Something is wrong. Megatune seems to show the correct manifold air temp. However the log does not. I used easytherm, and think I did so correctly. one problem is that is not smooth as you can see from the log, and also the value is way off. Does anyone know what i did wrong?

 

thats just the way the MAT (and CLT) sensors are, since they are an analog signal mapped to a course digital input, it ends up looking very choppy, especially when the temp is right between two of the digital input spacings. Check to make sure that both megatune and megalogviewer are displaying in the same units. MLV could be in *F and megatune in *C

 

Good point. I can see where choppy comes from.

I thought about the units, but could not find a place to change it in MLV. Megatune shows both CLT and MAT in F. and in MLV, CLT seems to read correctly (F) compared to Megatune.

 

There's nothing wrong with that log. MLV autoscales so you're looking at a 15 degree span on that graph. I'm pretty sure you can change that to a static scaling display in MLV. The logged representation of many of the runtime variables is in 256 discrete steps (8 bit value). The code itself keeps certain inputs to certain calculations in much higher resolution when it can make a difference. The tradeoff is execution speed of the code vs. resolution. This is NOT a "speed difference" a human observer would see in any way. The designers of all of the code variants have worked very hard to make the micro processor do a whole lot of stuff seemingly in parallel but is all interrupt driven embedded code. The program execution flow can go many different ways depending on what interrupt has recently triggered. What these guys have done (Bowling, Grippo, James, Ken, etc.) is very impressive. No, I don't get paid for that opinion but I am a consumer.

 

My apologies - if your values don't agree there IS a problem. I was merely commenting on the percieved choppiness. Toplessfc's advice is spot on.