its gonna b faster the colder out.... simple logic

 

Thanks

 

That's still missing the point. The point is that I can make my tune's mixture as flat as a board, but that flat line moves up/down based on ambient temperature. If I am comfortable with, say, 11.3 as my max WOT AFR, do I tune it in hot weather or cold? If I tune it for 11.3 in hot weather, I'm running too lean in cold. If I tune for 11.3 in cold weather, I'm giving up some extra power in the summer time.

If we can tweak the ECU's temperature compensation cals so that the AFR doesn't change with ambient temp, we can squeeze out the best power with a stable tune.

C6C6CH3vo: keep us informed as you do further tweaking of the temp compensation.

 

Actually, this isn't quite true. Your wideband isn't really measuring AFR, it is measuring equivilency ratio (effectively 1/lambda). From this it calculates lambda, then multiplies lambda by a fixed value for the fuel (14.7) to roughly estimate AFR.

Lambda is a fuel independant, relative location to stoichiometric ratio (1.0 in lambda). So, rather the air is hot and thin, or cold and thick, running at the same lambda (or the rough AFR calculated from it) SHOULD give you the same level of performance optimization.

The reason it often does not is that requivilency ratio is a *relative* measurement. It depends on the partial pressure of oxygen. This changes, fairly significantly, with altitiude changes and large pressure/temp changes. A Bosch sensor is originally calibrated at standard pressure and temperature. So, if you tune at the beach, then tweak and rally race in the mountains to the same numbers on a basic UEGO wideband, you are too rich and lose power.

However, there are a couple of ways to address this. Some wideband controllers offer a 'free air calibration'. In this case, the wideband will make the partial pressure correction for you (go to the mountains, do a free air cal, then tune back to the same numbers that had you cooking at the beach). The other alternative is to use a temp probe, a barometric pressure sensor, and a 'puffer' of 1% calibrated gas. At each location, check pressure, temp, and offset to cal gas, then use the offset and a correction table for your readings.

We actually have an engine builder for LSU aircraft engines in Europe that uses a variation of this with our widebands (http://www.ulpower.com). They omit the gas puffer, they just do a free air calibration on the wideband before engine start and record pressure and temperature. But since they can't do free air calibrations while the engine is running and the aircraft is climbing, they use temperature and pressure to calcuate a correction.

-jjf

 

I do exactly what jjf suggested on my race car when I log AFR. I live @ sea level, but I race in the high desert. So I do a free air calibration before I log AFR on my race car. This way I can get the most consistent AFR readings.

Whenever I log an EVO with the LM-1 I do a free air calibration on the sensor. IMO, the more you re-calibrate the sensor the more accurate the results.

 

One question is why are you worried about it when the engine does not knock?
With rich AFR (in the 11's), you are not running at the most efficent AFR for power. That would be in the 12's. Running rich instead you are controlling the heatup of the yet unburned gases by the combustion and compression induced pressure (and therefore temperature) increase. Running rich releases less thermal energy (most would be released at stoich) and therefore you keep the unburned end-gas (hopefully) below the self-ignition point. Knock is almost solely dependent on the temperature of the gases pre-burn. Rich running actally allows more pre-knock chemicals to form, that's why there is also rich knock, where making the engine richer INCREASES knock probability. Lower outside air temp causes lower IATs after the turbo which causes lower initial temperature before compression, which results in lower gas-temperature in the cylinder after compression. Hence you are further away from the knock threshhold. This means the AFRs can be closer to whats optimum for highest pressure release during combustion (12.5 to 13) and you can create more power. You get then more power not only from higher gas density (colder air), but also because of slightly better combustion. Same is true if you increase turbo and/or intercooler efficiency. Lower IATs allow to run somewhat leaner (therefore hotter burn), thereby compensating for the lower IATs, but creating more power than just from the higher air density alone.

Regards,
Klaus

 

Can someone post an update on this?
Is anybody still successfully using the air temp compensation table?

My afr's differ by around 0.5 as well... Its a bit frustrating as it is cold at night and hot during the day here these days, so temp swing is pretty big.

I mean there's definitely a need for this sort of correction as hot ammbient air will be more prone to lead to knock, but 0.5 or even 0.7 swing seems a bit too much.

 

my correction factor values are dead on. I run the same AFR no matter what the temp is outside. I was able to get this done correctly by tuning in xx deg weather, then logging constantly in cooler and warmer weather and making adjustments to the temp correction tables. For most other standalone ECUs, i use a temp correction excel spreadsheet to set up my correction facotrs, but for some reason, the correction math is off using it in ECUflash. So, i had to manually set it up. It took a lot of time and patience.

i'm still trying to figure out the math for us......it's not making much sense at the moment, but i'm still working on it.

CJ

 

What do you have for those values now?

 

^^Ditto...curious to see if my theory held out or not .