, if you know the air-fuel ratio you want to run- then you can do simple division to get the mass of fuel you want.

When you get into the hardcore thermo and gas dynamics- the numbers start getting messed up- but when you are getting an engine to run, one should not get hung up in that- and just hide those errors in the air estimation.

Having said all of that- don't get too hung up in it- that's just how the better computers do the calculations as opposed to simple look up tables. It means that with simple changes in some basic constants- most calibrations can be used from one engine to another.

BTW, you are going to have to wait until after the holiday for me to look at the software. Vacation and all...

ok, alfadriver, now that I have my laptop back up and running (hardware failures suck), I can finally send you some stuff.

first, the tuning software im using is tunerpro rt download at http://www.tunerpro.net/downloadApp.htm

and I will PM you the .bin and .xdf files for tunerpro. if the PM function is working for me today. im not seeing any tables remotely like what we are talking about, at least not labeled in a language I understand.

hope youre having a good vacation.

also, if anyone else wants the files, let me know.

actually, heres a link to the files: http://www.gearhead-efi.com/Fuel-Injection/showthread.php?394-1227727-1227730-ECM-Information-8D

im using the $8D.xdf (first one) and the AXCN.bin for my base maps. the stupid PM stuff wont let me attach files.....

so maybe, for my purposes in which I started this thread, we can get me up to the level of luddite. cause there is a WHOLE BUNCH of confusing E36 M3 in there. that makes no sense. conceptually, what alfadriver is describing does make sense, but when I go to open things up, I get completely bumfuzzeled and tempted to ragequit.

In reply to Dusterbd13:

I'm back from vacation- so go ahead and send me the files. it's the alfa(seventythree)@comcast(dot)net email that's in my profile. I'll try to get to the links later tonight- even though we have some things going on.

As for my discussions- I can put together a brief note about spark, but without a dyno, it's not nearly as easy as fuel is.

you have mail.

and id love to read up on spark. will have a knock sensor, so some safety is built in.

So to start with- need to reference some thermodynamics....

Again, not to worry about the actual physics, but if you do, great.

This is the ideal Otto cycle-

On that chart, the burning of the air/fuel is instant, at top dead center. But that just does not happen. It takes time for the spark to get the fuel to burn, it takes time for the fuel to do most of the burning, and it takes time for the last bits of fuel to burn. So a more realistic pressure-volume trace would look more like this-

And even that looks better than reality.

But the key is that you see that spark happens before top dead center, and the combustion finishes after.

The goal is to make the lower diagram to look as much as the upper one as possible- so that the work that the fuel can do is maximized.

And that last sentence is important enough to repeat- the work that the fuel can do is maximized.

I can go through a lot of more physics is people are interested, but I'm going to fast forward this to two different things- best spark and borderline spark.

Best spark is called MBT- which I've heard called a few different things- but for this, I'm going to label it Maximum Brake Torque spark. The work that the fuel can do and the torque produced are exactly the same thing- so MBT spark is when the work that the fuel can do is maximized.

Unfortunately, there are no good models that can even kind of predict where MBT spark is- chamber shape, piston speed, port design, number of valves, temps, turbulence, etc etc- all factor into what MBT spark will do.

When tuning MBT spark, what you do is advance the spark until the torque is no longer increasing. And most of the time, a couple of degrees of safety is put in that- since when reaching MBT spark- the amount of improvement is pretty flat- so when one finds that best of the best point, take 3 deg out.

If you have the luxury of tuning this on a chassis dyno- keep track of all speeds as close as you can- the spark is as real time as one can get, so you can tune the specific rpm as the engine runs up during a pull.

When running something less than wide open, the MBT does advance a little. At low loads, I've seen as advanced as 50 deg before TDC as MBT. And if you are going to drive the car on the street- this gain is important- as you will get a lot better fuel economy if you can run better spark at low loads. This part is a lot harder to tune on the street- as few dynos will want to do partial throttle runs. If you can- do it. You can run a couple of different throttle settings, and then draw a line between the different speed-load points to fill in the gaps.

Then there is borderline spark- when the engine just starts to knock. I'm not going to repost my old posts of what knock is, you all know it's bad.

Borderline will mean you can't advance the spark to MBT. Which means you just have to live with what the chamber will let you do. Like MBT, many factors influence where the knock will start- but compression ratio, temperatures, and turbulence have more influence here than for MBT- which is to say- make it more sensitive to knock.

Tuning it- well... when you are tuning- it would be good if you had a little switch handy that you can turn on when you hear knock- so that you can record the conditions you hear it. And this may sound odd- if you have a mic that you can put just over the engine- that is one of the better ways to tune knock. And tune the knock sensor. Knock sounds so different when it happens that you can have the sound low enough to not be a bother and still hear trace knock.

If you can believe it, that method is used when tuning the knock sensors mostly, too. That way you can know when the odd noise picked up on the sensor is real or not- a knock sensor is just a microphone anyway. It just touches metal so that the sounds travel quickly and directly to it. The reason many tuners tell you to be careful- the sensor picks up ALL noises (metal transmits vibrations really well)- so you need to know when it's knock or when it's just valvetrain noise. Or piston slap. Or whatever. And noting the PV diagrams I posted before, when knock happens is in a very specific time frame- during combustion- so you can make sure sensor is listening at the right windows.

BTW, there are now some reasonable pressure sensor systems out there for the home tuner. If you want to use one of those, that opens up a lot of possibilities in terms of tuning. Makes it quite possible to do a really good job of tuning spark to a couple of degrees of MBT and safe borderline without using a dyno.

So that's the long explanation.

Where do you start? If you have any spark data that someone else ran, start with it. Like on my Alfas, I would start with the distributor advance curves that someone else developed, and go from there.

Is there a way to subscribe to a thread, or favorite it for later reference, or can you just sticky this E36 M3?

In reply to gearheadmb:

There is when browsing on a mobile device (hit the Heart icon at the bottom of the page), but it appears that it is broken or missing for the PC version, at least on Firefox.

Thanks for posting alfa! This is a great read.

I too will be following...... not fully understanding, but following

In reply to oldeskewltoy:

It goes well with your porting thread.

This made my head hurt, i remember why my cars have always had a carb. But i like to learn so im trying to get the jist of it.

Nice job, Alfa! Even for someone who has tuned a few systems, the way you explained it is an awesome refresher, and it was explained in a much more practical way.. I wish you had written this up about 10 years ago! :)

dropstep wrote: This made my head hurt, i remember why my cars have always had a carb. But i like to learn so im trying to get the jist of it.

Carbs and fuel injection do exactly the same thing- estimate the air flow, and deliver fuel to match that air.

The way carbs do it is to take advantage of various designs of the venturi effect, and then various ways to meter the fuel to match that. If you can understand the why's of a carb, fuel injection is easy. Especially if you look into emulsion tubes as well as differing jets to try to be more accurate.

Carb designers were pretty clever.

alfadriver wrote: Carbs and fuel injection do exactly the same thing- estimate the air flow, and deliver fuel to match that air. The way carbs do it is to take advantage of various designs of the venturi effect, and then various ways to meter the fuel to match that. If you can understand the why's of a carb, fuel injection is easy. Especially if you look into emulsion tubes as well as differing jets to try to be more accurate. Carb designers were pretty clever.

Yep - I got into EFI after trying to figure out what it would take to modify a fairly common carburetor - something like a Holley four barrel or an Edelbrock - to maintain an accurate air/fuel ratio from cruising vacuum all the way to 15 psi of boost. I concluded that it would take some sort of precision micro lathe work, as the parts needed weren't available off the shelf.

So I don't crap up this thread... I've started a new one

https://grassrootsmotorsports.com/forum/grm/getting-the-tune-just-right-help-me-tune-the-fjo/108324/page1/

I'd greatly appreciate any assistance you can offer.......