pcmhacking.net

Hi all, I have been doing a log and calcVE update everyday for the last 2 weeks. It's a VNV6 on 12P. I believe I had the VE table very close yesterday, well the closest it's been anyway! Today I done another log and done the calcVE update to the VE table and then smoothed it, took it for another run and it was worse! I think my problem is when I'm smoothing the table manually. How do I know by looking at the 3D graph which points need touching? It seems I'm touching areas that don't need touching just to try and get a smooth table. How close does the VE table have to be to be able to say that it doesn't need to be touched again? How can I tell exactly which cells need to be adjusted and which ones to leave alone by playing back logs?

Once your close you will only be adjusting areas you didnt hit between multiple points, just to smooth small areas out. What might seem like a small 5% change to VE is a fair bit and can throw out the areas around that (remember the ECU interpolates between table points).

Maybe go back to your tune that got the target and actual pretty close, switch on closed loop and BLMs to see what it does. Just have a play around and will soon see what works and does not work. Could also leave it with closed loop and BLMs off, get some more data and average it out over a few days.

As suggested above Im going to switch on closed loop & BLM's to see if I can get closer to my final tune. So do I go for a drive and take a log then playback on the BLM table histogram on the running average view?

Then if there are any cells that are not reading 100, I need to go to those cells in the VE able and increase or decrease those numbers?

If this is the correct process how do I work out how much to adjust them by?

Cheers,
vc81

that's pretty much it, if you have a BLM of 106 and ve of say 53 > 53 * 1.06 = 56.18 (6% increase) likewise 97 would be 53 * .97 = 3% decrease. you could also change the histogram calculation to make it the 1.06 or .97 straight up if you wanted to save some time, but if its 106 pretty easy to convert in your head as you go.

Thanks for your reply holden202t. I've been logging blm's and adjusting each cell as needed using the calc above. I've been getting readings of 100.444 (2000 counts) etc so this means fuel is being added to bring the reading in that particular cell to 14.7. When I look at the wideband log historgram it is showing that particular cell is 14.2. How long does it take to get that cell to 14.7? If 2000 counts in that cell is not enough how many counts are needed?

Thanks,
Vc81

as long as you are around +/-10 of 100 its tuned i prefer on the slightly rich side of 100 to allow for aging/deteriorating fuel system and o2 sensors.you will find it wont sit on 14.7 even though its commanded as narrowband swing volts etc play a part in the real final ratio some tunes have a different volt settings for the o2 and can make the engine run richer or leaner than commanded depending on what style o2 sensor etc.call it done and go onto tweaking other parts

honestly something like 100.444 i wouldn't even waste my time on, thats why i made the blm converter spreadsheet, so i could format the numbers into zero decimal place or 1 decimal etc ....

also just want to confirm with your 2000 counts, are you showing the history average to get the 100.444 ?

Yea holden202t that count of 2000 was history average view.

Delcowizzid do you mean a reading anywhere between 90 to 110 is ok and fuel tuning is done? So a reading of 105.000 wouldn't be worth worrying about?

yeah thats how i read that comment!

thats correct.changes in daily weather will make blm's wander each side of perfect youll just chase your tail back and fowards.you could tune anywhere withen the max and min blm limits and it would drive ok once in closed loop but the closer you get it the better it will run untill its warmed up and go closed loop.