Engine runs best reading way to rich in wideband

[Mykk]

Just throwing this out there looking for hive mind genius to shed some light and insight on my tuning situation.


You know when tuning a carby you get the engine up to operating temp and then you equally turn in or out the idle fuel screws until the engine is idling the smoothest at it's highest rpm. Then you keep going until it starts to run rougher and then go back until it's smoothest again.


I essentially did the same thing with the standalone EFI on my project:



Late 90's BMW 4.4v8, the long block is factory stock, stock cams but the intakes are -4° from OE timing. Exhaust cams OE timing. Jag x308 Supercharger assembly on adapters, Weber IWP 510cc injectors. Paired cylinder injection, waste spark

I locked out the ignition advance at 12 degrees, disabled closed loop and manually started adding VE value in the idle cells until the idle was the smoothest, the exhaust note sounded best and the rpm was naturally the highest without changing the amount of idle air.

Once finished I looked at the wideband gauge. 12-12.3:1 AFR.

Unplugged the LSU4.9 and removed from the exhaust pipe, recalibrated the wideband to free air, reinstalled the sensor. Fired up. Idled at 12:1 AFR. Using the wideband gauges software I reset the parameters to default and updated the firmware (Innovate MTX-L+). Idles at 12-12.3:1 AFR

When I manually lean out the VE the idle starts to get worse around 12.5 AFR, when richer the idle starts to get worse around 11.7AFR

At idle with the 12:1 AFR it's not overly fumous, noticeable but not going to make you pass out on the ground.

I do believe in giving the engine what it wants, but at 12:1 AFR I'm afraid of washing the rings out or perpetually fouling plugs.


Speaking of plugs. I'd say the AFR ring has recognized the rich condition as well:





I haven't had an opportunity to see if it universally likes the rich AFR during all conditions, during cruise or power. But I'll experiment with that as soon as can too. I can tell you at lambda 1 during cruise it won't tolerate more then about 24 degrees of ignition timing until it starts making some knocking type sounds. (I say Lambda 1 because I do use the 14.7 AFR = lambda 1 scale despite the actual stoich being around 14.1 of our E10 fuel)

Thanks for checking out my dilemma, excited to hear your insight and feedback.

Edited by Mykk on Wednesday 13th November 15:41

[GreenV8S]

That's a neat installation. I use the same blower, but sadly without the beautiful Jag charge coolers.

I use an Innovate wideband AFR system as well. Don't take the readings at face value. Over the years I've used it, I've had two sensors fail in a way that gives plausible AFR readings that are just wrong by about 1 AFR, and I've heard of others seeing a similar problem. With that said, if you're seeing evidence of rich running on the plugs, check whether you may have air distribution problems. If you have cylinders runnng at different mixtures that would explain why you need to run everything rich to prevent the leaner cylinders from complaining. Have you confirmed the injectors are all flowing cleanly at the same rate?

My own experience of sequential injection is that it makes almost no difference at all compared to batch fired, but I suppose engines will differ and I suppose it's possible to have an engine be sensitive to fuel timing - that could be another reason why it might want to be slightly rich to prevent a cylinder from complaining.

[Mykk]

GreenV8S said:

....Over the years I've used it, I've had two sensors fail in a way that gives plausible AFR readings that are just wrong by about 1 AFR, and I've heard of others seeing a similar problem. ....

Thank you for the response, I've had LSU senors fail by giving intermittent lean spike readings. I can swap sensors and test.

I'm thinking of digging out the old gas analyzer sniffer to see if O2 percentage of the exhaust on the analyzer matches the wideband or not, I have a feeling they won't match.


I was also considering injectors, I've measured a dead time of around .6ms on the flow machine and in the vehicle it idles around 1.2-1.3ms of PW.

As a test I was going to lower fuel pressure and increase PW to see if idle cleans up while leaner.

Edited by Mykk on Wednesday 13th November 18:29

Edited by Mykk on Wednesday 13th November 18:30

[stevieturbo]

Really, it's a bog standard engine, there is little reason it should not idle perfectly well at or close to stioch.

And on a V, it's often worth having a sensor in both banks, so you can see what both are doing, rather than guessing.

but if you have no emissions rules to abide by, and doubttul it will be idling for hours on end, you could largely do what you want.

but considering most n/a engines will make peak power around high 12's or maybe low 13's....idling at 12:1 does seem rather pointless and unnecessary.

Of course, the other aspect is are you sure your readings are correct ? ( again, 2 sensors can be helpful there, even for idle aiming towards stioch, a narrowband would do )

[stevieturbo]

Mykk said:

I can tell you at lambda 1 during cruise it won't tolerate more then about 24 degrees of ignition timing until it starts making some knocking type sounds. (I say Lambda 1 because I do use the 14.7 AFR = lambda 1 scale despite the actual stoich being around 14.1 of our E10 fuel)

Thanks for checking out my dilemma, excited to hear your insight and feedback.

This sounds like something is very wrong then. It should cruise all day at stioch, light load like that will want a decent amount of advance, 30-40deg

I suggest you verify your timing is what you think it is, and if need be ensure cam timing again is where it should be.

[Mykk]

stevieturbo said:

....I suggest you verify your timing is what you think it is, and if need be ensure cam timing again is where it should be.

Thank you, I have verified and marked true TDC with a piston stop (double mark procedure) and have set up the trigger angle with a timing light. I'm confident the timing is accurate. I can check cam timing soon.

[tendown]

For the idling, try retarding the ignition from 12degs, then it should be more stable when less rich. You'll also probably need more air via the idle valve/whatever you're using.

[stevieturbo]

Mykk said:

Thank you, I have verified and marked true TDC with a piston stop (double mark procedure) and have set up the trigger angle with a timing light. I'm confident the timing is accurate. I can check cam timing soon.

What ecu and ignition setup ? And proper sequential or batch fire/wasted spark ?

[Caddyshack]

Such a lovely engine, I would go for throttle bodies and get it mapped by someone who knows the ecu inside out.

[stevieturbo]

Caddyshack said:

Such a lovely engine, I would go for throttle bodies and get it mapped by someone who knows the ecu inside out.

ITB's pointless considering it's supercharged.

[Caddyshack]

stevieturbo said:

Caddyshack said:

Such a lovely engine, I would go for throttle bodies and get it mapped by someone who knows the ecu inside out.

ITB's pointless considering it's supercharged.

How about throttle body?

I was more thinking about accurate ability to control the fuel more than a "dumb" carb?

I asked Mark Shead why it was worth putting ITB’s on my Cossie lump and he said it gave the ability to map for each cylinder, is that not the same for supercharged? I assume the inlet etc could flow more or less per cylinder etc or is it really marginal? He also seemed to suggest it might give snappier throttle response?

I know that you know your stuff, I am trying to learn, not argue.

Edited by Caddyshack on Wednesday 13th November 20:09

[GreenV8S]

Caddyshack said:

I asked Mark Shead why it was worth putting ITB’s on my Cossie lump and he said it gave the ability to map for each cylinder

That's an interesting answer. I'd have thought the conventional answer is that it tames the cam by effectively preventing charge robbing between cylinders. I thought that even on ITB setups it was usual to only have a single common load signal.

[stevieturbo]

Caddyshack said:

How about throttle body?

I was more thinking about accurate ability to control the fuel more than a "dumb" carb?

I asked Mark Shead why it was worth putting ITB’s on my Cossie lump and he said it gave the ability to map for each cylinder, is that not the same for supercharged? I assume the inlet etc could flow more or less per cylinder etc or is it really marginal? He also seemed to suggest it might give snappier throttle response?

I know that you know your stuff, I am trying to learn, not argue.

I can only assume you are misunderstanding him. Any fuel injected car with an injector in each port ( pretty much all of them ) you can tune individual cylinders. How many throttle blades you have makes zero difference to that. Pretty much any ecu will allow you to map each cylinder for fuel and timing. Even with ITB's, you cannot open each blade independently for each cylinder....and nor would you want to. It'd be a nightmare
You can fine tune air balance at idle etc but that's about it really, which can help with idle if using bigger cams etc.


Although unless you've a means of measuring what each cylinder wants ( whether EGT per cylinder, lambda per cylinder, or monitoring in-cylinder pressures, all of the above ).....you've no way to determine what each cylinder really wants or needs ( yanks will shout, read the plugs )
And at full and larger throttle openings....in terms of airflow, it's kind of irrelevant anyway as everything is wide open.

There is the possibility it can give snappier throttle response......although again, on a supercharged setup it's a moot point really as the blower will always be shifting enough air that I'm sure you'd never feel the difference.
And you could argue on a turbocharged setup as there may be inherent lag anyway....would you ever really feel any slight benefit as you're still waiting on the turbo. Maybe part throttle transitions/fast on off throttles where the turbo is still very active, but then if it's shifting loads of air anyway, that last small volume from single blade to inlet valve isn't a lot anyway

n/a, then ITB's are likely to make a difference.

[stevieturbo]

GreenV8S said:

That's an interesting answer. I'd have thought the conventional answer is that it tames the cam by effectively preventing charge robbing between cylinders. I thought that even on ITB setups it was usual to only have a single common load signal.

With ITB's at low loads, you get a very pulsating signal in each runner as pressure changes. This is really crap for a speed density MAP signal. So two options are usually offer a small link between all 4 to somewhat dampen/smooth the pulses out ( servo needs good vac too ), or some will just choose to ignore the MAP signal completely in the vac ranges, tuning off TPS vs RPM only, and once boost in the plenum is in positive pressure it starts to take precedence rather than TPS.
Or it can be a blend of all of the above.

Even cars with in intake with long runners and a single throttle blade, if you have a pressure takeoff from a single runner ( Subaru's fuel pressure ref port is like this ), it can be annoyingly erratic because of the pulses.
But measuring in the main plenum, it is much smoother.

I think even when Max had talked about is car with ITB's, he had created a system whereby he only took manifold pressure readings at certain points throughout the 720 cycle in each runner when the intake valve would be open. Sounds awkward and complicated that !!!

[Caddyshack]

stevieturbo said:

Caddyshack said:

How about throttle body?

I was more thinking about accurate ability to control the fuel more than a "dumb" carb?

I asked Mark Shead why it was worth putting ITB’s on my Cossie lump and he said it gave the ability to map for each cylinder, is that not the same for supercharged? I assume the inlet etc could flow more or less per cylinder etc or is it really marginal? He also seemed to suggest it might give snappier throttle response?

I know that you know your stuff, I am trying to learn, not argue.

I can only assume you are misunderstanding him. Any fuel injected car with an injector in each port ( pretty much all of them ) you can tune individual cylinders. How many throttle blades you have makes zero difference to that. Pretty much any ecu will allow you to map each cylinder for fuel and timing. Even with ITB's, you cannot open each blade independently for each cylinder....and nor would you want to. It'd be a nightmare
You can fine tune air balance at idle etc but that's about it really, which can help with idle if using bigger cams etc.


Although unless you've a means of measuring what each cylinder wants ( whether EGT per cylinder, lambda per cylinder, or monitoring in-cylinder pressures, all of the above ).....you've no way to determine what each cylinder really wants or needs ( yanks will shout, read the plugs )
And at full and larger throttle openings....in terms of airflow, it's kind of irrelevant anyway as everything is wide open.

There is the possibility it can give snappier throttle response......although again, on a supercharged setup it's a moot point really as the blower will always be shifting enough air that I'm sure you'd never feel the difference.
And you could argue on a turbocharged setup as there may be inherent lag anyway....would you ever really feel any slight benefit as you're still waiting on the turbo. Maybe part throttle transitions/fast on off throttles where the turbo is still very active, but then if it's shifting loads of air anyway, that last small volume from single blade to inlet valve isn't a lot anyway

n/a, then ITB's are likely to make a difference.

Thank you