pcmhacking.net

Okay, then what does cause the spikes to occcur?

Is your cal setup correctly now? As mentioned earlier, you need to have the stoich setting at 14.7 and then the target AFR also at 14.7 for the areas you want closed loop to work.

Is you O2 sensor working correctly? If its lazy it wont work well. What are your BLMs like at the moment? If they are miles out you might have some work to do first. Narrowband VE learn uses the INT value and instead of adding to the BLM value when the INT reaches a threshold point it adds/subtracts from the VE table, resets the INT and starts again.

Yep, the cal is close, the AFR table is mostly 14.7. The O2 is good, except a lean spike will kill it momentarily.

Currently have both wideband and narrowband learn enabled.

Still learning, and logging today (can post them if you want to see...)?

Idling is bit rich, maybe timing could go higher, torque is okay but seems a bit flat.

Be great to have your input on them at this point?

Being a reasonable cam oxygen dilution from overlap can be a problem at low RPMs and narrowband sensors get lazy from low temp too. Increase the min VE learn RPM to maybe 2k RPM. Dont have both narrow and wideband learn on either, if you are really wanting to play with it just turn off the wideband one.

But seriously I would just turn both off and tune from BLMs for the areas in closed loop, much easier. It only takes a couple of drives with logging and you will have it closer than messing with VE learn and smoothing the cells you dont hit often enough.

To make it easier, and if your wideband is setup right, try this ADX. It has a new item 'calc VE'. You can (roughly) calc the required VE from the error in your target to actual AFR on the fly which makes RT tuning easier for idle, and if you are careful can then use the history table to do more of the VE table from logs - just look at the running sample count and running range in the table to decide if the data is bogus.

Thanks VL400, will do as you suggest... But I'd like to try with one learn on first, why not turn off narrowband Instead leaving Wideband on?

BTW cam is one step above 'mild street', and intake valves are bigger with stronger springs.

That's a cool ADX from VL400.

Engine isles at 55-60 kpA, does that mean some defining limits need to be changed, like 'IAC - Idle MAP Threshold' that's got a range 14 to 34 kpA?

And, 'BLM Params - Idle Cell [High Vacuum/Low MAP] Learn Conditions'? I've already changed these ones to 40 & 60...?

VL400 wrote:To make it easier, and if your wideband is setup right, try this ADX. It has a new item 'calc VE'. You can (roughly) calc the required VE from the error in your target to actual AFR on the fly which makes RT tuning easier for idle, and if you are careful can then use the history table to do more of the VE table from logs - just look at the running sample count and running range in the table to decide if the data is bogus.

If you use this "rough" method to fine tune VE beyond idle, you're looking to only SELECT a cell with a low Running Range value, AND a high Running Sample Count - above 30.

As your tune gets 'closer' you'll find that the SELECTED values will almost match what is already in your VE Table. Avoid changing extreme values that don't, look right among the surrounding values.

Expect only a few changes. i.e just use it for your idle

i wouldnt use the wideband ve learn it looks sketchy in the logs ide run the narrowband

delcowizzid wrote:i wouldnt use the wideband ve learn it looks sketchy in the logs ide run the narrowband

Yeah,.. I was just thinking that too. When I display a chart of a wideband learned VE Table it's like every second column is updated - is that what you meant?

Or was that the one I put through the wideband converter spreadsheet... can't recall, but the wideband learn isn't smooth, it's got extreme values 'peaks & troughs'. What could cause that?