Soo, in reviewing a recent data log, I am getting negative calculated load values, after I let off the throttle and the motor is engine braking (Not under the "throttle hang thingy" and my TPS is reading 13 (No throttle). The question that I have is that in looking at my AFR's, under that condition, my AFR's are reading 10.1-10.3's.. What the Hell? I thought that unless I am in the throttle hang thing, I should be reading very high AFR's, ie. only pushing air through the motor? :iamreallyconfused: Or am I just forgetting about something?

 

what are your modifications?

normally when you let off the gas the afr runs up to the twenties. if you leave a slight pressure on the pedal you will see a rich condition.
how log do you read those rich numbers.

afraid I am not much help, can only think you have some intake mod

your title might not be to indicative of the question you are asking,

 

Negative load happens in reverse, I once hit negative 7000 rpm in reverse

 

I have observed negative load. It is a function of the load calc not being perfectly in sync with the actual airflow. Think of it in terms of this:

The accepted formula was hz / rpm * 852, right? So you let off the throttle after a particularly energetic pull. Lessay ran it up to 7K and then let off suddenly to idle load (but actually HIGH vacuum that you cant see at idle). Idle airflow is usually around 32hz or so.

32 / 7000 = .00457 *852=3.894

The simple fact is that at that rpm and the the increase in vaccum the calc'd load will end up at close to 0. But we arent using just that formula for the calc'd load, we are using an algorithm that apparenly causes EvoScan to post negative numbers. I havent bothered to ever break it down, but I would surmise the negative load happens if the actual load were less than 1 or some low number like that.

C6C6- I had -39 counts of knock from doing that as well.

 

The calculated load I came up with using IPW and AFRMAP is rubbish, it was a stopgap until we found a way to output the real load. It can be negative because if the injectors are turned off, the latency is subtracted from a zero IPW giving a negative number.

Best thing to do up to load 159 is to use the load logging that is standard. Above that use the timing as long as you have no knock or patch for 2 byte load. There are lots of things going on to develop advanced logging/realtime mapping with refinements to the protocol, so ultimately that will be better still when it is ready.

 

Yeah the more highly modified the car is, the more problems you have with the calculated load. We have done a decent job adapting the load calculation, but we have better tools to work with.

Load Calc is good if you still use a stock rom and UNDERSTAND this value is a guideline, I still use it, but don't "Live by it" since I'm familiar enough with maps and load to know what cells a car might fall into.

As john said, Negative load is possible when IPW is at 0, or the InjPulsewidth column is disabled while logging.

 

Hi,

How accurate is the load calculated really? (I hope it's not accurate)

I was out tuning and didn't look at this parameter in the log since I have a boost gauge. Turns out the peak of load calculated value is a lot higher than I expected..

I was expecting a peak of 1,6-1,65 bar but according to load calculated I was peaking about 300 (which would be 2 bar!!?)

 

Intake and MAF scaling change calculated load from real ECU load drastically. MAF smoothing has no effect but I was able to tune out most of the nontunable AFR map values with the smoothing that otherwise would have been able too

 

I get the same 10.2 (or so) AFR reading when under engine braking on my LC-1. I'm not sure why it is showing a rich reading, but I'm assuming that it just can't display as lean as it really is.

 

Although this has been discussed already, in this and other posts, I wanted to point out that a load value is NOT the same as the boost level, so 300 is not 2 bar.

Anything that alters the MAF input will affect the load value. You are correct, the Load value isn't totally accurate, but its calculated from data that is choppy and not completely a precision value from the ECU. Proper 2byte load (either using the current 2 byte patch, or any new 2 byte patch) will yeild the value used by the ECU instead of the value calculated.

FWIW the values have always been close enough for me to tune a car. As when I tune, and refer to this number (after setting injector latency and scale properly) as a general lookup, and I always work all of the cells around it, and it works very well. Honestly though, I could tune just as well without knowing this as over time, you get a feel for it.

 

Ok, what is "2 byte load"?

I see in the log that there was a very short peak for load calculated to 308!! I guess that this is a bit higher than recommended even if it's almost instantly dropping to 270..

 

What software are you using? The LC-1 and LM-1 switch modes when there is so much O2, and so little fuel, that its reading the free air value %, and NOT AFR. If you calculate the same value in free air mode, you will yield an AFR of 10.5 or something super rich, when in reality, the calculation and mode has changed to read Oxygen% in the free air, which I think is roughly 20% at sea level.

Mitsulogger handles this properly, as does Logworks. Not sure about Evoscan.

 

I get the 10.2 reading in EvoScan. I'll try it in LogWorks II and I bet I'll get the switch you're talking about.