Targeted Lambda vs load and RPM?

[jddoxtator]

Done some research and i fully understand the Ford ECU mapping points now. However, how the ECU applies specific lambdas to specific load and RPM ranges is still a mystery to me.

I see on the fuel tab that there is an option for setting the base calculation lambda and also a manual over ride lambda, but there does not appear to be any way to fine tune lambda in relation to both load and RPM.

As far as I'm aware, this is all handled in the back ground by the ECU based off that base point for lambda calculations. Manual override, I am assuming, ignores these calculations and just forces a steady state lambda, which is not what I want to achieve.

Now there is Power Enrichment that you can set with relation to load and RPM. But obviously that is only in effect when the conditions for Power Enrichment are met.

What I would love to be able to do is idle at around 1.0 lambda and cruise around 0.95 lambda.

I have set my base fuel number to 0.95 lambda, but from what I can tell it still targets 1.0 lambda. Though, I haven't done a full VCM RPM monitor run yet.

Manual override obviously works, but then you are just always at 0.95 lambda outside of Power Enrichment.

[engineermike]

Why do you want to cruise at .95?

[jddoxtator]

So I have been told, optimum power in a naturally aspirated setup is anywhere from 0.88 - 0.97

I figured cruise is where you return to and launch from most of the time, so being slightly rich here may help.

[engineermike]

I don’t think anyone takes that approach.

[Plimmer]

Try the table Open Loop then Fuel Base Cold table. I think that is the one used when not in PE mode.

[jddoxtator]

Doesn't really matter if no one takes that approach, being able to do it is the crux of the issue. It will enable finer tune control.

Yes, Plimmer. I think you are right. Fuel Base Cold seems to be what i am looking for. Handles all temperatures up to 115C and 87% load. There are certainly areas of the map that are concerningly lean with advanced timings. Cold launches could be dangerous.

Very simple change. Just adjusted all cells over 0.95 down to 0.95 and left all cell values below 0.95 intact to keep low temp fuel protections while extending protections into the higher temps and loads.

[engineermike]

The cal already pulls timing to compensate for lean condition.

[K44]

I've taken the same approach trying to eliminate some low speed knock. The engine seems much happier running a bit rich. Idles smoother and the AFR gauge on the dash is a lot steadier.

Still not sure this is the way to go but I have found no downside so far.

[jddoxtator]

Actually, if we add to the fact that the ECU adds timing based on richer fuel lambdas, running at 0.95 lambda should improve torque at lower loads and RPM ranges as the fuel lambda table will increase timing.

Relying on the fuel lambda table to pull timing may be problematic because it doesn't do anything until things are leaner than 1.0 lambda. Which may or may not be a problem depending on spark advance set.

Actually interesting thought. Does anyone know why the Fuel Lambda table actually starts pulling timing when fuel goes richer than 0.82? Is this just to keep torque numbers under control? A built in limiter of sorts?

Edit: Yep that did the trick! We are working with real meat and potatoes now boys! You can hear the advance correction occurring at idle as it is a little more poppy sounding, less sowing machine sounding. Zero knock retard. Torque throughout the lower range should be improved.

[engineermike]

Running .95 lambda would actually run a little LESS timing at light load IF the spark source is borderline, but that’s not going to affect actual torque response because the logic accounts for that in the torque control.

The lambda borderline table will actually pull timing when going from 1.0 to .95 (the starting point before knock advance comes in), then start adding it back as you go richer. The lambda borderline modifier isn’t there to manipulate torque or limit anything; it’s there because it’s a characteristic of the engine. At light load, you are probably running mbt anyway, so the borderline modifier won’t affect it at all.

In fact, there is an mbt lambda modifier as well, so if both borderline and mbt drop a degree then all you’ve accomplished is a 5% reduction in fuel economy.

You’d have to log spark source, mbt, borderline, torque request, and brake torque to really know how all these are interacting.


I don’t think the ford logic works the way you think it does.

[jddoxtator]

Is the borderline spark Lambda and MBT spark lambda not an inverse relationship?

If not, it means the ECU is actually adding timing in the much leaner lambda 1.1 rows, which makes no sense from a safety standpoint. It would make sense to advance down low where the numbers are positive at 0.7 lambdas, but pulling timing in the middle would make very little sense unless you goal is to pass that fuel through for some other purpose?

This is why I figured the table was an inverse relationship. But it does kind of make sense if they are thinking the only reason to increase fuel is to cool the cats and exhaust sensors. Advancing timing if that is your goal would not make sense then.

Edit: I have noticed something in particular. Lambda efficiency table in Torque Model seem to line up with the Lambda borderline and MBT spark tables. Apparently this is a multiplier for the torque table based on current Lambda. If this is correct, it is telling the ECU to expect more torque in the slightly richer lambdas.

Table values
Lambda multiplier
0.70 0.993
0.80 1.019
0.85 1.028
0.90 1.029
0.95 1.020
1.00 1.000
1.10 0.938

Would this not suggest that as Lambda is decreasing, the ECU is expecting more torque? Obviously super rich Lambdas lessens combustion temperatures and efficiency, so the drop off at very rich makes sense.
So the question is, where is this torque coming from then? If it is not the Lmbda tables allowing for more advance is it just the sole effects of the fuel or does the increase multiplier to the torque actually affect torque in some unknown way?
I just find this very unintuitive if this is working the way you are saying it does.

Edit 2: Ok, compared to a first gen Raptor ECU and lambda values are indeed direct not inverse. Unless something has changed in ford ECU's in those few years you should be correct.

Edit 3: Tried out a Lambda spark table in the direct inverse. It did lead to some spark retard on 0 to 60 pulls about 4.5 degrees, So i just flattened these adder tables out to zero.

[veeefour]

MBD and BRDL inverse? No no, think about one being the starting point, second being the cap. Most calibrations would
keep BRDL higher compared to MBT for low load and idle, as spark works in BASE there and that is the way to keep it like that.
Nothing to do with inversion whatsoever. Running 0,95 lambda or changing the stoich from 14.09 is not the best idea unless you ruining on "exotic" fuel.
MBT max values are equal to 100% Brake TQ ever produced at high load conditions. Idle and low load works under different scenario.

But there are multipliers as mentioned...stock car stock injectors AFR stable around 14, no way it will ever be used as its easy to keep the lambda at desired level.

TQ multi for AFR is not even used besides cat protection events in stock car. This changes if you have big injectors that are not able to keep stoich at low load,
running rich at idle might trigger TQ cut solely because of THAT table. This something you need to address if you go F/I bic boi powah.

[engineermike]

Again, I really think you need to try to understand Ford's logic better. It will all make a lot more sense. These engine controls don't work like anything traditional.

Tuning these works best if you start out with 2 basic assumptions: 1. Everything in the calibration does what the description says (borderline lambda is NOT used as a tool for controlling torque, cat protect is used to enrich to cool the cats not some other mechanism, etc), and 2. Everything in the tune is calibrated accurately for the stock engine and expected fuel.

Therefore, the MBT lambda correction is used to accurately calculate MBT timing. MBT timing is used when it's the lowest calculated spark lead and will lead to the maximum efficiency. It's also used to calculate brake torque at all times, as the "distance" from MBT is used to look up the torque ratio. So, if your MBT lambda correction is wrong then the brake torque calculation is wrong as well, which can lead to a host of other issues. The Borderline lambda correction is also very accurate from Ford and zero'ing this out will only cause other problems. Rather than making it what you think it should be, it would be better to understand why it is the way it is. Whatever changes you're trying to make are better done where they are meant to be done, rather than destroying accurate tables that Ford provided.

Lambda MBT and Lambda Borderline corrections are not inverse of each other. In some cases, they track somewhat and in others they don't. These are accurately calibrated to each engine.

Gasoline engines generally make more torque at .85 lambda than they do at 1.0. You see that in the lambda efficiency table. However, it might pick up 3% torque at .85 while using 15% more fuel, so don't think the fuel economy will be improved. Secondly, the entire premise of the logic is that it is controlling torque. So unless you are at WOT and WOT start/end is forcing the blade open, then the torque is going to be what you request (unless you've messed up the calibration). So at anything below about 90% throttle, nothing you do with timing and lambda control will affect the engine torque output at any given pedal position. This goes back to understanding the Ford logic. If you want more part throttle torque, then there is a place to get that, but trying to fool it into make more part throttle torque by manipulating spark and lambda isn't the right way and will cause other issues.

You seem to think running 1.0 lambda (or 1.1) is dangerous. It's actually not, as long as the calibration is accurate. GM started running 1.0 lambda at WOT decades ago. Yes, the borderline table adds timing at 1.1 lambda because there is less fuel available and less heat in the chamber to cause knock. Yes, it takes timing away at .85 lambda because that's usually where cylinder pressure and heat peak, and it makes more torque as a result.

[jddoxtator]

Alright, so if I want to gain anything. I have to do it in borderline tables and then make sure there is enough breathing room in torque tables and better to leave fueling and lambda tables at stock.

So essentially, the only changes that are required are are base calculation parameters when actual engine mods are done to the engine and the ECU will just do all the heavy lifting, for the most part?

For instance throttle body surface area, manifold volume and long tube header O2 delay.

Edit: back to stock fuel tables and just the mods from the worked example on HPA course. Plus the optimum power still being enabled. May tweak these borderline map setting depending on what I see.

[engineermike]

I think you're getting there.

You'll want to watch logged spark timing, knock retard, spark source, borderline timing, and MBT before making a lot of changes. What usually happens at WOT is the timing starts at borderline then starts adding knock advance until a) it hits another limit such as cyl pressure or MBT, or b) you get a knock event and knock advance backs off. Depending on which happens determines what your course of action is. If you're getting to MBT or knock already, then the best you can expect to do is get there sooner or faster for more initial torque after the hit.

[K44]

Mike, are the MBT tables from Ford accurate? While chasing some low RPM knock I added 100 octane fuel to rule out false knock. Now I hit MBT at WOT. Ford has it limited to 26. I have seen tunes with it around 30. Is 26 degrees MBT timing on the coyote?

Thanks and I hope this is not a hijack.

[engineermike]

K44 you know mbt is a bell curve so if you’re within a few degrees either way, it would be very hard to measure a difference. My best GUESS is that Ford defines mbt by “<X% torque reduction when timing is retarded from mbt”, which would err on the low side but still not sacrifice much of anything.

I too have heard of some running timing higher than mbt, but I’ve tested some high boost combos and found the same torque and power at timing values below what most big tuners run.