Jump to content

E throttle cl idle control


a6_quattro

Recommended Posts

Hi all

Could this be caused by the integral term not being zeroed out when going from open loop to closed loop?

In my case for example, I have added values for AC and Power steering, the integral is quite large when going from open to closed loop causing the throttle percent go from 3.7 to 5.0.

This is just a thought...

Link to comment
Share on other sites

  • 4 weeks later...

It would be very nice if the Link team could investigate this issue and fix the bug in the next sprint.

In my opinion, if you have support in the software that claims to handle closed loop idle control, then you have to make it work.

Its extremly frustrating to have this behaviour on a street car and driving in town for example. When going from open to closed loop the rpm suddenly rises because of the E throttle ISC CL Trim add value of no reason.  

So far I'm very satified with the product except for cl idle control with e throttle.

Have tried many different PID settings for the throttle so that is rouled out of the equation. The respons and behaviour of the throttle is good.

Link to comment
Share on other sites

I have the same issue in my 2jzgte stock E throttle,open loop idle control is ok,but go to closed loop  the E throttle ISC CL Trim add  a little value of no reason,caused  idle speed can't  fall to target value.vehicle speed signal has been connected and set correctly.
This issue has troubled me for a long time.

Link to comment
Share on other sites

I've expierienced the same problem. Besides it seams no possible to find good PID Settings which control idle very precise and makes the throttle enough fast with less overshoout at big changes. 

I use a 82mm throttle on  2,2 liter engine, means idle is controlled within 1-1,5%. Even with idle control turned off,  the PiD controller oscillate slowly arround target especially when in gear, because TP seams to jump once integral has summed up enough to overcome friction. Think the push-pull driver must apply current from both sides to control and hold the throttle with more force arround idle. 

I've used Nichols Ziegler methode, measured step response, phase margin etc. to adjust PID's. I also tried Link VDO PID (which oscillate heavy after a step btw!) and spent a lot of time to find sensible PID empirically. Unfortunatelly nothing seams to work well for idle. 

Bad E-thrrottle idle control was discussed a lot over past years. Think it's really time that link engineers investigate into this Problem. 

 

Link to comment
Share on other sites

26 minutes ago, a6_quattro said:

Should you interpret the silence from Link as "buy another ecu"?

No really not, I'm sure Link will do something. Every ECU manufacturer has some problems.  Except this problem the LINK G4+ platform runs very stable, and reliable. Overall the ECU's are still absolute great! DBW idle control is the only thing which is not ideal atmo. Think it even work ok on small throttles. From a control aspect a big DBW throttle compiend with a small engine is the most difficult thing to control. 

I just asked my system control engineer docent (worked for 20years+  for OEM's). The main problem ist the friction. I will explain it on a example. Target TP = 2.0 %

Actual TP= 1,7% 

The PID controler increase the Output DC%, but the TP stays on 1,7% until the static friction is overcome. Now the dynamic friction is much lower, which means the TP jumps to a much higher value like 2,7% and consequently overshots the target. Now the same process starts to the other side. This is called stick-slip effect. 

I'm aware off two strategy which could be applied. 

- Target signal must be be moved always by a small amount so TP does allways a small oscillation arround Target. Consequently there is only dynamic friction and no stick friction. The disatvantage of this methode is higher mechanical wear. 

- second strategy is to use push and pull driver at the same time. They work agains each other, so force is applied to both sides of the throttle. So the throttle is active hold in position from both sides and not just spring against one side motor force. With this methode the stick friction force is much smaller in proportion to the close and open force applied by the motor. Disadvantage, it's quite complex to engineer the control algorithm. 

 

Link to comment
Share on other sites

4 hours ago, Steve said:

You can also try to lower the frequency the dbw works at to help reduce "sticking".

Syvecs have a nice strategy where it gives the tb a little "kick" to get it moving

I have the factory DBW Throttle body manual from Bosch. 500HZ is the lowest recommendet frequency. Most OEM system work a higher frequencys. 

Lower frequency means also slower control, which we doesn't want. 

Syvecs strategy sound similary to the small oscillation strategy i mentioned. 

Link to comment
Share on other sites

I'm tired of "work arounds" that never work because the toolbox is to small for DBW. And for the issue that I described above you can't possible do a work around.

 I'm fully aware of the different compromises that the different aftermarket ecu:s has, had VEMS before this ecu. So one of the reasons I bought Link Fury was the spec it had. If it claims to have some particilary support it should have it, simple as that.

If Link have had knowledge of this flaw  for a long time then it's really time to fix it.

 

 

Link to comment
Share on other sites

16 hours ago, a6_quattro said:

Should you interpret the silence from Link as "buy another ecu"?

It is actually a long weekend holiday here in NZ.  In the weekends I will typically reply to threads I can in my spare time on a cell phone but obviously cant view logs etc on that.

From what I can make out from your comments above I dont know why the hold timer/hold step is activating, that may be a bug but I will try to take a look at your log next week.

 

15 hours ago, mapper said:

- Target signal must be be moved always by a small amount so TP does allways a small oscillation arround Target. Consequently there is only dynamic friction and no stick friction. The disatvantage of this methode is higher mechanical wear. 

If you set your E-throttle deadband to zero thats exactly what would happen.

 

15 hours ago, mapper said:

- second strategy is to use push and pull driver at the same time. They work agains each other, so force is applied to both sides of the throttle. So the throttle is active hold in position from both sides and not just spring against one side motor force. With this methode the stick friction force is much smaller in proportion to the close and open force applied by the motor. Disadvantage, it's quite complex to engineer the control algorithm. 

6

Im not sure where you heard this but it seems like a fantasy to me?  It is just a single pole DC motor - think of it as a coil of wire.  You cant "apply current to both sides" - no electricity would flow and nothing would happen.  You can only apply current in one direction to drive it open, or reverse the current to drive it closed.  There is no way to "clamp" or "hold" a traditional DC motor using electronics.  The normal strategy for a E-throttle is they have a spring which normally holds the blade open to about 5-10%, so the ECU has to apply a very small DC to drive the motor closed to idle position, the only thing that is holding the blade still is the balance of motor power against the spring pressure.

 

16 hours ago, mapper said:

Actual TP= 1,7% 

The PID controler increase the Output DC%, but the TP stays on 1,7% until the static friction is overcome. Now the dynamic friction is much lower, which means the TP jumps to a much higher value like 2,7% and consequently overshots the target. Now the same process starts to the other side. This is called stick-slip effect. 

This should not happen at idle due to the spring that I just mentioned, as soon as you reduce motor DC the spring will overcome the motor torque so it will move to a new point of equilibrium.  Of course excessive friction could be a problem but generally in a throttle body the spring is several orders of magnitude stronger than the friction effect.  

 

As for general E-throttle idle control we dont get many reported problems, considering we sell more than 100 plug-in ECU's for factory E-throttle cars every month (Subaru V10, 350Z, Mini, Audi TT), there is a huge volume out there seemingly working ok...  Maybe it becomes more of a problem if the butterfly is not well sized for the engine? 

I suspect A6 Quattro's problem described above with the hold step may be a genuine bug but on the whole we usually see quite acceptable control.  Certainly, my E-throttle daily driver behaves well, as well as the rest of the test cars at Link HQ.  As far as I have seen there are no improvements scheduled for E-throttle idle control in the next firmware release but if there are genuine bugs causing control problems these will certainly be looked at. 

Link to comment
Share on other sites

Hi Adam

You're right I missed to think about it's a singel spool DC Motor. This strategy needs two coils or a stepper motor. 

But the small moving strategy is not the same as just zero deadthband. Zero deathband makes the control always work, but doesn't change the target signal. The strategy i was talking about does apply a small oscillation to the target, so to the controler input. 

I expierenced the same problem with hold step as A5_quattro. I will check if i find a log.  

Not right sized dbw to engine size is relativ. We run 800hp+ on a 2,2l engine. This is calling for a 76mm+ throttle body. I agree that this makes it much more difficult to control, but somehow we need to find a strategy in aftermarket to control it. 

 

Link to comment
Share on other sites

Adam

Now when I got your attention, I can record a log with other settings on PID and idle control. Don't hesitate to ask as I'm driving the car every day with exactly the same problem.

As I mention for you by mail, my idle is almost like OEM so that's NOT the problem. Throttle control is good following target so that can't be a problem with that settings in my eyes either.

The problem is coming from open and going to closed loop THEN there is a value added randomly.

Sometimes it's a small negative and when coming from open loop at next stop there is suddenly a positive value added.

Tried with all possible settings on idle control with 2D table and 3D, zeroed out all kind of load compensations an so on.

When such things happens randomly, usually there is something strange going on in the code or the strategi being coded up.

To sum up:

1. Not all possible PID settings being tried out but the one being tried provides a good throttle control following target as it should

2. To investigate why this happens I tried different settings and also zeroed out all possible settings, I think, that could interfer with the idle control

3. Idle control on both closed and open loop is good

4. When going from open to closed loop the E throttle ISC CL Trim randomly adds a arbitrary value

Focus around the E throttle ISC CL Trim and what's happens there and not the PID settings is my advice as an novice mapper.

Regards

Hans

Link to comment
Share on other sites

Hi Hans,

I have had some time to test your file on the bench.  Strangely when I try to reproduce the same conditions shown in your log I do not get any ISC CL trim added.  I also donot see any reason for it in your log/settings.  Does this occur regularly for you?  Is it always a similar value or is it random as you mention above?  I will get engineering to take a look once I have a bit more to go on.

Link to comment
Share on other sites

Hi Adam

Yes, it happens regulary and it looks like it's a random value, sometimes it's negative and sometimes it's a positive value.

I've noticed that the ISC Offset (%) holds a value that changes over time despite that idle control is in open loop in the log file.

Just a thought, could it be this that introduces a value to Trim? Is this some kind of "learnt value" thats taken from last time it was in closed loop?

Maby thats correct, couldn't find anything in help file and its extremely difficult to guess when I don't know anything about the strategi around this.

Just trying to help role out the obvious.

Hans

Link to comment
Share on other sites

Ok, I have been able to reproduce this now although not consistently, I too get something fairly random, sometimes positive, sometimes negative.  I have passed my test data on to the engineering team to take a look.

Link to comment
Share on other sites

  • 2 weeks later...
15 hours ago, a6_quattro said:

Adam

Whats the status on this?

 

No updates yet sorry, I have put as much pressure on it internally as I can.  It is now down to where it fits into the priorities in the Engineering dept.

 

5 hours ago, 434josh said:

I am very interested too as I am about to setup my extreme with dbw and closed loop idle control.

For some reason this doesnt effect everyone, you might as well give it a try.  If you have a problem you can always change to openloop, with ignition idle control added it usually gives quite acceptable results too. 

Link to comment
Share on other sites

I have som kind of same problem with my thunder on a v12 engine. Cl throttle is adding and pulling away procent of the throttle. If in closed loop and pulled away 5 %on idle i only got 95% opening on full throttle. - 3% Will be 97%on full throttle and so on. I had this issue over a year without fix, lambda correction bank by bank dosen't work either. I had mail contact that ran out in sand. Sad because it's a good ecu.. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...