Is lambda and knock sensors necessary on a race engine
Discussion
I am running a E46 M3 S54 BMW race engine on standalone ecu with 12.5/1 compression ratio (Standard is 11.5/1) and was thinking if the car is mapped properly on race fuel or mapped properly for pump fuel it is not necessary to run Lambda control
The engine is also not running knock sensors but has vanos retained
What is the difference / advantages between wideband and standard lambda
Any info would be much appreciated
The engine is also not running knock sensors but has vanos retained
What is the difference / advantages between wideband and standard lambda
Any info would be much appreciated
In a word no. At least IMO 
But you want logging for sure to know you are running optimally.
Lambda is only for closed loop stoich fueling which on a race engine probably isnt used much.
That being said I believe for lemans etc running leaner at WOT is done to conserve fuel. Also of course perfect combusions is at stoich, so if you have a well designed engine that may be the optimal ratio for MBT.
The question therefore is more like do you want to run in open or closed loop?
If your ECU can run widebands in closed loop, then thats far better than NB aka lambda sensors.
So really it depends on engine design, cooling and what type of racing you are doing as to what your optimal engine management solution will be..
But you want logging for sure to know you are running optimally.
Lambda is only for closed loop stoich fueling which on a race engine probably isnt used much.
That being said I believe for lemans etc running leaner at WOT is done to conserve fuel. Also of course perfect combusions is at stoich, so if you have a well designed engine that may be the optimal ratio for MBT.
The question therefore is more like do you want to run in open or closed loop?
If your ECU can run widebands in closed loop, then thats far better than NB aka lambda sensors.
So really it depends on engine design, cooling and what type of racing you are doing as to what your optimal engine management solution will be..
No, they're not necessary.
If the engine is mapped properly you will never need to look at the lambda values. It'll just be right. Closed loop control is pointless on a racing car, as you will not spend much time in the areas where closed loop is useful.
Lambda sensors don't just work around stoichiometric - wideband sensors are 'accurate' from AFRs 10-16 or thereabouts.
If you DO run a wideband lambda sensor, be aware that the best air:fuel ratio isn't a fixed value, and the actual value at any given RPM/Load can only be determined on a dyno when you get a power reading. Using a lambda sensor to set the AFR at a near-constant value throughout the rev range won't be 'bad' (assuming you choose a sensible value - 14:1 at idle and cruise, 12.7:1 at WOT etc etc), but might not be perfect.
I had a terrible misfire on my racecar, so I ended up fitting a wideband sensor and playing with the map to try and find the cause and effect a cure. Several people on here helped with that learning process. With bigger injectors, tweaks to the map, tweaks to acceleration fuelling etc etc the misfire has gone. I am now at the stage where I will remove the sensor to save weight/amps as it is of little benefit now.
If the engine is mapped properly you will never need to look at the lambda values. It'll just be right. Closed loop control is pointless on a racing car, as you will not spend much time in the areas where closed loop is useful.
Lambda sensors don't just work around stoichiometric - wideband sensors are 'accurate' from AFRs 10-16 or thereabouts.
If you DO run a wideband lambda sensor, be aware that the best air:fuel ratio isn't a fixed value, and the actual value at any given RPM/Load can only be determined on a dyno when you get a power reading. Using a lambda sensor to set the AFR at a near-constant value throughout the rev range won't be 'bad' (assuming you choose a sensible value - 14:1 at idle and cruise, 12.7:1 at WOT etc etc), but might not be perfect.
I had a terrible misfire on my racecar, so I ended up fitting a wideband sensor and playing with the map to try and find the cause and effect a cure. Several people on here helped with that learning process. With bigger injectors, tweaks to the map, tweaks to acceleration fuelling etc etc the misfire has gone. I am now at the stage where I will remove the sensor to save weight/amps as it is of little benefit now.
tristancliffe said:
No, they're not necessary.
If the engine is mapped properly
And that, my friends, is the crux of the matter! Now, even if you map it properly on day1, do you know that on day2 it will still be absolutely spot on??If the engine is mapped properly
Even in F1, where you would have thought they would be the very definition of "properly mapped" they re-optimise fuelling during the race with wide band sensors......
The simple fact of the matter, a decent wide band sensor will be able to get your exhaust AFR to within 1 or 2% of target repeatably from day to day, any other calibration (including say baro/humidity corrections etc) will struggle to do that.
Regarding the knock sensors, well it depends if your engine is knock limited! If you only ever run at high rpm, and have sufficient knock margin then clearly knock sensors are no benefit. If however you run in the mid range (such as out of slow corners etc) and here your engine is knock limited then a proper KCS will allow you to stay closed to this limit than an open loop system.
Slinky said:
How much weight are you going to be saving with that removal?
Surely, from a data logging perspective, having lambda is an important indicator?
Surely, from a data logging perspective, having lambda is an important indicator?
Slinky said:
How much weight are you going to be saving with that removal?
Surely, from a data logging perspective, having lambda is an important indicator?
My car is at least 10kg over the weight limit, so currently on a weight reduction programme (me and the car). Electrics are total loss, so saving amps important too. Surely, from a data logging perspective, having lambda is an important indicator?
Lambda trace has been good for ages. Don't see how useful it is if I never make any changes based on it anymore. Probably 80% of race engines (I.e.club level, money more important than last percent of performance) never run a lambda sensor ever. I would consider oil pressure far more important.
I don't F1's choice have much in common with my own. I have a smaller budget and a bigger stomach.
I would estimate the weight savings for the lambda sensor and loom to be approx 300gm tops, and if you don't connect the probe heater (which you won't need on a race car running at high load) then there is no power saving to be had either. For me, knowing my fuelling was optimum is easily worth 300gm !! (i.e incorrect fuelling could easily loose you 10-20bhp, something that a 300gm mass penalty would not offset)
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff