Idle Control on Mercedes M104.9xx

[pric]

Hello everybody!

My Installation is kind of finished since last year. (steady improvement and fiddeling around)
However i never got the idle control working as i would like it to.

The M104 with MAF (from 1991 on) uses DBW if equipped with ASR.

If not equipped with ASR (like mine is) it uses a throttle plate actuator aswell, which has a limitation to around 25% TPS.
The actuator part of it is independent of the DBC. (free run)

The factory ECU uses an H Bridge to control the throttle plate.
At least thats my guess because the plate is never fully shut(it has a dampened stop with which you can force it shut manually).
And i can't get it to idle below 1050rpm with -2° of ignition advance
Factory idle is between 600 and 800 rpm at 80...90°C CLT (with 6...10° of advance), and up to 1500 rpm on warm up.

It has a TPS pot, and a seperate inverted TPS pot just for the Idle mechanism.
It has an idle switch as well. (idle control is enabled when switch is pressed)

Right now i'm just using the TPS and an AUX port for the flap motor.
The problem is with the AUX Port i can't control it fine enough, because of the closing spring of the flap.
I need like 25%+ DC to open the flap, but then it's opened 20%TPS right away which is way too much.
You can then decrease the DC and it closes relative to the DC.
Next Problem, as the Idle is activated via TPS percentage, it shuts down the idle control when the actuator opens the flap, which then shuts down the actuator, and gives you a ramping up and down of the rpm.
You can't get around this because the fuel cut etc. is also dependet on the tps value.

I thought about using a relay with the idle switch, that cuts the tps signal wire, when the switch is pressed...
(can't use it directly cause it's a NO contact)
But then i can't still close the throttle more, or control the DC fine enough...

Any thoughts on this one?

Greetings and thanks for the advice beforehand...
Paul

[RHD]

are you using the external DBW controller?

the standard EMU is not designed to run DBW directly, and i think that will do everything you require if i understand the setup and problem correctly?
lots of DBW throttles dont fully close in that same way, only the BLACK version EMU can to it internally

[pric]

It is Drive By Cable, so a standard throttle body in that way.

Like this non ASR throttle body:
https://www.youtube.com/watch?v=oluUNq4Yznc
(except mine doesn't have cruise control either, so even a little simpler and less possible opening thru the actuator)
(looks like this: https://www.youtube.com/watch?v=Pyncnd0E2YY)

Instead of an idle air control valve the throttle plate is actuated by a gear motor.
However the gear motor can only open the throttle plate about 20°

To control the plate nicely it has a feedback pot for the throttle opening angle the gear motor makes.
And an idle switch to engage the idle control... (and probably fuel cut on the original ecu)

What i was thinking is to use the stepper out, as this is an h bridge and should be able to invert the polarity of the voltage and therefore alowing me a lower idle. And cut the tps signal via a relay triggered by the idle switch.
(bad thing about this is, i would only need the stepper a, and stepper b is then blocked by the software)
But how many steps does a "permanent magnet dc motor" have? LOL,
They are "infinite" because of the commutator doing its thing...

thanks for the reply!

[RHD]

ok now thats how i initially thought it worked but theh had second thoughts. You can not do it with the stepper OPs they are not an H bridge and they have very low current capabilities. Use an AUX op and install a flyback diod!

I see the stop inside the throttle looks like it has adjustment. I think the only way you can do it is to adjust that to either the highest RPM you could need or the lowest RPM you could need depending on the spring tension and available travel etc. Then treat it like a standard PWM idle valve PID settings to always pull the correction against the spring in the same direction. The frequency will need to be very high and i dont know if theres any way to use the inverted position sensor. It could be very sensitive so the numbers in the PID settings could be very small.

I would start by adjusting it up to say 1200rpm then using the PWM idle control to try and pull it down.

[pric]

It is already controlled via AUX1 configured as pwm valve with a flyback diode.

The open stop has a hard limit without any spring it's limited by the halfgear lever or whatever it's called.
(I could jam something in there to limit the opening angle)
The closed stop is adjustable.

My main problem is that the plate has to be able to move in both directions, to get it to idle lower...
So the output has to be invertable.

I'll probably adjust the closed stop, so that the throttle plate is fully shut.
And use the idle switch to interrupt the tps signal and keep it as "pwm valve"
Because when im pulling it down, i can't pull it up any more.
With cold engine it's struggling to keep running, because it needs more air then the normal opening angle can provide.
(the first 20 seconds or so)
Even though it would probably work better once it has warmed up, if i do it how you suggested.

If this doesn't work out i will get a pwm idle control valve.

Thanks for your help!

[gregpe]

my impression is that is symilair to JZ when throttle is mechanically onen only little just in case if electric throttle fail.
you would need to use DBW function in your case and have a throttle controlled with H bridge.

[RHD]

It is already controlled via AUX1 configured as pwm valve with a flyback diode.

The open stop has a hard limit without any spring it's limited by the halfgear lever or whatever it's called.
(I could jam something in there to limit the opening angle)
The closed stop is adjustable.

My main problem is that the plate has to be able to move in both directions, to get it to idle lower...
So the output has to be invertable.

I'll probably adjust the closed stop, so that the throttle plate is fully shut.
And use the idle switch to interrupt the tps signal and keep it as "pwm valve"
Because when im pulling it down, i can't pull it up any more.
With cold engine it's struggling to keep running, because it needs more air then the normal opening angle can provide.
(the first 20 seconds or so)
Even though it would probably work better once it has warmed up, if i do it how you suggested.

If this doesn't work out i will get a pwm idle control valve.

Thanks for your help!

yes thats basically what i was trying to explain but i would adjust the stop until its open enough to idle when its cold then try to pull it shut more once it warms up but it depends on the travel of the spring loaded stop etc... but in either case so all the control is in one direction only.

[pric]

@gregpe:

On the 2JZ you have basically two throttle plates, one is dbw and the other one is dbc.
In my case i only have one throttle plate, which can be actuated a small amount (maximum 20° or so) to control idle.
It's not a stepper motor it's a normal dc motor, so i can't use the stepper out.

Thanks, though!

@rhd:
I thought about it and doing as you said is smarter. >> Only closing it is probably better then only opening it.
Then i don't have to mess with the TPS input at all.
I will modify the open stop then so it is even more open than it already is, to allow for cold start and cold idle.

The question is, to which angle should i calibrate the zero point then.
Fully shut 0% TPS, or neutral stage 0% TPS? (does it matter?)
And should i tick "reverse" in the idle control parameters?
What does that mean anyway?
100% DC in the Map is 0% DC at the actuator, 80% DC in the Map is 20% DC and so on?

Thanks again!

Greetings
Paul.

[gregpe]

etsc throttle in 2jz is a flyby wire including clutch and fail-safe, if dbw fail with cable you able to open throttle about 20%... strange system...
regarding your one i think you should use DBW module not stepper outputs and you will be able to control both direction of throttle plate... but perhaps mercedes has different strategy..

[pric]

So i tried both ways.
Closing and Opening it via AUX1.
(I made the idle control active all the time, by typing in 100% in both tps below and above, so i don't have the hassle with tps shutting idle control of)
Neither of them work, both for the same reason.
Once warm closing via aux works fine.
But i cant open it wide enough (without modifying/destroying the stock throttle body) to allow for enough air on cold engine. (these TBs are too damn expensive, to realy modify them...300€+ used!)

Opening via AUX without closed stop, also works fine, until you go hard on the throttle.

Because of the spring there is so much "backlash" for the electrical signal.
When working against the spring i need 26% DC to crack the throttle plate open.
But in the other direction one percent less duty cycle can be enough to stall the engine/decrease the rpm tremendously.
So there is never the same DC on the motor for the same rpm because of the "backlash".
So the closed loop did all the work, but it's not fast enough.
I tried 50% feedback, and upped the kP and put some differential component in it.
Right at the edge where it began to swing a little bit, it worked pretty well already with ignition controll enabled.
I put all my knowledge from control engineering courses in it i took a long time ago...
However it was still a 50:50 chance if the IAC can catch it when going off the throttle.
There are too many factors playing a role for the idle, as to make the closed loop work reliably without position feedback.

Thats the reason why nearly every control valve in industry has a positioner,
and why this throttle actuator has a position feedback signal...

I gave up on it, i orderd a three and a two wire idle pwm valve...

We'll see