Sorry for the misleading thread

When I run logs with mitsulogger and evoscan I am getting very different load numbers. Shouldn't these programs be using the same formula to calculate load? It gets confusing when you are trying to tune a cell on a map but are getting two different readings with each program.

I guess I'd rather trust evoscan because that is what Im getting my wideband reading from, but it makes me wonder when mitsulogger says im hitting 290++ load and evoscan only says Im hitting 250 calc. load.

Maybe I am misunderstanding something

 

The load numbers are calculated differently on both Programs.. Mitsulogger uses Injector scale and Injector voltage latency values to calculate load. Evoscan uses some embedded value for Injector scale, and the car's voltage to calculate.

Mitsulogger reads higher than Evoscan because of this calculation, in general it reads about 20% higher, I'm not sure why because the calculation is what was provided to me and was considered to be accurate. You can use whichever calculation you want in Mitsulogger if you feel one is better than the other..

The bigger question is which of the two load values is closest to your car? Do you have stock injectors and does your reflash have the injector scale or the latency values altered?

 

One more thing to point out - I read lower load with evoscan v0.98 than I do with my v0.97 which reads lower than the mitsulogger


Interesting thing is that at lower RPMs my maps correspond to output more with lower load levels than read. At higher RPMs where my AFR is flatter and leaner and my timing looses it's readable precision (22 degrees vs 4 degrees) I couldnt tell you how close it is.

 

If you look at the new Evoscan Data.XML file, the injector scale is 464 instead of 513..... I always wondered where that calculation came from as its making assumptions about a fixed injector latency instead of using a known value and interpolating from voltage.

I would actually prefer to use a calculation similar to the Evoscan one, if interpolating injector latency from the 12v and 14v values was used..

 

Jack- I have the stock injectors, scale, and latency.

Last night I made a couple pulls on evoscan and made some changes with the assumed load from it and almost perfectly dialed in my fuel map with one try, so I guess it works.

Props to you on mistulogger though Jack, its so easy to use and has never crashed on me like evoscan. If it had the programming for my lm-1, I would probably only be using that.

Thhaaanksss

 

Its interesting you mention that, in Evoscan .98 (not sure about earlier versions) his formula says the injector scale is 464 and not 513.. a bit of fudging for the stock injectors but unfortunately makes scaling for larger ones much more difficult (editing the load formula in the XML file is difficult enough for many people)

I'll try to spend some time sorting out the load calulation in Mitsulogger, I suspect the problem now is specifically that the latency used is for the 14v value, when it can vary from 13.5v or so.. and if your 12v and 14v settings are different than stock, the "Curve" between the two is also no longer the same, and it throws off the calculation.. Since there is currently no interpolation at all, then it will read a little higher all the time.. I promise i'll address that.

 

That makes sense now. Do you have plans to add the LM-1 for widebands?

 

Yep, I am hoping to get it supported in the next week or two.. Not only the LM-1 or LC-1 but any of the devices that can be logged with the MTS chain.