84-89 Idle Ignition setup changes and ideas

[scarceller]

In the 84-89 911 cars the idle ignition is set about 3 degrees advance (BTDC).

But I have done some interesting testing and want to get a few thoughts on it.

The goal is to idle at 880RPMs and find the optimum ignition advance. I have a chip emulator in my DME and can change any map in real-time with car running so this makes trying different idle ignition values easy.

Then I also have a WideBandO2 system that monitors AFR as well as amount of intake air by measuring the 0-5vdc output from the AirFlowMeter.

So here's the testing:
- Set idle ign at 0deg (TDC) and I see 11.2 Liters/sec of intake air at 880RPMs
- Then I start advancing timing and at 10deg advance (BTDC) I see 9.8 Liters/second for the same 880RPMs.

Pushing ignition advance any further past 10 deg does not lower intake air, the sweet spot seems to be 10 deg advance.

I find this method of setting up Idle Ign interesting because it helps optimize/pinpoint the best ign timing for the engine.

If we think about this approach, we need to achieve 880RPMs (the target idle) and if you can do this with the least amount of intake air then we have hit found the very best ign advance. If we fire the ignition to late (after TDC) then we waste the energy from the air/fuel mixture. And if we fire to soon (BTDC) we also waste energy and strain engine components.

Any thoughts on this type of tuning? I think one can learn a lot by looking at Intake Air for the given unit of work which in this case is to idle the engine at 880RPMs.

I think you can also apply this type of tuning to other parts of load maps.

[techman1]

Sal,

Are you also measuring injector firing / gas consumption? Do you also see a drop in injector pulse width? (is that how the dme controls volume?)

[scarceller]

techman,

Simple answer is yes.

The DME calculates fuel directly from air flow, so if air flow goes down from 11.x L/sec to 9.x L/sec so does fuel. In my testing I also look at AFR and my AFR is setup to 14.2 and all testing is done in O2-Open-Loop (no O2 signal feedback).

As I advance timing from 0-10 Deg I see air flow dropping till I hit 10 deg adv. I also see AFR staying at 14.2 the entire time.

Originally Posted by techman1

Sal,

Are you also measuring injector firing / gas consumption? Do you also see a drop in injector pulse width? (is that how the dme controls volume?)

[techman1]

Sounds like very efficient research.

I am guessing your observation is due to more efficient burn of the mix, so are you seeing a increase in egt because of this? I know, hard to measure at idle, but I am just curious.

Thanks for all you contribute!

Neal

[scarceller]

Neal,

I don't have EGT sensor (something I'm looking to add). But I suspect you may be correct that EGT goes up as intake air goes down but I can not confirm this at this time.

I think the theory is solid, What I think happens here is that you want to achieve full flame-front burn across the cyl so that it reaches full burn at around 10-15 deg ATDC this gives you max cyl pressures.

If you light the fuel mixture to late (like at 0deg TDC) then the flame front does not achieve max cyl pressure till way to late, maybe 20deg or more ATDC and here the energy from the burn is simply lost.

This theory is simply to do the unit of work (to idle engine at 880RPMs) with the least amount of intake air /sec and thus the least amount of fuel.

The reason this intrigues me is that this same principle may work for PartThrottle map tuning as well.

Originally Posted by techman1

Sounds like very efficient research.

I am guessing your observation is due to more efficient burn of the mix, so are you seeing a increase in egt because of this? I know, hard to measure at idle, but I am just curious.

Thanks for all you contribute!

Neal

[techman1]

My thoughts also. Now that would be neat, but I definitely see a need to monitor egt/head temp.

But wait, think this out. More work happening in the cylinder, less unburnt gases going out the exhaust. On second thought, my guess is: cylinder head temp goes up a little, egt at the exhaust port goes down and overall unburnt gasses (got a tailpipe snifferA)) goes down. It would be interesting to see how exhaust gas composition is affected with teh changes you are doing!

Neal

[burgermeister]

I think you'd really like to know the maximum pressure inside the cylinder. Obviously the engine idles more "efficiently" at 10degBTDC. But I also bet peak cylinder pressures are significantly higher - it's the way the efficiency is increased.

And of course the margin for detonation is reduced. Engine could be running hotter, intake air could be hotter, more carbon buildup, more oil burning - lots of things could reduce the amount of advance the engine can tolerate. timing for your engine that day may not work for other engines another day.

Last, I imagine you are limited to an interpolated curve for ignition advance. I don't know how the tables look for a 911, but the idle advance may need to be compatible with throttle tip-in advance just off idle, which may not be 10 degBTDC for all conditions.

Probably pointless advice ... but hey, you asked for opinions!!

[scarceller]

This is good feedback.

The 84-89 cars have different ign tables for idle a PT, I'm only messing with the idle ign for the time being. No issues on tip in since this would only apply to the PT Ign.

My only observation was that factory sets 3BTDC and I found going to 10BTDC gave me smoother idle and I noticed 10% less intake air required for 880RPMs. The way I figure is that I just improved fuel economy at idle by 10% but hey this is only at idle and most likely no one cares.

Originally Posted by burgermeister

I think you'd really like to know the maximum pressure inside the cylinder. Obviously the engine idles more "efficiently" at 10degBTDC. But I also bet peak cylinder pressures are significantly higher - it's the way the efficiency is increased.

And of course the margin for detonation is reduced. Engine could be running hotter, intake air could be hotter, more carbon buildup, more oil burning - lots of things could reduce the amount of advance the engine can tolerate. timing for your engine that day may not work for other engines another day.

Last, I imagine you are limited to an interpolated curve for ignition advance. I don't know how the tables look for a 911, but the idle advance may need to be compatible with throttle tip-in advance just off idle, which may not be 10 degBTDC for all conditions.

Probably pointless advice ... but hey, you asked for opinions!!

[Eagledriver]

You have found the optimum advance for maximum torque at idle. There are no worries about temperatures or cylinder pressures, you're at idle so they are trivial. You are correct that this will provide maximum fuel efficiency at idle. There maybe higher levels of CO or NOX however. Probably they will still be way below required numbers as long as the engine is in good shape.

-Andy

[ischmitz]

Sal,

Just for reference the advance on the 964 3.6 is right around 0 degrees during idle. Having played with MS-II systems a little bit trying to get stable idle I assume the reason for the slight retard over optimum is to provide a "knob" that can quickly be tweaked by the DME if the idle speed falls below the setpoint (undershot). If you set idle advance to the optimal 10 degrees you effectively rob the DME of this adjustment as a means to catch an undershot.

The engine reacts quickest to the ignition advance. If the ICV has to open you need to wait for the mixture to follow. In the meantiime the engine might have stalled. Just a thought.

Ingo

[Mark McClure]

This is very interesting reading but I am a bit confused.

I am thinking that getting the best burn at idle provides very lillte benefit due to the fact that it is IDLE ie not doing anything therefore there is no need for torque or really fuel consumption......not sure about everyone else but I limit the time I let my car idle for.

I can see the benefits of...

1. cleaner running and therefore less carbon build up,
2. less engine stress due to optimum timing,
3. better transition to off idle...maybe....?

If the method can be used across the whole rev range under load conditions then the re are obvioulsy huge benefits.

Mark.........

[scarceller]

Mark,

Bingo you got it, I started by observing the relationship between air flow and work being done at idle. But where I'm headed is to apply what was learned to Part Throttle tuning.

For example let's say the unit of work is to move the car down a strech of road in 3rd gear at 55MPH. I can easily advance timing and watch Air Flow drop for the same unit of work (55MPH in 3rd gear). Once you achieve the lowest air flow for the unit of work you just found peak torque.

This is not a replacement for dyno tuning but can certainly be used for limited street tuning to get in the ballpark.

Originally Posted by Mark McClure

If the method can be used across the whole rev range under load conditions then the re are obvioulsy huge benefits.

Mark.........

[911st]

Sal,

Way cool work that is based on real info. I look forward to any findings or conclusion you achive.

It might be that living with a little less than ideal idle quality and fuel economy is the price we pay for better throttle response.

We have a large manifold volume between the cylinder and the AFM that isolates and slows fueling response. With throttle we probably at first see a lean surge or condition until the system catches up with what is happening (shift from idle map and the AFM sensing a change in air flow).

Thus, we might find we get a little better throttle response if we run a little fat at idle so we are at an ideal power AFR when we blip the throttle. This was so significant with CIS that I could observe it on a Motec Wide Band moniter. I would lose 1 to 2 points of AFR before the metering function could respond.

I to also wonder if it is a valid strategy to improve throttle response with a quick change in timing from a late timing postition to an ideal timing.

Thus, maybe a less than perfect idle quality is the cost of performance.

Not sure, just a thought.

[911st]

Sal,

Can you educate me how much the stock ignition retards spark as the motor temps rise?

Thx.

[scarceller]

I have dis-assembled the 1989 program code and what I have found is it retards timing if Intake Air Temps rise - there is a map on the chip for this. But I have not found any retard based on CHT, this does not mean it doesn't I just have yet to find any code for it. I'm still looking at the ignition code and I have not seen anything to do this.

I have read and heard about this feature of timing retard based on temp and I think folks had the general idea but it's based on intake air temp from the intake sensor on the AFM and not CHT.

Originally Posted by 911st

Sal,

Can you educate me how much the stock ignition retards spark as the motor temps rise?

Thx.

[911st]

That sound right.

Do you know how much hoter intake temps pull timing?

I think the 91-94 C2Turbo can move timing plus or minus 3 deg. Just wonder what the 3.2 dose.

AFR also may effect ideal timing. A richer mix can burn faster to a degree I suspect.

Thx.

[scarceller]

The IAT Ign Table has 3 values like this:

Temp C = data
---------------
-10C = 20
45C = 16
65C = 14

What I have yet to figure out is how the code uses the values to influence timing. Also, from what I see in the code it is very possible this table is ONLY used to influence WOT timing and NOT idle or PT timing.

Originally Posted by 911st

That sound right.

Do you know how much hoter intake temps pull timing?

I think the 91-94 C2Turbo can move timing plus or minus 3 deg. Just wonder what the 3.2 dose.

AFR also may effect ideal timing. A richer mix can burn faster to a degree I suspect.

Thx.

[911st]

Thanks Sal.

Do we know if the delta in 'data' is equal to timing degrees? That is, is the differance between the high and low 6 deg of less timing advance?

That iwould be about the same as the 3 deg +/- on the C2t.

[scarceller]

It's to early for me to comment on the ign correction based on IAT or CHT. I simply am not that far along in the code dis-assembly yet. All I know is they do some ign correction based on IAT and they may also do some on CHT but not yet 100% certain and I don't want to quess.

But during cold start they do alter ignition based on CHT, this I know for certain. They actually advance ign to as much as 20 deg advance with real cold engine (like down near 0C CHT). This makes sense because cold fuel takes the flame front much longer to fully burn the fuel so they get the flame let extra early. The correction lasts till CHT reaches 60C or so after this the ignition is not corrected.

But your question is about ign advance under HOT conditions like hot IAT or CHT and this part of the code is still under investigation.

Originally Posted by 911st

Thanks Sal.

Do we know if the delta in 'data' is equal to timing degrees? That is, is the differance between the high and low 6 deg of less timing advance?

That iwould be about the same as the 3 deg +/- on the C2t.

[911st]

Sal,

Also wondering, there is three maps I think. Idle, normal, and WOT.

Is there a contact at the TB that trigers the move from idle to the normal map? I forget.

If so, is it possable to have idle set at say 14.5/1 and the normal map support an instant 13.2/1 AFR as soon as the throtle is first opened?

Thx again.