# ROC and Delta math modifications

## Recommended Posts

Hello!
1.  Can the Map Delta, Tps Delta and Engine speed ROC math be changed?  Specifically the time constant?

2.  I assume the above question is no.  So i've started to try to write my own channel but I'm not smart enough to get the results i'm looking for.  I've started with the equation of CNT(A)   A being Tps Delta.  It is not showing me negative values on - delta values.  I'm trying to lengthen the time of the delta calculations and see if that gets me the data I'm looking for.

Outside of overrun fuel cut conditions every car I've tuned for 25 years on every standalone platform, during slow decelerations ie throttle reduction to keep distance from a slower car on highway, Lambda is always well below my targets.  When steady state, or when accelerating through the same areas of the map, the lambda is where I want at .9.  While I try to use 02FB to correct it, I think if I can write a math channel that modifies the time base of the 3 channels listed then I can use that math channel as an axis on 4D fuel and correct what I've wanted to for YEARS.  I tested this morning on a customer car and wrong the above math channel and entered it as an axis for testing.  Now I just need to perfect the math.

I also tune OEM ecu's and for years have been wondering how they combat this condition during closed loop fueling without the 02's even twitching since fueling is perfectly reaching targets.  Multipliers in the background i'll never understand i suppose

Thank you for any guidance!

##### Share on other sites

This sounds like the Decel Enleanment. I'm still learning and playing around with ideas to address this myself. the 100ms thing is a constantly updated number that calculates the rate over the last 100ms, so, while it may be useful there's still usually too much fuel going in, but it doesn't last as if you took foot off from 100% tps to 50% tps it would register -50% ROC and then it would be 0%  ROC within 100MS - we need something a wee bit smarter. Maybe something with the estimated air in cylinder, with some thesholds - rather than the 100MS Delta..

Currently I'm messing with a 4d table that can pull up to 60% of the fuel out..  screenshot attached, I will try the higher TPS% areas with more aggressive numbers if it does make a difference (the area on the right  in the screenshot, with the 0s).

Ultimately .. Decel Enleanment is a feature I'd love to see. I had hoped the async injection would sort of cover it.

##### Share on other sites

You could use the knock tps/MAP/RPM deltas as they have user configurable time based filters. You might need to turn normalised knock mode on the see them.

The rich drift when you initially lift off is some of the fuel puddle evaporating, oem's have very complex strategies to reduce this effect since to meet emissions targets they need to be as close to lambda 1 as much as possible, but in the aftermarket world it really isn't going to be achievable with the typical hardware, budget and intelligence involved.  Decel enleanment or "injecting less fuel" is only a very small part of the wall wetting strategy, the problem is the fuel is already in the port clinging to the walls, so pulling fuel out only helps a small amount a few cycles later after the bulk of the puddle has evaporated.  OEM's will have the fuel injector spray pattern and injection timing well optimised to minimise the puddle size in the first place, they will have many other tricks to keep the MAP and puddle as stable as possible such as slow the throttle movement down, move the camshafts and use EGR etc.

##### Share on other sites

• 2 weeks later...
On 5/19/2024 at 12:46 PM, Adamw said:

You could use the knock tps/MAP/RPM deltas as they have user configurable time based filters. You might need to turn normalised knock mode on the see them.

I've manage to get some OK results using this (see screenshot with graph - the rich dip was much better than usual)). I've gradually bumped up and I'm keeping this all outside of the delta of CLL. I recognise getting it like OEM is impossible bit some progress: I managed to get under 10L/100KM recently on a drive that involved a 2 hours of mountain road blasting and 20 minutes of 0 mileage idling. I was also thinking of trying the 'AFR target error' instead of the TPS % on the X axis.

I recognise this approaching the point of diminishing returns, too

## Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

×   Pasted as rich text.   Paste as plain text instead

Only 75 emoji are allowed.