oversteer Posted October 13, 2019 Report Posted October 13, 2019 Had the ethrottle brick and log code 75 yesterday....unfortunately i was not logging when that happened... I switched the throttle to always on and took some logs today, the sub and mains seem to track properly, but aps and tps do not really, tried changing proportional and derived from 7 & 35 to 8.5 & 40 , not sure it helped ! Car uses a Subaru bp5 Lib pedal and ls3 90mm TB. Is e Throttle behavior acceptable ? have attached tune and 2 logs, on log 7 & 35, one 8.5 & 40. eth8.5and40.llg eth7and35.llg EG33Lib2.pclr Quote
oversteer Posted October 13, 2019 Author Report Posted October 13, 2019 *update* Just started car up again with the hope of a little drive tuning, but ethrottle is not happy....log is a train wreck ! ethrot err.llg Quote
Adamw Posted October 13, 2019 Report Posted October 13, 2019 What PID was in use when it became unstable in the last log? Normally 7 & 35 works ok on the LS throttles. You may have to reduce P further. Quote
oversteer Posted October 13, 2019 Author Report Posted October 13, 2019 8.5 0.051 40 How close should aps and tps track ? I can see idle control trying to do it thing, i would hope the error ratio ignores this ? Quote
Adamw Posted October 14, 2019 Report Posted October 14, 2019 21 hours ago, oversteer said: How close should aps and tps track ? AP sub and main need to match. TP sub and main need to match. TP main and TP target needs to match. There is no requirement for AP to track TP as in many scenarios there is no relationship between these two at all. Your error is because TP main does not match TP target. Use p= 7, I = 0.05, D = 35. You may need to drop P to 6 or increase D a little if it still becomes unstable with those settings. Quote
oversteer Posted October 15, 2019 Author Report Posted October 15, 2019 Yeah that's pretty much what i got yesterday playing around with it, 6.5, 0.051 37, however that PID of the throttle is not what actually caused the code 75 this latest time ! if you look at the log when the error counter started you will see the TPS bounce/oscillate even with aps at zero(the car did this when i set throttle to always on when testing that could be heard) and it seems it was a hysteresis thing cause by setting the throttle map with 9 at 0(an attempt to implement some base idle position there rather is the idle control system)....seems it doesn't like this, having the first number so close to second(10) ! Quote
Adamw Posted October 15, 2019 Report Posted October 15, 2019 21 minutes ago, oversteer said: Yeah that's pretty much what i got yesterday playing around with it, 6.5, 0.051 37, however that PID of the throttle is not what actually caused the code 75 this latest time ! if you look at the log when the error counter started you will see the TPS bounce/oscillate even with aps at zero(the car did this when i set throttle to always on when testing that could be heard) and it seems it was a hysteresis thing cause by setting the throttle map with 9 at 0(an attempt to implement some base idle position there rather is the idle control system)....seems it doesn't like this, having the first number so close to second(10) ! No, it is PID related. Even with APS at zero the throttle is still controlling the position in closed loop. In PID tuning, when you have an oscillation that increases in amplitude with time it is called an "unstable system" - it will never settle down or fix itself unless you interrupt it (this is why the ECU shuts off power in this situation). With any PID tuning there is always a compromise between stability and response time, DBW control is particularly tricky because for acceptable response you need to be working very close to the limit of stability - you will see some oscillations but their amplitude should decrease quickly with time. Quote
oversteer Posted October 15, 2019 Author Report Posted October 15, 2019 Hmm I don't know, we are not talking a small oscillation finding it position, the throttle was bouncing rapidly, 9 times a sec in log and it stopped when i changed the throttle map. PID wasn't much different than it is now also. Quote
Brad Burnett Posted October 15, 2019 Report Posted October 15, 2019 need more D. Ive had hell with some LS based throttle bodies. The increase in derivative gain will help dampen the oscillations. Quote
Adamw Posted October 16, 2019 Report Posted October 16, 2019 21 hours ago, oversteer said: Hmm I don't know, we are not talking a small oscillation finding it position, the throttle was bouncing rapidly, 9 times a sec in log and it stopped when i changed the throttle map. PID wasn't much different than it is now also. Notice how the oscillation starts small and increases until it is interupted by a change in set point - this is classic PID instability. You will also notice the Motor DC is exactly 180deg out of phase with the throttle position (i.e the motor DC oscillation peaks exactly when TP oscillation is in a peak trough - the yellow cursor shows this). This usually indicates the oscillation is due to too much proportional. I still wouldnt rule out needing more D as Brad says however. D has an effect kind of like "putting on the brakes" as you approach the target so it can have a similar effect to reducing P. Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.