- Mon Aug 24, 2020 2:23 am
#44958
• Large engines (compared to 2.0L and under) Appear to prefer MAP based load sensing.
◦ Modifications to My engine include
▪ Edelbrock E-Street Heads
▪ High Ratio 1.8 roller rockers
▪ Mild Cam – Produces good vacuum and smooth Idle but much better Torque and HP
▪ Standard 58MM Throttle body
▪ upgraded 24Lb Injectors.
▪ Combination produces a dyno Proven 330 HP on a Carburetor - a bit more on EFI
• I spent an inordinate amount of time attempting to tune my 91 Mustang 5.0L using TPS only as the load sensing. (Alpha N) While I was able to get a reasonable running engine it was never consistent from one run to the next and TS Autotune could never seem to figure out what the engine wanted as every session modified the entire VE MAP often resulting in messing up what was previously a pretty good result from a previous tuning session. After many, many sessions I finally managed to get a descent tune so then turned my attention to AE.
• Once I turned my attention to Acceleration Enrichment things got REALLY interesting. With TPS load sensing it was not possible to get the engine to smoothly accelerate under a quick throttle stab. In neutral, off idle, this manifested as a minor but obvious hesitation. On the road under load this minor hesitation resulted in what felt like the ignition shutting off and the switching back on with a vengeance. Under heavy throttle in first or second this was violent enough that I was concerned for the safety of the drive train. In reviewing logs it looked like the engine was going lean so the apparently obvious solution was to add fuel to AE.
After many months of playing with settings, driving, adjusting settings, drive some more, rinse and repeat. I saw a comment from a forum user to the effect that larger engines work best on MAP based load sensing. So, having conclusively proved that TPS load sensing does not work I switched to MAP based load sensing. After changing all relevant settings to MAP the engine fired right up but ran terrible. A quick adjustment of the “Normal” idle cells brought the idle into a stable mode. Spent another couple of minutes tuning cells in neutral up to about 2500 RPM. With that done I switched on Autotune and went for a low speed drive in my neighborhood. The difference was immediately obvious in how the engine ran and in how quickly Autotune was able to pull the VE table into line. A couple of highway tuning sessions further proved out this initial impression and resulted in a nicely tuned VE map.
• MAP based Acceleration Enrichment.
With a reasonable VE table under my belt and the engine feeling a lot happier living under MAP based load sensing I turned my attention to AE. Initial trials produced the same results as under TPS based Alpha-N Load sensing. So took some logs of the occurrences and sat down with a bag of cookies and a coffee and studied the logs to determine what was going on. Initially it still looked like we had the same leaning out under a heavy throttle application which puzzled me as I had AE set to the Maximum allowable as a legacy from the TPS based setup. Up to this point my study process on the log had been to jump directly to AE events and review what was happening at that point in time. But seeing nothing obvious I decided to watch the entire log in “real time” so simply hit the play button and let it run.
Reviewing the log in real time made it immediately clear what was going on. By watching the VE table on the right and the Log on the left simultaneously I noted that each and every time the log showed an AE event I immediately lifted off the gas resulting in a vertical drop in the VE table trace. This was followed by a leaning out of the AFR but with a 1.5 to 2 second delay.
Hmmm, why would the engine continue to run rich for 1.5 to 2 seconds after shutting off fueling back to idle??? In pondering this in combination with the fact that it always happened in conjunction with an AE event I finally understood what was happening. As shown in this Screen Capture.
Untitled.png (26.81 KiB) Viewed 9072 times
The graph on the right still has the TPS settings I was using to overcome the severe hesitation. The right shows my current AE settings under MAP based AE Note On the TPS side TPS DOT threshold is set to 30% and fuel loads are essentially maxed out. Compare this to the MAP based DOT (which is essentially the Default curve supplied in the Speeduino initial tune file).
Bottom line is I was pouring so much fuel into the engine on a throttle stab that the engine was sucking in liquid gas which extinguished the spark until it had time to dry out. The leaning seen in the log was the result of the O2 sensor seeing atmospheric levels of oxygen coming down the pipe as a result of the fact that there was zero ignition happening.
When I initially Implemented the MAP based AE I set MAPdot to 50% but was still getting the hesitation under what appeared to be steady state throttle. So back to the log. The two partial screen captures below reveal what was happening.
-27.png (7.47 KiB) Viewed 9072 times
Looking at the white throttle trace it appears that with minor variations the throttle is held steady but as shown in the green trace AE kicks in???
+28.png (8.46 KiB) Viewed 9072 times
A fraction of a second later AE is fully engaged Why when the Throttle was still held steady??
The answer is revealed in the red MAPdot readings. Comparing the top reading to the bottom reading shows us that there is a total MAPdot change of 55.938%. Having initially set the MAPdot threshold at 50% this fluctuation was sufficient to kick in AE which, in a steady state throttle was sufficient to kill the spark. The small variations could be the result of a bump in the road or electrical noise, but what ever the cause I had my answer.
The solution here was to move the threshold back to the default tune recommended 80% threshold.
So, what I learned from all of this is that as has been endlessly stated by countless others is Logs are your friend but you need to couple the logs with your impressions, interpretation and understanding of what the engine is telling you.
Hopefully this information helps others in the future
Things Learned Tuning Speeduino
On a Modified Ford 5.0L
The following are observations I have after spending two summers learning to tune my 91 Mustang 5.0L V8 on Speeduino EFI.
I make no claim that these observations are cast in concrete truths, rather they are what in the end worked for me.
They are offered in the hope that, as happened to me, comments made by other forum users in combination with my own observations about what my engine was trying to tell me, in the end resulted in a killer tune.
Perhaps buried in this is a the missing nugget you need to get your tune done.
That said, Still with me?
Read on.........
On a Modified Ford 5.0L
The following are observations I have after spending two summers learning to tune my 91 Mustang 5.0L V8 on Speeduino EFI.
I make no claim that these observations are cast in concrete truths, rather they are what in the end worked for me.
They are offered in the hope that, as happened to me, comments made by other forum users in combination with my own observations about what my engine was trying to tell me, in the end resulted in a killer tune.
Perhaps buried in this is a the missing nugget you need to get your tune done.
That said, Still with me?
Read on.........
• Large engines (compared to 2.0L and under) Appear to prefer MAP based load sensing.
◦ Modifications to My engine include
▪ Edelbrock E-Street Heads
▪ High Ratio 1.8 roller rockers
▪ Mild Cam – Produces good vacuum and smooth Idle but much better Torque and HP
▪ Standard 58MM Throttle body
▪ upgraded 24Lb Injectors.
▪ Combination produces a dyno Proven 330 HP on a Carburetor - a bit more on EFI
• I spent an inordinate amount of time attempting to tune my 91 Mustang 5.0L using TPS only as the load sensing. (Alpha N) While I was able to get a reasonable running engine it was never consistent from one run to the next and TS Autotune could never seem to figure out what the engine wanted as every session modified the entire VE MAP often resulting in messing up what was previously a pretty good result from a previous tuning session. After many, many sessions I finally managed to get a descent tune so then turned my attention to AE.
• Once I turned my attention to Acceleration Enrichment things got REALLY interesting. With TPS load sensing it was not possible to get the engine to smoothly accelerate under a quick throttle stab. In neutral, off idle, this manifested as a minor but obvious hesitation. On the road under load this minor hesitation resulted in what felt like the ignition shutting off and the switching back on with a vengeance. Under heavy throttle in first or second this was violent enough that I was concerned for the safety of the drive train. In reviewing logs it looked like the engine was going lean so the apparently obvious solution was to add fuel to AE.
After many months of playing with settings, driving, adjusting settings, drive some more, rinse and repeat. I saw a comment from a forum user to the effect that larger engines work best on MAP based load sensing. So, having conclusively proved that TPS load sensing does not work I switched to MAP based load sensing. After changing all relevant settings to MAP the engine fired right up but ran terrible. A quick adjustment of the “Normal” idle cells brought the idle into a stable mode. Spent another couple of minutes tuning cells in neutral up to about 2500 RPM. With that done I switched on Autotune and went for a low speed drive in my neighborhood. The difference was immediately obvious in how the engine ran and in how quickly Autotune was able to pull the VE table into line. A couple of highway tuning sessions further proved out this initial impression and resulted in a nicely tuned VE map.
• MAP based Acceleration Enrichment.
With a reasonable VE table under my belt and the engine feeling a lot happier living under MAP based load sensing I turned my attention to AE. Initial trials produced the same results as under TPS based Alpha-N Load sensing. So took some logs of the occurrences and sat down with a bag of cookies and a coffee and studied the logs to determine what was going on. Initially it still looked like we had the same leaning out under a heavy throttle application which puzzled me as I had AE set to the Maximum allowable as a legacy from the TPS based setup. Up to this point my study process on the log had been to jump directly to AE events and review what was happening at that point in time. But seeing nothing obvious I decided to watch the entire log in “real time” so simply hit the play button and let it run.
Reviewing the log in real time made it immediately clear what was going on. By watching the VE table on the right and the Log on the left simultaneously I noted that each and every time the log showed an AE event I immediately lifted off the gas resulting in a vertical drop in the VE table trace. This was followed by a leaning out of the AFR but with a 1.5 to 2 second delay.
Hmmm, why would the engine continue to run rich for 1.5 to 2 seconds after shutting off fueling back to idle??? In pondering this in combination with the fact that it always happened in conjunction with an AE event I finally understood what was happening. As shown in this Screen Capture.
The graph on the right still has the TPS settings I was using to overcome the severe hesitation. The right shows my current AE settings under MAP based AE Note On the TPS side TPS DOT threshold is set to 30% and fuel loads are essentially maxed out. Compare this to the MAP based DOT (which is essentially the Default curve supplied in the Speeduino initial tune file).
Bottom line is I was pouring so much fuel into the engine on a throttle stab that the engine was sucking in liquid gas which extinguished the spark until it had time to dry out. The leaning seen in the log was the result of the O2 sensor seeing atmospheric levels of oxygen coming down the pipe as a result of the fact that there was zero ignition happening.
When I initially Implemented the MAP based AE I set MAPdot to 50% but was still getting the hesitation under what appeared to be steady state throttle. So back to the log. The two partial screen captures below reveal what was happening.
The answer is revealed in the red MAPdot readings. Comparing the top reading to the bottom reading shows us that there is a total MAPdot change of 55.938%. Having initially set the MAPdot threshold at 50% this fluctuation was sufficient to kick in AE which, in a steady state throttle was sufficient to kill the spark. The small variations could be the result of a bump in the road or electrical noise, but what ever the cause I had my answer.
The solution here was to move the threshold back to the default tune recommended 80% threshold.
So, what I learned from all of this is that as has been endlessly stated by countless others is Logs are your friend but you need to couple the logs with your impressions, interpretation and understanding of what the engine is telling you.
Hopefully this information helps others in the future
)
- By hoopstar