Auto VE tuning
Discussion
Yeah in theory it should be smooth, but sometimes its hard to get it exactly. Usually get its close +/- 1 or 2% or so and call it a day. Some of the variance will be down to intake air temp modeling. You can hammer the table with the smooth function if you like. But it will probably end up a little lumpy. Closed loop will sort most of it out you just need to make sure WOT is accurate. Perfection is almost impossible, hence why the factory use the MAF, MAP and O2 sensors to help QA and control their modelling errors.
you can get a smoothing table from efilive, you punch in your ve values and it then smooths the gaps. I had mixed response with it and found you are better off just doing the AutoVE manually. You will get steps, jumps some spikes etc.
My approach to get it final was to select the spark map, as I didnt have time/money/resource I used the HSV spark table. Next was to do the AutoVE but to filter out all the tables below 80kpa which enabled me to tune for WOT. Then the rest was tuned under normal driving. Ended up with a nice table, I could have optimised it a bit more at +5800rpm but tbh whn are you ever in that range?
Unlike Rich I went back to MAF closed loop but with LTFT disabled. This was on an LS1 not with the newer e40/38's which I believe is a b
h to work on
My approach to get it final was to select the spark map, as I didnt have time/money/resource I used the HSV spark table. Next was to do the AutoVE but to filter out all the tables below 80kpa which enabled me to tune for WOT. Then the rest was tuned under normal driving. Ended up with a nice table, I could have optimised it a bit more at +5800rpm but tbh whn are you ever in that range?
Unlike Rich I went back to MAF closed loop but with LTFT disabled. This was on an LS1 not with the newer e40/38's which I believe is a b
h to work onMadMaxHSV said:
Thanks Rich.
It's probably pretty close just being to OCD.
Speaking of IAT, is there a table for VE adjustments or is it done automatically using gas constants?
Older ECU you had a custom table.It's probably pretty close just being to OCD.
Speaking of IAT, is there a table for VE adjustments or is it done automatically using gas constants?
Newer ones dont. But you can sometimes hack things with IAT modification to injector flow etc.
However ideally the tables for charge blend should deal with changing IAT, but often not all the tables are available.
Im using Injector flow vs IAT at present, but know its not ideal.
I am trying to get the idle sorted on the E40. The Efilive idle tutorial is only for the LS1's so none of the tables discussed are applicable.
I've just been out today and just upped the desired idle and with the cam it sounds fantastic at 850rpm.
Still a little niggly to start though.
Like you MM, I got the AutoVE started and got about 2 hours of logging only for the laptop to throw a wobbly.
Glad you started this thread. I am bookmarking it for future reference. Maybe we could bounce stuff around?
I've just been out today and just upped the desired idle and with the cam it sounds fantastic at 850rpm.
Still a little niggly to start though.
Like you MM, I got the AutoVE started and got about 2 hours of logging only for the laptop to throw a wobbly.
Glad you started this thread. I am bookmarking it for future reference. Maybe we could bounce stuff around?
Edited by MyM8V8 on Sunday 28th June 20:20
Depends on what its doing around idle. Idle is the hardest to get right on a modified engine. All the load points are easy. Its idle transitions that suck. Also cold and warm starts etc.
Timing, fueling and airflow obviously mess with it. Fueling hopefully is closed loop, so that leaves airflow and timing. Usually with a cam you will want a little more timing, this is because it takes longer to burn near idle because overlap diluting the charge. But maybe only a couple of degrees or so.
Also airflow may need to go up a little, which automatically will add fuel again not too much.
Id probably look more to airflow with a cam, then move onto spark once you have it as stable as possible with airflow changes.
Timing, fueling and airflow obviously mess with it. Fueling hopefully is closed loop, so that leaves airflow and timing. Usually with a cam you will want a little more timing, this is because it takes longer to burn near idle because overlap diluting the charge. But maybe only a couple of degrees or so.
Also airflow may need to go up a little, which automatically will add fuel again not too much.
Id probably look more to airflow with a cam, then move onto spark once you have it as stable as possible with airflow changes.
toothrot said:
have you? Have to say the older ecu = a gem for tuning, seems the newer ecu's still are a bit of mystery even over on the forums!
I even have a problem keeping up with Gelfs AutoVE tutorial on the E40's. Still can't work out how you process the info from 2 x wb's easily. It'll come.it's the same with the auto boxes, all the stuff there is for the 4L60's. All the tables are differtent although the theory is the same.
I think the LS1 is relatively easy, beacause there arne't as many tables. I think the memory in LS1 is 540k? E40= 1 mb, E38/67=2meg
ringram said:
Depends on what its doing around idle. Idle is the hardest to get right on a modified engine. All the load points are easy. Its idle transitions that suck. Also cold and warm starts etc.
Timing, fueling and airflow obviously mess with it. Fueling hopefully is closed loop, so that leaves airflow and timing. Usually with a cam you will want a little more timing, this is because it takes longer to burn near idle because overlap diluting the charge. But maybe only a couple of degrees or so.
Also airflow may need to go up a little, which automatically will add fuel again not too much.
Id probably look more to airflow with a cam, then move onto spark once you have it as stable as possible with airflow changes.
understand about the load points but idle is one area where o2s really help too. Gareth, might be worth just trying with one wb keeping it simple? Did you go pre or post cat with your wb's? Timing, fueling and airflow obviously mess with it. Fueling hopefully is closed loop, so that leaves airflow and timing. Usually with a cam you will want a little more timing, this is because it takes longer to burn near idle because overlap diluting the charge. But maybe only a couple of degrees or so.
Also airflow may need to go up a little, which automatically will add fuel again not too much.
Id probably look more to airflow with a cam, then move onto spark once you have it as stable as possible with airflow changes.
Yeah you can go by one wideband. I did my current engine with just one. Fueling on vixypys dyno was pretty well bang on.
Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
ringram said:
Yeah you can go by one wideband. I did my current engine with just one. Fueling on vixypys dyno was pretty well bang on.
Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
Can you find me one.;)Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
MyM8V8 said:
ringram said:
Yeah you can go by one wideband. I did my current engine with just one. Fueling on vixypys dyno was pretty well bang on.
Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
Can you find me one.;)Just use the results from one to start with. Though ideally you could create a custom pid to add both together and average the result.
But Id worry about that down the line once you have got to grips with the result from one. There should be premade pids for single wideband inputs etc.
ringram said:
Scan tool, under PIDs "O2-Wide"?
You need to choose the serial wideband for the brand you have. Also under "External" the Wideband input for the serial connection.
Ill give you a run down on Saturday
I have that PID already. Thats for picking up the serial info from the LC-1 on the V2. What you mean is a calculated PID to average the two banks.You need to choose the serial wideband for the brand you have. Also under "External" the Wideband input for the serial connection.
Ill give you a run down on Saturday
Gassing Station | HSV & Monaro | Top of Page | What's New | My Stuff