MegaSquirtSanta - Custom Modifications / Firmware
#42
OK I have the early beta of the Santa release, which includes all the features of the AC Idle up release plus MAT correction! Initial results look promising, but they do require careful tuning, because it changes the balance for the closed loop PID.
A picture of the settings menu:
Initial semituned results here:
The idle duty will really rise with increased MAT, but I noticed that the range of valve motion stayed more or less the same. Meaning PID is not exerting more effort at higher temps, and that the MAT correction is actually correcting for the less dense air.
Subjectively, the idle felt more stable at high temps. I turned off the A/W intercooler pump and let it heatsoak to 160F, but idling was rock solid. Killed the engine, started it up, no heatsoak starting issues. I still need to tune it because at high temp, the correction was a bit too much for the no-AC idle, and dragged the idle 50 higher than the target. But honestly, I couldn't really tell, it only came up in the scatter graph.
Great job Santa!
A picture of the settings menu:
Initial semituned results here:
The idle duty will really rise with increased MAT, but I noticed that the range of valve motion stayed more or less the same. Meaning PID is not exerting more effort at higher temps, and that the MAT correction is actually correcting for the less dense air.
Subjectively, the idle felt more stable at high temps. I turned off the A/W intercooler pump and let it heatsoak to 160F, but idling was rock solid. Killed the engine, started it up, no heatsoak starting issues. I still need to tune it because at high temp, the correction was a bit too much for the no-AC idle, and dragged the idle 50 higher than the target. But honestly, I couldn't really tell, it only came up in the scatter graph.
Great job Santa!
#44
Santa,
- Whenever closed loop is re-entered there should be a "closed loop re-entry target RPM adder". This stays for like 1 sec, and then the target ramps down to the normal target over say, 3 sec. For example, target will initially be 1200 RPM upon re-entry. This target will stay for 1 sec, then ramp down to 850 over 3 seconds. This will solve the stopping-at-a-stop-sign idle dip, and it will also work well for stepping on the clutch when lugging the engine at 600 RPM.
- Whenever closed loop is re-entered there should be a "closed loop re-entry target RPM adder". This stays for like 1 sec, and then the target ramps down to the normal target over say, 3 sec. For example, target will initially be 1200 RPM upon re-entry. This target will stay for 1 sec, then ramp down to 850 over 3 seconds. This will solve the stopping-at-a-stop-sign idle dip, and it will also work well for stepping on the clutch when lugging the engine at 600 RPM.
We also mapped the VE curve from start to finish and positioned our idle rpm targets in the middle of downward VE slopes. If the idle valve overshoots target-rpm the VE drops away, adding a natural tendency to return to the target RPM. Lower then target RPM->VE increases. We would let it hop between a couple of these set points instead instead of ramping it down smoothly.
#47
Finally got to try the clutch/neutral closed loop lockout option today... wow, very nice! Much more stable feeling during low speed "parking lot/traffic light" conditions. It does however highlight the need for accurate calculation of the idle valve value during CL idle entry. "Last known good value plus dashpot" needs to be improved...once that is licked and you start closed loop idle with the correct idle duty...we will be very close to seamless...
#48
The VE dips were not tuned in, we just used the normal intake and exhaust resonances. Spent quite of bit of time building the VE curve by mapping the engine every 100rpm or so to a constant Lamdba. The end effect is similar to building fueling and spark slopes but more effective.
#50
Just wanted to add that this works a charm. It was the only way we could get a stable idle on a FS-car (supercharged single.cyl, no flywheel).
We also mapped the VE curve from start to finish and positioned our idle rpm targets in the middle of downward VE slopes. If the idle valve overshoots target-rpm the VE drops away, adding a natural tendency to return to the target RPM. Lower then target RPM->VE increases. We would let it hop between a couple of these set points instead instead of ramping it down smoothly.
We also mapped the VE curve from start to finish and positioned our idle rpm targets in the middle of downward VE slopes. If the idle valve overshoots target-rpm the VE drops away, adding a natural tendency to return to the target RPM. Lower then target RPM->VE increases. We would let it hop between a couple of these set points instead instead of ramping it down smoothly.
#51
Junior Member
Thread Starter
iTrader: (1)
Join Date: Jun 2011
Location: Australia
Posts: 178
Total Cats: 3
You can plod along in gear and it won't enter CL idle at all, never, no mater even if you go off throttle completely and essential the idle is pulling the car along - great for traffic cruise etc.
Avoids the rpm climb when you slam the clutch in, as the pid hasn't been fighting the load, it just casually enters CL pid and continues to idle normally.
A very simple od that took no time for me to code and make work.
You need to wire the clutch switch to pe0 / flex input.
G
PS - I sent you a PM and you've flatly ignored me, wassup with that?
Avoids the rpm climb when you slam the clutch in, as the pid hasn't been fighting the load, it just casually enters CL pid and continues to idle normally.
A very simple od that took no time for me to code and make work.
You need to wire the clutch switch to pe0 / flex input.
G
PS - I sent you a PM and you've flatly ignored me, wassup with that?
#52
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,493
Total Cats: 4,080
isnt this what the RPMdot part of the lock settings are for? or is this another part of the code that ms2 doesnt have? MS3 has an adder to change the RPMdot lockout setting when the a/c is on.
if anything when this happens there would be an idle droop not a climb, as your chugging along at say 1200 rpm in 5th gear, then CL would activate and would want to lower the idle to your target and start closing the valve. Then when you hit the clutch you send the rpms to stalling range...I fought this until i tuned my settings correctly. It took no time at all to make it work
if anything when this happens there would be an idle droop not a climb, as your chugging along at say 1200 rpm in 5th gear, then CL would activate and would want to lower the idle to your target and start closing the valve. Then when you hit the clutch you send the rpms to stalling range...I fought this until i tuned my settings correctly. It took no time at all to make it work
#54
Junior Member
Thread Starter
iTrader: (1)
Join Date: Jun 2011
Location: Australia
Posts: 178
Total Cats: 3
if anything when this happens there would be an idle droop not a climb, as your chugging along at say 1200 rpm in 5th gear, then CL would activate and would want to lower the idle to your target and start closing the valve. Then when you hit the clutch you send the rpms to stalling range...I fought this until i tuned my settings correctly. It took no time at all to make it work
With this mod, none of the above happens.
G
#56
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,493
Total Cats: 4,080
actually it does in the code, but no idle up adder that you havent added yet.
can never think of a time i'd ever need this code function. Correct me if I'm wrong but this is just preventing an idle spike if you happen to be costing without throttle while the idle code is enabled up a hill for some reason and then hit the clutch in?
No, the scenario without this mod is ... you are in 1st gear, idling along, you come to a slight incline and the idle is dragged down to 800rpm, the PID code responds and opens the valve more to compensate and in turn raises idle due to the load, and cycle repeats - with the valve opening all the time to compensate.... eventually, bang you slam down the clutch (to avoid hitting the car in front, or stall) and boom the rpm spikes until the PID catches it etc.
#57
Crap.. that should have been VE dips.
The VE dips were not tuned in, we just used the normal intake and exhaust resonances. Spent quite of bit of time building the VE curve by mapping the engine every 100rpm or so to a constant Lamdba. The end effect is similar to building fueling and spark slopes but more effective.
The VE dips were not tuned in, we just used the normal intake and exhaust resonances. Spent quite of bit of time building the VE curve by mapping the engine every 100rpm or so to a constant Lamdba. The end effect is similar to building fueling and spark slopes but more effective.
#58
Junior Member
Thread Starter
iTrader: (1)
Join Date: Jun 2011
Location: Australia
Posts: 178
Total Cats: 3
Also, I think it makes tuning a little easier (I guess)....
G