When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Did you have your A/C on during the cranking? It sorta looks like the A/C compressor was on and it turned off when the dip happened, idle duty drops 10.2% instantly which matches the A/C compressor idle up duty. The closed loop idle handoff is fine and doesn't seem to be causing it. What also happens is your VVT Duty shoots way up (hits the 1800 column in the table) which allows in lots more air and doesn't help stability as well. That's why the VVT table I have has the 1500 column next to the 1600 column, stops it from advancing as soon as the RPM goes above 1k RPM, starts when it advances at 1500 so I usually don't hit that during idle or during startups.
You can check if the VVT is causing it by changing the coolant on value to something like 220C when testing hot starts so its disabled entirely.
Thanks! I will test that tomorrow after work. I played around with a few settings tonight and I believe it may be VVT causing it, but I will find out tomorrow and report back. I didn't have the A/C on, but I think my cooling fan did kick on, so it may have been that if that could also make the duty cycle drop 10.2%. I will have to see how high my idle goes during hot start up, because it does jump up higher than I'm used to, but not sure exactly how high; will have to pay more attention to that. Thank you, I will let you know how it goes!
If its the cooling fan kicking on, changing the fan on coolant temp to 203 and the fan off temp to 197 or so is safe and should stop the fans from turning on immediately after a hot start, as it will cool a few degrees usually. Defroster also turns on A/C compressor
If its the cooling fan kicking on, changing the fan on coolant temp to 203 and the fan off temp to 197 or so is safe and should stop the fans from turning on immediately after a hot start, as it will cool a few degrees usually. Defroster also turns on A/C compressor
Great, thank you! I want to see if first it may be the VVT settings, since I suspect it might be. I'll let you know tomorrow evening what I find!
If its the cooling fan kicking on, changing the fan on coolant temp to 203 and the fan off temp to 197 or so is safe and should stop the fans from turning on immediately after a hot start, as it will cool a few degrees usually. Defroster also turns on A/C compressor
So, looks like it is the VVT settings. I got the car up past the 140 for the CLT temps, and then shut it off and turned it back on, and it made those oscillations. I then changed the activation temp to 220 degrees, started the car, and it started up just fine. The car did die again at 140 degrees for some reason...it hadn't done that before but this time it did again. I'm wondering if it could have something to do with the VVT table? I will have to try scaling it the way you mentioned and see what happens.
You could try the VVT table I have on my tune and see if it works for you, has a bit more advance across the board and usually dyno's pretty good from what I can tell online. 29 advance at 1600 is nice too for power.
Hard to tell why it stalled without a log, could just be VVT related as well, no idea.
If you want to check the effectiveness of your PID settings for VVT, you can log both the VVT Target Angle and the VVT Actual Angle and see how fast it reaches the target and if it oscillates around the target.
Last edited by HowPrayGame; 10-19-2020 at 08:01 PM.
You could try the VVT table I have on my tune and see if it works for you, has a bit more advance across the board and usually dyno's pretty good from what I can tell online. 29 advance at 1600 is nice too for power.
Hard to tell why it stalled without a log, could just be VVT related as well, no idea.
If you want to check the effectiveness of your PID settings for VVT, you can log both the VVT Target Angle and the VVT Actual Angle and see how fast it reaches the target and if it oscillates around the target.
Thanks! I'll give that a try. I will try logging my current PID settings first, and see what I get there. I'll upload the log for you to see as well, since you have much more experience than me when it comes for this. I appreciate you working through this with me, definitely a learning curve to it.
So with everything unchanged, here are my two logs. VVTLog1 is a hot start around 176 degrees. It looks like VVT starts to activate during startup. VVTLog2 is a start from around 134 degrees to 140 degrees where the car then shuts off. It looks like VVT starts activating around when the RPMs drop rapidly and then the car stalls out. What do you think?
The log with the stall only seems to have 8 data fields logged so I can't really see what's happening. Total guess but sometimes the Neutral/Hold Duty value takes hold and commands the VVT to do stuff at idle even when its off, try changing it from the 39.6 or something your basemap has it as and try 19.6 instead. If the log has the duty jump to around that value then the car stalls that's probably it
The log with the stall only seems to have 8 data fields logged so I can't really see what's happening. Total guess but sometimes the Neutral/Hold Duty value takes hold and commands the VVT to do stuff at idle even when its off, try changing it from the 39.6 or something your basemap has it as and try 19.6 instead. If the log has the duty jump to around that value then the car stalls that's probably it
Thank you! I will give that a go tomorrow. Sorry about the log, I was trying to condense it down to what I thought would be needed, but if it happens again I can take a better log. Thanks for the help, I'll let you know how it goes!
The log with the stall only seems to have 8 data fields logged so I can't really see what's happening. Total guess but sometimes the Neutral/Hold Duty value takes hold and commands the VVT to do stuff at idle even when its off, try changing it from the 39.6 or something your basemap has it as and try 19.6 instead. If the log has the duty jump to around that value then the car stalls that's probably it
Just to give you an update, I gave that a try, and it had the same behavior. So, I decided to play around with a few of the VVT settings this evening, and I think I may be getting it somewhere good! So I was able to get it to stop oscillating, but then it had a SHARP drop in idle RPM between 1800 and 1000. Car did stay running though. When I took a log, I saw that my RPMs creeped a little over 1800, so VVT activated, hence the drop. I set the minimum CLT to 220 as a test, and the sharp drop was gone! Looks like I now need to tweak my VVT table. I'm thinking of using yours and seeing how it goes; do you notice any behaviors with your VVT table that you don't like, or is it a pretty solid table you would say? I have uploaded my current tune along with a log of the behavior I mentioned just so you can take a look if you'd like.
Uploaded tune still looks like it has Inj. Timing being modified by VVT Angle? Turn that off and it will fix the stall/weird behavior whenever VVT kicks on at your coolant temp of choice, its the Adjust Inj. Timing Based On setting like I have it in the last photo I uploaded.
You may want to change this setting under Closed Loop Idle settings from 1 second to 2 seconds on the PID delay, the car isn't following the crank to run taper it should with the IAC Valve, just going directly into closed loop so it drops pretty significantly. My car did that for a while before I figured out that low of a PID delay will negate the crank to run taper entirely resulting in a big dip right after startup.
Hitting VVT Advance during startup (in my experience) will do what you describe. I don't have anything too weird happen with my current table even though it sorta hits the 1600RPM column during startup. It looks like you are getting VVT advance around 1k RPM which may be due to your maximum/minimum VVT angles being set not perfectly (the ~275 and ~318 values), some calibration may be required there as well.
Something else to keep track off for a sharp drop off of RPM right after startup, my car will hit over-run fuel cut when I start the car hot and it results in immediately hitting 2k RPM > Dipping to 1400 RPM then tapering off to the target RPM as it normally does. You don't seem to have over-run enabled yet so that isn't an issue, if you want to get better fuel economy having fuel cut off when you aren't on the throttle is pretty nice for fuel economy.
Uploaded tune still looks like it has Inj. Timing being modified by VVT Angle? Turn that off and it will fix the stall/weird behavior whenever VVT kicks on at your coolant temp of choice, its the Adjust Inj. Timing Based On setting like I have it in the last photo I uploaded.
You may want to change this setting under Closed Loop Idle settings from 1 second to 2 seconds on the PID delay, the car isn't following the crank to run taper it should with the IAC Valve, just going directly into closed loop so it drops pretty significantly. My car did that for a while before I figured out that low of a PID delay will negate the crank to run taper entirely resulting in a big dip right after startup.
Hitting VVT Advance during startup (in my experience) will do what you describe. I don't have anything too weird happen with my current table even though it sorta hits the 1600RPM column during startup. It looks like you are getting VVT advance around 1k RPM which may be due to your maximum/minimum VVT angles being set not perfectly (the ~275 and ~318 values), some calibration may be required there as well.
Something else to keep track off for a sharp drop off of RPM right after startup, my car will hit over-run fuel cut when I start the car hot and it results in immediately hitting 2k RPM > Dipping to 1400 RPM then tapering off to the target RPM as it normally does. You don't seem to have over-run enabled yet so that isn't an issue, if you want to get better fuel economy having fuel cut off when you aren't on the throttle is pretty nice for fuel economy.
Oh shoot I didn't realize that I still had that on, thanks for pointing that out! I'll make the change on the PID delay as well and see how it goes as well as look up how to calibrate VVT; I've never done that before, so I'll need to research how that is done, or how to get those values. Thanks for the tip on over-run! I haven't ever used that before, but I will look into it to see if its something I'd want to try out. I guess it wouldn't hurt to just mess with it. I'll let you know tomorrow after work how it goes with these changes.
Wanted to give you another update. So I changed the Inj timing to "Off" and then increased PID delay to 2 seconds. That didn't fix it, so I did as you said and went through a VVT calibration. The way I did it was setting the duty % under "test" to 100, then the min and max angle to 0. Revved the car to around 3k to 4k RPM, took a datalog, and recorded down the highest value as my max. I then did the same but put the duty % to 0 under "test". Recorded this as my minimum. After turning off the test and putting my new min and max values, that seemed to have fixed the issue! However...I'm now experiencing a very slight fast dip in idle RPM after revving the engine. I revved the engine in place to around 4k RPM, and then noted the behavior. After revving to 4k RPMs, the idle then "dips" around 1k to 760 RPM, and shoots back up to around 930-940 RPM. I took a datalog, and you can see this occur starting from 5.6 seconds to 6.9 seconds. I've uploaded my datalog along with my newly revised tune.
I want to say again, thank you for helping me out and giving me advice; I really appreciate it! I think we are getting close to a decent idle!
Looks like a really good return to idle, that little dip should be fixed by raising the closed loop idle initial value table a tad so it hits a bit higher than 850 RPM (900ish is good) and then can settle down to your target idle gently.
Just for some insight, I am looking at if your car is currently in closed loop idle during the dip (Status 2 = 128) and it is not, and the IAC Duty % seems to match your closed loop idle initial value table. Raising the area of the table it is referencing (850 RPM X between 170/190CLT) should hold the idle slightly above 850 until closed loop idle kicks in and can bring it down to target. That should make it close to OEM idle, maybe even better. Can't recall how my OEM ECU idled, its been a while.
Thanks! So just for a bit of clarification, if I raised these values in the 850 RPM column and experimented with different values in this part of the table, it may help with that small dip I'm experiencing? Sorry, I know its a bit of a noob question.
EDIT: Sorry, just realized what you are saying! You meant the values from 170 to 190 CLT at 850 RPM.
Yep that's it, the car should be referencing the Closed Loop Idle Target RPM and Coolant for the IAC Duty % it sits at for a second or two before actually entering closed loop idle.
I think it's actually the higher cells you want. If you go look at a datalog you can see a "closed loop target RPM" value while even while not in closed loop idle. It is usually higher than the final target idle RPM, probably because of dashpot adder, mine so happens to be 1138 RPM most of the time.
To get it to start out with more duty cycle so it tapers into a lower idle smoother you should adjust whatever cells you have megasquirt targeting. 1138 or the 1100 row in my case.
Truth told, megasquirt probably never uses the 850 RPM cells, or much of anything but the highest RPM rows. It will have entered closed loop at a higher target and so grabbed an initial value to start modifying much earlier. Once it's done that it won't use the table again. You might be able to scale the table so it has more higher RPM cells, too. Might make it more consistent when it's cold and the target RPMs would then usually be a lot higher since the base target idle while cold is usually 1100RPM+
I think it's actually the higher cells you want. If you go look at a datalog you can see a "closed loop target RPM" value while even while not in closed loop idle. It is usually higher than the final target idle RPM, probably because of dashpot adder, mine so happens to be 1138 RPM most of the time.
To get it to start out with more duty cycle so it tapers into a lower idle smoother you should adjust whatever cells you have megasquirt targeting. 1138 or the 1100 row in my case.
Truth told, megasquirt probably never uses the 850 RPM cells, or much of anything but the highest RPM rows. It will have entered closed loop at a higher target and so grabbed an initial value to start modifying much earlier. Once it's done that it won't use the table again. You might be able to scale the table so it has more higher RPM cells, too. Might make it more consistent when it's cold and the target RPMs would then usually be a lot higher since the base target idle while cold is usually 1100RPM+
Originally Posted by HowPrayGame
Good point, the IAC Duty % in the log is 29.8 when the car dips in idle so yeah it doesn't actually reference the lower cells
Thanks guys! I'm thinking of scaling the table 900, 950, 1000, 1100, 1200, or 950, 1000, 1100, 1200, 1300 and then adjusting the higher RPM cells. I'll give it a go after work and see what I can do!
So something interesting happened. I started the car a few times after it warmed up, and it didn't happen this time. I didn't change anything, but it looks like its idling fine (although a little rough). I wonder what happened that its now stabilized. I guess I will see if it stays that way! Decided to take the car out for some auto tuning with VEAL, and noticed that coasting with no throttle applied while in gear causes the car to kind of jerk back and forth. I also probably need to play around with accel enrich, since it does go very lean when lightly applying throttle.
10-18-2020, 09:39 PM
0
