Altezza E-Throttle Idle Issues

[Jappachino]

Hi, I have a 1998 Toyota Altezza with a TALTX plug and play ECU installed. My tuner and I are chasing a problem with my throttle body - at idle (at operating temperatures), we have it set so that the target tps is 1.8% and the target RPM's to be 850RPM. Every so often, I would come to a stop light, clutch in and notice my idle sitting higher than it is supposed to, and the tps% is sitting at something higher like 2.2-2.8%. When I blip the throttle, sometimes it goes down to 1.8ish where it is supposed to be and the idle comes down, sometimes it goes to another value in the ballpark of 2.2-2.8% and keep idling high. I also powered on my ECU without firing up the engine, put the e-throttle to "run mode" manually and inspected the behavior when the e-throttle is not running - the tps main and sub voltages would vary by ~0.05V or so each time I blip the throttle or even go on and off it gently,and the tps% also having some variance anywhere from 1.4% to 3.8%, which is worse than when the motor is running... 

So far I've verified that my reference power to the TPS and APS is always stable at 4.95V, I have re-calibrated APS and TPS, verified that the voltages of each sensor (TPS and APS, main and sub) is within spec of what toyota asks for at the rest and full position of the pedal, tried a brand new NTK TPS, tried replacing the whole throttle body from a donor motor. 

The altezza's "e-throttle" is unique (and a pain) where it is something like 60% mechanical for a fail-safe to be able to limp the car to a shop upon e-throttle control failure... I've set the mechanical throttle plate set screw so that the plate is at 1.5% tps when the dbw motor is powered down, and verified that my car can idle happily with that 1.5% by turning the e-throttle in the software OFF and firing up the car. (Weird thing is that this 1.5% TPS reading with the e-throttle OFF is ALWAYS consistent down to the decimal place... ) 

This issue causes some rev hang at times too, and is small yet annoying. My question is, on a dbw setup like this, how accurate should be the tps% when my foot is off the gas? Is it supposed to go back smack on 1.8% which I have it set at, or is it natural to have ~0.5% variance in the "off throttle" or 0% APS position? Could this be a tune issue? 

I will post my setup file in a bit when I am home, and try to get a data log that demonstrates this behavior. 

Any input will be appreciated, thank you in advance. 

[Jappachino]

I actually have a log file here, but the data BEFORE 15:28 should be disregarded as I initially forgot I was running a tune where I was messing with idle target %'s, and stopped to tweak it back to what it SHOULD have been. You will see at the end when the ECU reconnects I turn on the E-throttle with the engine off and go on and off my pedal at various speeds, demonstrating the jumpy tps values.

 

I will leave this here for now as its very late and I will come back tomorrow with the current tune file and a better data log.

https://drive.google.com/drive/folders/1bqd-zBh5m8p4sYmJXOGNMVhA94Ys8BFt?usp=sharing

 

[Adamw]

Need the tune too.  Also in the next log, set the PID to E-throttle 1 so we have that data logged.  It looks like you dont have enough (or any) Integral.

[Jappachino]

Hi, finally got home and I have here a tune that my tuner JUST tweaked a little bit more, and a log of my drive on the highway/city home. Please see link below for the tune and log:

https://drive.google.com/drive/folders/1HXAI5NHYSn0Ltmm7BFFtHj9AdY-uFBYa?usp=sharing

@Adamw, my PID (for now) is set up at 5.5, 0, 25 respectively, I totally missed your message today and did not log the PID values on my drive  I will be sure to do so next time. Also, would the "I" value impact where my idle TPS% lands at?

Thank you,

[Adamw]

Yes, your problem is no integral.  Integral is the factor that slowly builds up to remove the last bit of error that is always left with proportional control only.  Try integral at about 0.1.

[Jappachino]

 

58 minutes ago, Adamw said:

Yes, your problem is no integral.  Integral is the factor that slowly builds up to remove the last bit of error that is always left with proportional control only.  Try integral at about 0.1.

Oh interesting! I will try that tomorrow after work if I have time!

[Jappachino]

Hey @Adamw,

I went for a drive changing the integral gain to 0.1. The tps falls back on a consistent 3.0% (+/- 0.1%) now when I am off throttle which is good, but there is still a lot of rev hang. 

HOWEVER. I was surfing the forums and found this old thread with a comment by @Scott: 

Mentioning a PID setup he had for an Altezza. (PID = 7.00, 0.059, 45.00) I thought okay, let's try what he had there... and the car runs much nicer! I feel like the throttle is more responsive and consistent feeling. This being said, I still have a few issues:

- The idle TPS% being at around 3.0% now, my car idles really high at around 1500RPM. Is this now just an issue of getting my tuner to tweak my idle timing to make my motor run at a lower RPM at that throttle percentage?

- I still THINK there is a bit of rev hang, I actually have not looked at my logged data and will do so after posting this message.

- When I start my car when it's warm, my idle initially hunts like crazy until it slowly comes down to 1500RPM, where it was happily idling at from my first run.

Please find my log files in the google drive folder, alongside my tune with the PID set up as what Scott mentioned.

https://drive.google.com/drive/folders/1SaEchFd7qc-o1Ua_5ew3wY7DH_HR9GaT?usp=sharing

Thanks again!

 

[Jappachino]

I also found this:

 

PID is set at 4.75, 0.005, 35 with max/min clamp at 95% instead of my 90%.

Hey @Davidv, sorry for digging out a really old thread. Do you by any chance have what you had your PID eventually at for your Altezza when you fixed your wiring in the end?

[Vaughan]

29 minutes ago, Jappachino said:

The idle TPS% being at around 3.0% now, my car idles really high at around 1500RPM. Is this now just an issue of getting my tuner to tweak my idle timing to make my motor run at a lower RPM at that throttle percentage?

Now that your E-Throttle is actually getting to it's target you'll need to setup your idle base position table properly. Maybe try looking at old logs to see what the tps was while idling previously to get some idea

[Jappachino]

@Vaughan currently my idle target is set as below:

and I know as a fact that my car will idle happily at around 850 RPM with my e-throttle powered down (so sitting on the idle set screw position, which is 1.5%) should I try tweaking these values closer to 1.5 or even right at 1.5 at operating temp? It's weird that its SUPPOSED to be targeting 2.0% but still sits at 3.0%, would this be something I should further play around with using PID settings?

[Adamw]

1 hour ago, Jappachino said:

It's weird that its SUPPOSED to be targeting 2.0% but still sits at 3.0%, would this be something I should further play around with using PID settings?

PID has nothing to do with the target, PID controls how the throttle motor responds to error.  Target comes directly from your tables. 

The ECU isnt even trying to control idle in your log because the MAP lockout is set to zero.  You can look at idle status to see what it is doing.  When it is in "Hold" mode the target will be Base position+ offsets (dash pot offset in your case adds 1%).  The MAP lockout needs to be somewhere above your normal idle MAP. 

Try the attached tune, I have made some changes back to more normal settings.  Note I have changed lots so it will likely need tweaks and may not work nice right away, but it will be a lot closer starting point than what you have now.  You can see what changes I have made by using the file compare function.   

Also, you have large counts of APS error accumulator, it looks like you havent done an APS calibration, so do that and a TPS calibration once you have loaded this tune.  Do a new log of a cold start, warmup to normal temp and then some throttle blips etc and I will make some further tweaks.

 

 

 

Dylan Altezza basic idle starting point.pclx

[Jappachino]

7 hours ago, Adamw said:

 Also, you have large counts of APS error accumulator, it looks like you havent done an APS calibration, so do that and a TPS calibration once you have loaded this tune.  Do a new log of a cold start, warmup to normal temp and then some throttle blips etc and I will make some further tweaks. 

I did calibrate the APS when I installed this donor throttle body, but I will definitely try again. And thank you, I will try it out later this evening and report back. 

[Jappachino]

Actually, I managed to pop out of work to do this quickly (Perks of working from home haha)

I re-did the APS and TPS calibration, then fired up the car COLD. The idle is actually at target from cold start, and I did not touch my accelerator at all. When it got past 60 degrees ECT, I started doing short blips, some more gradual than others, and noticed it's doing the thing where it is trying to hold the idle at 1400ish RPM's and refuses to come back down. Also, I noticed the idle still hunts ever so slightly, it doesnt bother me too much but I assume this is something that I can tweak one of the P, I or D values for?

Please see attached log of all this.

Thanks!!

Cold start to temp idle - Adam's tune.llgx

[Adamw]

Try this.  Just a warm log will be ok.

Dylan Altezza basic idle V1.2.pclx

[Jappachino]

Hi @Adamw

Wow, with your V1.2 tune the car idles SO much nicer now. I have uploaded a log of my cold-ish start and warm start at the link below.

https://drive.google.com/drive/folders/1ybos0BAxXGOMuZkL_Gmp8nz5g6Y5ZOvT?usp=sharing

Things I observed:

- On the "cold" start, on the initial fire up of the engine the RPM's went up, dropped below target (~800RPM) and then stabilized at the target. Is this normal?

- (Again, the "cold" start log) Around 1:37, the idle started hunting a bit, so I gave the throttle a blip and she came back down to idle happily at the new target RPM.

- (Both logs) I noticed that when I blip the throttle, the revs go up very quickly but coming down it is quite slow and more gradual - is this because we are still working to tweak the idle to behave more consistently first?

[Jappachino]

Bump, is there anything else I should further adjust? 

[Adamw]

23 hours ago, Jappachino said:

- On the "cold" start, on the initial fire up of the engine the RPM's went up, dropped below target (~800RPM) and then stabilized at the target. Is this normal?

Change the idle base position for the 30°C cell to 3.8.

 

23 hours ago, Jappachino said:

- (Again, the "cold" start log) Around 1:37, the idle started hunting a bit, so I gave the throttle a blip and she came back down to idle happily at the new target RPM.

Try dropping the idle actuator integral gain to 0.03 or 0.02.  It may also be fuel related but your lambda is not working in the log to tell.

 

23 hours ago, Jappachino said:

- (Both logs) I noticed that when I blip the throttle, the revs go up very quickly but coming down it is quite slow and more gradual - is this because we are still working to tweak the idle to behave more consistently first?

A couple of things here.  It looks like idle base pos is still not quite right when warm.  Lower the cells for 70°C and above to 1.8%.  Secondly, you need to tune the dashpot effect to achieve the best balance between fast return to idle and minimal undershoot. Try dropping dashpot offset to 0 as a quick test, some engines dont even need it.

[Jappachino]

Apologies for the continuous questions, I'm still new to all this :/

1 hour ago, Adamw said:

It may also be fuel related but your lambda is not working in the log to tell.

I noticed that too, is that supposed to happen? My lambda reads when I'm driving but seems like it stops when I am off the accelerator... Is this a setup issue?

1 hour ago, Adamw said:

It looks like idle base pos is still not quite right when warm.

What should I be seeing to know that the idle position is at the correct %? Is there a range of ignition timing at idle/Lambda I should be seeing to verify this?

1 hour ago, Adamw said:

Secondly, you need to tune the dashpot effect to achieve the best balance between fast return to idle and minimal undershoot.

Ok, I will try some trial and error and see if I can get it dialed in better  Am I understanding this correctly - higher dashpot % = slower throttle plate closing when I suddenly get off the accelerator?

Thank you!

[Adamw]

37 minutes ago, Jappachino said:

I noticed that too, is that supposed to happen? My lambda reads when I'm driving but seems like it stops when I am off the accelerator... Is this a setup issue?

Looking at your log you have lambda error 16 which is usually power supply related.  Either lack of voltage or a particularly noisy supply.  If wires are correct size then try the capacitor as shown in the quick start guide.

 

38 minutes ago, Jappachino said:

What should I be seeing to know that the idle position is at the correct %? Is there a range of ignition timing at idle/Lambda I should be seeing to verify this?

In this pic you can see the ecu needed to move the throttle to 3.9% to achieve the target RPM, but the base position for that temp was only 3.2%.  

 

40 minutes ago, Jappachino said:

Ok, I will try some trial and error and see if I can get it dialed in better  Am I understanding this correctly - higher dashpot % = slower throttle plate closing when I suddenly get off the accelerator?

Yes, the dashpot offset is how much extra the throttle is opened when the RPM is dropping towards idle and the hold and decay adjust how quickly this extra offset is removed.  In the old days cars with carburetors has a small device on the linkage called a dashpot (a bit like a shock absorber) to close the throttle more slowly near idle.   

[Jappachino]

Hey @Adamw!

I hopped in my car and loaded up a tweaked tune where I changed:

- Dashpot offset to 0.1% (undershoots ever so slightly off throttle but honestly I really have to be paying attention to notice)

- Idle target tps% at 30C ICT to 3.8%

- Idle target tps% of everything above 70C ICT to 1.8%

Also, my lambda sensor started working again mysteriously... I will keep an eye on it though. My tweaked tune and log of cold start + drives around the block while tweaking dashpot offset %'s are in the link below.

https://drive.google.com/drive/folders/126-de4Jfwsx3r3LbFxhdkuFnSt_I_Mop?usp=sharing

Though it is sending out a reading, I noticed my lambda spiked at times:

This only happens occasionally when the overrun fuel cut is active. Is this something of concern, or is this normal?

Also, there is one instance at 5:07 where my aps error % spikes to 22, could this be because of a bad aps?

[Adamw]

Obviously the spike to max lean (lambda 10) is normal when a cut is present.  The lambda goes to zero when it is in error condition, so where it drops to zero at the 4:32.5 mark is most likely the same power related issue as above.

[Jappachino]

17 minutes ago, Adamw said:

so where it drops to zero at the 4:32.5 mark is most likely the same power related issue as above

Aahh okay. I will hopefully get around to wiring the capacitor in sometime this weekend and maybe that will fix it?

Does the E-throttle/APS/TPS behavior look to be functioning properly to you? I only took it around my block so it was mostly just stop&go driving, but it FELT a lot better than what it was initially when I started this thread. Only odd thing that I noticed looking at the log was the APS error. 

[Adamw]

Did you do the APS calibration like I suggested earlier?  The file compare shows no changes to TPS or APS calibration, it has exactly the same values in there as your original map which would be a pretty impossible coincidence.  

So you set E-throttle to set up mode, click the little wrench icon on aps calibration, then follow instructions on screen, pushing pedal to floor etc, then set E-throttle back to ON mode, then store?

[Jappachino]

On 8/18/2021 at 8:56 AM, Jappachino said:

I will try it out later this evening and report back. 

I did re-calibrate it again back at this step, should I be doing so on every iteration of the tune? 

[Adamw]

No you dont need to do it everytime, but is doesnt look like it has been done at all.  Did you do it as per my instruction?