Throttle "hang"

Howdy. Recently updated an older Daytona Prototype to an M150 from the older Bosch 4.3. This is a fairly high strung 5L LS based motor. I was able to get it up and running pretty easily, and after a little bit of idle crank and airflow tuning it starts and idles all day. Note this is far from a tuned file, Ive only done basic setup and a few tweaks so it idles happily. Note I am using the default throttle servo opening vs airflow which is way off for an LS3 throttle body that I am using, but I have had no issues on another car that I have this setup on.

However when I rev it a little (either with the pedal or by pushing on the shift lever) the RPM hangs, throttle shows more open (aligning with the hang) and it doesnt want to enter back into idle mode. Its probably good enough to get it to a dyno where I will be around more experienced tuners, but wanted to see if I could resolve it before getting it on the rollers. Also more fun and a better learning experience when you have to fiddle! Thanks.

Try this.

Your Idle Actuator Throttle Aim Minimum is quite high at 45%, Your Idle Mass Feed Forward Main has values around 92% at idle speeds, which means that the Idle Control will be aiming to have the throttle open to around 40% when idling. This seems to be excessively large to me, compared to other M1's that I have worked on.

First, the Throttle Pedal might not be calibrated. In your logged data, the Throttle Pedal has the value of -2.9 for most of the file. There is a Throttle Pedal Sensor Diagnostic "Closed Warning".

Now I think the reason the throttle moves at all the the Throttle Aim State becomes "Gear Shift". The Throttle Aim State is never "Pedal" which would be the pedal requesting the throttle be changed.

Your Throttle pedal settings look really odd. Only a little over 1V of range, and an offset around 2.1 - 2.8v for main/tracking. Something is not right here.

In the idle control, the feedward table has 95% of a 45% open throttle. You might want to get the FF table a little more inline with the actual throttle position to reach that target RPMs.

That should give you something to fiddle with.

First of all, thanks to both of you.

Black- This is my second LS car, and I think its related to the design of the throttle body. The first 30%+ of servo movement has a tiny change in airflow. I think its to help the car idle well.

David-

I fixed the -2.9, added throttle springs and forgot to re-cal. The range is small due to the way the linkage is setup. it was a non DBW car with a 15+year old footbox. I didn't think this would be an issue other than it being "weird" so to speak. (although any play, etc would cause issues). I wanted to keep the linkage short, simple to reduce risk of failure and/or parts adding clearance/play. I could obviously add some complexity to the linkage with a ratio change. Obviously I use "digital resolution" in the AtoD, but the drivers foot probably bounces around more than that anyway ;)

feedforward, i am somewhat close. takes between 36-38% servo position to idle. I will bring that closer and report back! I think I recall you tuning some LS cars and are familiar with the servo vs air flow at low throttle positions.

As far as the throttle moving, in the log I used downshift request to do the blip but it does the exact same thing if I manually blip the throttle. I tried it both ways to possible better understand the issue.

Jordan,

Let us know how you get on.

When you say "The first 30%+ of servo movement has a tiny change in airflow."

Check your throttle area table. If that table doesn't reflect the airflow of the throttle body, it will skew this behavior dramatically.

What size is the TB?

Stock LS3 throttle body.

I did not change the throttle area table since I don't have flow bench data and it would be a guess. Note that I followed the same process when I tuned the LS in my Porsche and it worked great (street car, so drivability is important) other than the tune looking weird because of the large servo position at idle. at 30% you are at the edge of this reduced area, so you start to get into the normal flow region pretty fast. Especially on this high strung engine that idles at ~1400rpm (Bosch had it setup with an 1800rpm idle!)

I do think I need to do another loop of basic idle tuning and then adjust the feedforward tables correctly. That is my plan Friday when I return to town. Based on the explanation above and my further research I think that is my issue. I got it close so it idled great, but because of the error b/w feed forward and reality the return to idle is delayed.

Ill be sure to post back. This engine is a challenge and I am looking forward to the learnings.

Sounds good.

Separate from rev hang...the somewhat similar engines I've tuned also had very high idle speeds, but seemed perfectly happy idled WAY down. Monitoring coolant pressure, oil pressure, temps, everything looked great when idled down, they weren't getting hot, crankcase pressure looked fine, they burned clean, so I was left wondering if the level of control was simply better now, or an engineer had done it for some reason specific to the race engine that I wasn't aware of. Depending on the level of sensor data and what was known of each engine, then I'd make a call on how high to leave it.

That said, part of me would wonder if it was something far less technical, like they had a driver that stalled it a few times so they kludged the tune for them.

I suspect this can continue to be idled down as I get the tune more in line. Not sure how to explain it, but it sounds comfortable at 1200 all day. You don't get too much driveline noise from the engine torque being not consistent between power strokes (cam lope etc).

Ill keep an eye on the other factors, but I suspect that everything will be fine lower. I don't like going much below 1200 on an oil restricted pushrod engine like this since its likely the top end/lifters/etc do see a wonderful amount of oil. But I have no data to back it.

The Bosch is also Alpha N with a cable throttle body with no idle air control valve (at least as it was delivered to me). It was not the easiest car to drive, so I imagine the high idle helped with making sure it would start at idle across a wide range of temp/pressures and easier on the driver (as you noted)

Yup just relaying that while you can easily idle it down now, there may have been valid reasons for idle speed being up a bit, outside of the shortcomings of their controls. Sounds good!

Thanks David for the files, they were great references. Now I understand how the ecu handles the feed forward table everything makes sense. I made the servo area table adjustment (thanks David) and then tuned the feed forward table. Car seems to be close at idle, AF ok (closed loop is off given the significant cam overlap). The car also doesn't hang anymore, since the feedforward table is now calculating a closer servo position. However it now dies after a rev (no big deal, I haven't even tried to work on it) but an odder issue is that it will idle for a bit and then die, as if the throttle servo is HYPER sensitive. In the attached log you can see the servo go from 5.2% to 5.1, and the engine instantly starts to die. The control algo bumps it back to 5.3, and throws timing at it but it can't recover. In the car it is not a stumble and die, but rather a pretty fast off. With the servo being so insensitive around idle (by mechanical design) I am surprised. Thoughts?

Thanks! Ill be on the dyno next weekend (hopefully) with some savvy motec goes in the building but really would like to have this basic stuff sorted. And the learnings are fantastic.

Achieving net zero torque idle on an 8 cylinder engine with very little inertia and frictional losses involves creating very little torque per cylinder. You get to a point where anything causing a little less torque output isn't sufficient to sustain engine operation and it shuts off like a light switch. I've run into this on similar engines. Raising idle so you have a little more margin is one bandaid sometimes used. Dialing down idle throttle PID control and letting timing do more work helps.

The 0.1 of throttle plate motion isn't doing a ton unless that's literally closed vs. open.

Where's it going for tuning?

Ya, I was surprised. I can't imagine that 0.1% movement of the throttle servo would cause anything dramatic. I started on another part of the car today but ill get back to fiddling with the tune Thursday. Thanks for the advice, more stuff for me to play with. Also want to check the intake for vac leaks (I didnt build the motor, car was technically running when it was dropped off)

I intend to tune it (mostly) myself using GSPeed's dyno, so I will have his tuner on standby for help and to do a final look over my work. Or if I am pulling my hair out, hand him the laptop and ill pull up a chair and a whiskey ;)

I am not charging the customer for the tuning since clearly I am not a professional at it, so its not a huge deal if we have to pay Louis' guy to help. Though my goal is to come out with some serious knowledge and so far so good. Ive tuned 5-10 LS cars back 10+ years ago, a few LT1 stock ECUs 20 years ago, and a FAST EFI back in the day. Then the M150 in my Porsche with Collin's help (he did the full throttle stuff, I did everything else).

All sounds good man. Best of luck and let us know how it goes. Hopefully I'll see it at COTA soon.