idle control maps

[nickbmw]

A few questions that i can't find a definite answer through the help function.

Is the DC error correction table active when PID control is enabled or PID control overrides it?

What's the intended use of the idle RPM ref table? When do we nned to adjust it?

Regarding the PWM battery correction. When i turn on the lights and the coolant fan, the idle drops more than 100rpm from the target. I am watching the battery voltage in the data logging, but the voltage drop is very minour, 0,3V aprox. How do we correct this properly? Do i need to rescale the volt axxis with 0,2V increments between 13 and 14V?

[Karel]

Stock OEM Ecus are using "load" input from big power units like Headlamp, Powersteering, rear window heater, fan etc.....
you can use some correction for it too if you know what to do.
I am personally not a Friend of idle valves. I use basicly no idle valves on racing cars. I am using just ignition adjusted idle.

Why ?

IF you will fast responding idle control systém you will end with huge diferences in opening of the idle valve and it will affect you vakuum at idle, and this will end in huge changeas in VE during idle.... so it will easly start to change AFR and End in some unstable state. If you make huge IGN advance changes, it will nto affect your AFR at all.


Just idea.......

[nickbmw]

Stock OEM Ecus are using "load" input from big power units like Headlamp, Powersteering, rear window heater, fan etc.....
you can use some correction for it too if you know what to do.
I am personally not a Friend of idle valves. I use basicly no idle valves on racing cars. I am using just ignition adjusted idle.

Why ?

IF you will fast responding idle control systém you will end with huge diferences in opening of the idle valve and it will affect you vakuum at idle, and this will end in huge changeas in VE during idle.... so it will easly start to change AFR and End in some unstable state. If you make huge IGN advance changes, it will nto affect your AFR at all.


Just idea.......

It's a every day car, so not having an idle air control valve is not an option for me, ignition controlled idle is not sufficient, especially during warm up. My car is idling like stock in most cases. The coolant fan kicking in is easy to correct, there is a setting for this, but for other electrical loads like the soft top and headlights, the pid takes over to correct the idle rpm dropping, the PWM battery correction table is not enough to correct this because the voltage drop is barely measurable.
The problem with the pid is that when it is enabled, letting off from 3-4k rpm the idle drops to 650-700rpm very often before it stablizes at 850. With the pid disabled, this doesn't happen. I am not very happy with the way the pid works actually. If there was a time delay for the pid to activate after the idle control is active, i believe it would be much better. As it is now, i am watching how the pid acts when the rpm drops, and it is applying negative correction before the rpm reach the idle rpm target point. I am trying to improve this using the idle activation DC increase, but it doesn't seem to help at all. You don't feel the rpm dip when there are no loads because it is minor, but when there is load, eg coolant fan or A/C on, then it dips lower, to 500-600rpm, and it starts to become annoying. Raising the fuel resume point to 1400rpm is fixing this, if i go lower it will dip and then come up and stabilize with a/c on.

[Karel]

There is no problem how PID works... the problem is on your side.... if you cant adjust it, it will not work good enough.... I am also not the best by tuning the PID systems, so i am trying to do it nearly perfect on open loop control and then just allowing small adjustments for the PID systems.

Ok, another idea for you. Try to disable the PID and set manualy the idle valve duty for your desired RPM. After finding the value, you can use PID again but set the founded duty as minimum PWM. So the system will not be able to underrun the desired idle RPM....

understand ?

[nickbmw]

There is no problem how PID works... the problem is on your side.... if you cant adjust it, it will not work good enough.... I am also not the best by tuning the PID systems, so i am trying to do it nearly perfect on open loop control and then just allowing small adjustments for the PID systems.

Ok, another idea for you. Try to disable the PID and set manualy the idle valve duty for your desired RPM. After finding the value, you can use PID again but set the founded duty as minimum PWM. So the system will not be able to underrun the desired idle RPM....

understand ?

Μy idle is setup perfect at full open loop (igniton control and PID disabled), steady and nice at 850rpm with warm engine, with 5 degrees advance. I don't think it can be done any better. When i activate the PID it is never applying more than 1% correction with the deadband set at 50rpm, and the ignition control is barely adding or subtracting any ignition advance. I could change the pid parameters (prop., integral, deriv.) to make it faster or slower responding, but it is always a compromise, it is at the default settings at the moment.
With A/C on, i am adding 5% dc to keep the idle rpm target, and again PID doesn't correct more than 1-2%. But if i drop the fuel cut resume below 1400rpm, it will dip and then come up and stabilize. It will do it even with 1400rpm fuel cut resume sometimes.
The same with the coolant fan, adding 3% dc keeps the targer perfecty fine, with or without PID.

Ofcourse, there are chances i am missing a detail somewhere, and this is the point of asking questions, and discussing with other people.

[Karel]

OK last idea. What about to use analog input as correction input ?

Use so resistor divider to achieve like that:

Headlamp OFF, Coolant Fan OFF = 0V , correction 0%
Headlamp ON, Coolant Fan OFF = 1V , correction 2%
Headlamp OFF, Coolant Fan ON = 2V , correction 4%
Headlamp ON, Coolant Fan ON = 3V , correction 6%

again just example.

I think there is always a Way, its just question how to make it.....

[nickbmw]

OK last idea. What about to use analog input as correction input ?

Use so resistor divider to achieve like that:

Headlamp OFF, Coolant Fan OFF = 0V , correction 0%
Headlamp ON, Coolant Fan OFF = 1V , correction 2%
Headlamp OFF, Coolant Fan ON = 2V , correction 4%
Headlamp ON, Coolant Fan ON = 3V , correction 6%

again just example.

I think there is always a Way, its just question how to make it.....

Thank you very much for this, when i read your idea the first thing that came into my mind is that the S2000, like most Hondas, has an electrical load input on the stock ECU. From what i read online, it outputs 0-5V, the more the load the higher the output, so it should be perfect to use for what i need, but i will measure first with my multimeter to confirm.

[RHD]

if you have a "load signal" yes that should work well the problem is that the actual load comes about from the added resistance for the crankshaft to turn the alternator and unfortunately normally there is no way to measure or compensate for this.

You can set the PID so that it will only add positive compensation this should stop the rpm from dropping so low at times

The RPM DC ref table is what the idle valve does once the engine has moved out of idle control parameters.

The Idle error correction works all the time whether the PID is on or off, personally i find it adds a lot of instability and hunting up and down so i dont use it very often

There is a VE /DC correction table if your finding the afr changing with idle control, but thats usually only needed if you are using Alpha-n fueling

[nickbmw]

if you have a "load signal" yes that should work well the problem is that the actual load comes about from the added resistance for the crankshaft to turn the alternator and unfortunately normally there is no way to measure or compensate for this.

You can set the PID so that it will only add positive compensation this should stop the rpm from dropping so low at times

The RPM DC ref table is what the idle valve does once the engine has moved out of idle control parameters.

The Idle error correction works all the time whether the PID is on or off, personally i find it adds a lot of instability and hunting up and down so i dont use it very often

There is a VE /DC correction table if your finding the afr changing with idle control, but thats usually only needed if you are using Alpha-n fueling

Thank you for the nice response.
Ι checked the load signal with the stock ECU last night, it output around 3.6V with no load, and drops close to 1V with full load. The problem is, it doesn't work with the standalone ecu, there is no voltage. I talked with a friend, and we concluded that it needs a pullup and a pulldown resistor to work, which are inside the stock ecu on the stock setup.
Indeed, i have the PID set already to not be able to decrease the DC more than 2%.
How does the RPM DC ref table affects engine operation? I guess it only affects vacuum with closed throttle? At low rpm i have set at the same values needed for idle operation, but i think i will increase it a bit there so it will be 2-3% higher than the needed dc for idle
I have the idle error correction set at zero as well at the moment. I did experiment with it a bit, i will play a bit more with it when i have the time.
I have some variation in the idle afr depending on the iat readings. Now i am using the chargetemp table, but i don't really like what it does. It's rich until the engine reaches operating temp. If i don't use the chargetemp table for idle, then when the intake heatsoaks in traffic, it will become lean from false high iat readings and the idle quality goes down. I have requested another strategy, where the chargetemp table could be a blend between airbox/ambient temp and iat sensor, but noone seems to pay any attention. The current strategies are insufficient for very low loads and idle, it is impossible to have perfectly stable AFRs or close to stable in all conditions, ecu is very sensitive with iat readings. One of the first things i did was to use a fast response sensor and relocate it on the charge pipe (right in front of the throttle body)

[WHPZach]

Did you post your suggestion in the bugs and suggested features area on the forum?

Also, you should be able to achieve satisfactory results with the air charge estimation and the normal CLT/IAT trims. No model is perfect, ever, not even OEMs, so don't discount the value of closed-loop O2 feedback.

[nickbmw]

Did you post your suggestion in the bugs and suggested features area on the forum?

Also, you should be able to achieve satisfactory results with the air charge estimation and the normal CLT/IAT trims. No model is perfect, ever, not even OEMs, so don't discount the value of closed-loop O2 feedback.

I do use ego feedback, with custom pid values which make it respond much faster with very little overshoot. I agree, no model is perfect, the model which will offer the least afr variation will be the best.
I did 3 times, but i dont get any reply. This is what i have posted 3 times in the suggested features area, which i believe will improve things a lot and is simple for the developers to do :
You could offer the posibility for a second air temperature sensor, let's call it airbox temperature sensor or ambient air temperature sensor, along with the main iat sensor placed on the inlet manifold. So we can set the blend between these two sensors in the chargetemp table, instead of the coolant temp sensor. The coolant temp is more or less stable during all year and weather, so when we use the chargetemp table and set the clt sensor influence more the fuel calculations, we virtually just chose to ignore the iat sensor for whatever reason, and the afr is unstable. Two air temperature sensors will be much better for fuel calculations and can completely solve all the problems that come from iat sensors that are heatsoaking and give false readings to the ecu at low loads, rpm and speeds.

[dazjb]

You can set the PID so that it will only add positive compensation this should stop the rpm from dropping so low at times.

Is this done by changing the 'Max feedback + -' settings in the 'Idle PID Control' parameters section?

[nickbmw]

You can set the PID so that it will only add positive compensation this should stop the rpm from dropping so low at times.

Is this done by changing the 'Max feedback + -' settings in the 'Idle PID Control' parameters section?

Yes

[RHD]

How does the RPM DC ref table affects engine operation? I guess it only affects vacuum with closed throttle? At low rpm i have set at the same values needed for idle operation, but i think i will increase it a bit there so it will be 2-3% higher than the needed dc for idle


I have requested another strategy, where the charge temp table could be a blend between airbox/ambient temp and iat sensor, but noone seems to pay any attention. The current strategies are insufficient for very low loads and idle, it is impossible to have perfectly stable AFRs or close to stable in all conditions, ecu is very sensitive with iat readings. One of the first things i did was to use a fast response sensor and relocate it on the charge pipe (right in front of the throttle body)

Teh rpm REF DC is what the ICv does when the engine is OUTSIDE of idle parameters, ie above min tps setting and above idle control rpm etc... in other words when you are driving around normally

the air that enters an engine while ideling typically moves very very slow it will all fit through about a 5-6mm hole so i cant see airbox temp being relevant, thats the point of having a 3d charge temp table when the air is slow the temp will be more matched to CLT specially if the manifold is heat soaked and then when the flow is faster it will tend to sway more to the IAT. You can adjust the split with the 3d table until its works very well, there is also a separate IAT compensation table if you still require some more direct compensation.

[nickbmw]

How does the RPM DC ref table affects engine operation? I guess it only affects vacuum with closed throttle? At low rpm i have set at the same values needed for idle operation, but i think i will increase it a bit there so it will be 2-3% higher than the needed dc for idle


I have requested another strategy, where the charge temp table could be a blend between airbox/ambient temp and iat sensor, but noone seems to pay any attention. The current strategies are insufficient for very low loads and idle, it is impossible to have perfectly stable AFRs or close to stable in all conditions, ecu is very sensitive with iat readings. One of the first things i did was to use a fast response sensor and relocate it on the charge pipe (right in front of the throttle body)

Teh rpm REF DC is what the ICv does when the engine is OUTSIDE of idle parameters, ie above min tps setting and above idle control rpm etc... in other words when you are driving around normally

the air that enters an engine while ideling typically moves very very slow it will all fit through about a 5-6mm hole so i cant see airbox temp being relevant, thats the point of having a 3d charge temp table when the air is slow the temp will be more matched to CLT specially if the manifold is heat soaked and then when the flow is faster it will tend to sway more to the IAT. You can adjust the split with the 3d table until its works very well, there is also a separate IAT compensation table if you still require some more direct compensation.

The true/actual air intake temp is a blend between ambient temp and iat sensor temp, not a blend between the iat temp reading the coolant temp. The coolant sensor temp is being used only because it is a stable reference point to use to smooth out large changes and fluctuations of the iat sensor.
Airbox/ambient temp is very relevant to the actual intake air temp. The intake air temp sensor on the plenum can, and will output higher temps than the actual, during heatsoak, because the sensor itself absorbs a lot of heat from the component it mounts on, an aluminum intake manifold for example, and the airflow in the plenum is not enough to cool it down and make it read realisticaly, at low loads. But it will rarely/almost never output temps lower than the actual. When moving slowly in traffic and idling, it will output much higher temps than actual. How do you know it's actually higher? Because the afr is getting lean, which shows that the iat reading is false and is throwing out the fuel dose calculations. So when it's 30 degrees outside, and the iat sensor on the manifold reads 50, but the λ at idle goes from 1 where we set it, to 1.1, it's obvious the true temp is not 50, it is much closer to ambient. The coolant temp is stable all year, but the ambient/iat temp not. By setting the chargetemp tables between airbox/ambient temp and iat sensor temp, you essentially set how much heat the air absorbs from the hot intake components, in the path it has to take to reach the cylinders, but using realistic inputs/data. Coolant temp is not realistic, the intake manifold is nowhere close to 90 degrees, and the delta of the coolant temp compared to the ambient temp is large.
My English isn't good enough to put down my thoughts exactly the way i want, but i believe you get my point.
Furthermore, chargetemp blending using the coolant temp is forcing you to use false VE numbers in the VE tables, and the lambda is always rich until the engine warms up. A chargetemp blend of iat sensor and ambient temp would also correct this.