Where does one find this?

 

Try this. Place in the Universal Data Log Formats directory. This works with EvoScan v.97. There were some very odd decimal definitions in the version that I based this iteration on that I have cleaned up a bit.

 

I was looking at the website and was curious if anyone knows how they go about calculating HP and TRQ.

Anyone use this feature?

 

Open a log file choose tools/powergraph. Then select EVO 5 speed (if that is your evo). It calculates it automatically from the datalog file.


MalibuJack can also elaborate on this feature.

 

Sorry.

Let me clarify.

Would anyone happen to know what calculations they use?

What data to extrapolate these figures.

 

My guess is that they probably just compare statistically to other similar log results and give your car's closest match respectful torque band.

If true then (probably not) plotting WHP is just a calculation of torque and RPM being the only required calculation (torque(RPM)/5250)

 

It calculates power similarly to how a dyno does it. It takes the samples of data, and calculates the power based on the rate of change (in speed or rpm, I forget which) of engine speed over time.

There are some dynamics it takes into account, final drive gearing, the cars mass, frontal area (Aerodynamics since that will affect the numbers at high speeds), and of course, if you use them, the current elevation/humidity/temp to take air mass into account.

Its all based on formulas very widely available..

The big problem in earlier versions of Evoscan (Its probably fixed in the new version, but i have to verify this since I got a sample log last night) was its timestamps were not in a format usable by data log lab, and needed conversion..

So if you get the logs to load, and their based on LOGID, then the sample rate doesnt reflect a consistent rate and therefore delta-time cannot be calculated.. If the time/seconds columns were merged and formatted differently, or (as I did in my earlier conversion tool) calculated the seconds from the start of the log, then the HP calculations work fine if you have adequate data.

Now, I do know that the newest Evoscan deviates from the older DLL definition a bit, my new app works with it, but some of the log header info has changed. I will update and create a new evoscan definition and test it if I can this week.

 

FWIW I have been in contact with the author of DLL who wants to officially support these tools and include the definitions, but I explained to him that prior to .97, everyone using DLL was also using my converter, and will supply him with logs. You guys can do the same if you want to send him .97 log files with all the parameters typically used enabled. (Obviously leave out the 4 evo4 ones since those are for his DTC's)

 

just to clarify for myself, im new to the evoscan program and trying to learn how to use it... im interested in calculating the HP and TRQ as well... following this thread, i have gone to tools and dont see the option for the power graph... do i need to install that file first that someone posted earlier in this thread and then the option will be there ? or is there something im missing.

 

Evoscan itself doesn't have any HP calculations that I'm aware of.. Assuming the HP calcs are fairly simple, I don't see why someone couldn't calculate it on the fly in a custom log entry (I plan on doing it with MitsuLogger, which is my MUT logger app)

In Data Log Lab, there is a power graph feature under tools (Power Graph) where you can plot the current view of your data.. It has some "glitches" but will generally work pretty well if you trim a few samples from the ends of your runs.

Here's the problem, since I haven't had a chance to do any logging with .97 of Evoscan, I do know it doesn't require my Log Analyzer tool any longer to calculate load, but the log analyzer tool also reformatted the time stamps a little bit to make it work with Data Log Lab.. If the new version fixed the timestamps so they are seen as timestamps, then DataLogLab should correctly import the data, if it cannot, you will load correctly using LogID, only that is a sample number value, and Evoscan does not have a consistent sample rate and therefore it can vary significantly. This is what throws off the ability to calculate power in DataLogLab, if you find no matter what you do, you get 50,000whp or some wild number, its probably an issue with timestamps. (you can tell if the numbers in the graph at the bottom axis don't look like times and runs look longer than 15 seconds or so..


Now, with all that said.. I spent a little time fixing up the DataLogLab definition file, and it should work with Evoscan .8, .9 using my Log Analyzer 2.1, It will work with Mitsulogger V1 Beta 3 or later, and it should work with EvoScan .95 and .97 without my log analyzer.. I also fixed a decimal point glitch that the folks at Dataloglab pointed out, which appears to fix all the loading issues you might get if you get a bad load calculation or an abhorrent value. Anyone who has had issues in the past can try this new definition (you should update it anyway) since its no longer based on the original definition file that Laakness put together for me which had some duplicate ID entries and had problems loading some columns.


Do keep in mind that this definition file is a work in progress.. And I'm only specifically focusing on supporting my own tools, such as the Log Analyzer 2.1 for Evoscan, and Mitsulogger, however Evoscan .97 logs are similar enough that they should work right away, or with a few simple changes.

If you find they need to be altered, and you can do it yourself, make sure to post your changes up here and I'll make sure their "Correctly" merged into the definition. Since There are log headers that are named differently, but contain the same data for two different log types, they can be set up so it doesnt matter what kind of log your loading.

 

I just downloaded DLL and the evoscan.txt file.

I then did a "power graph" with DLL.

It seems pretty consistant - although I'm not sure how accurate it is.

Apparantly with my mods I'm putting out ~ 305 awhp and 296 trq.

I've never been to a dyno and haven't run the 1320' yet. I'm going to the track this friday so I'll get a rough estimate how accurate it is.

 

ok...where do i get the dll file from that you downloaded kapolani... and what folder do i put the info in on evoscan ?

 

Is tire pressure and overall circumference accounted for? Even with a the same tire size a 10 psi decrease in pressure can decrease the tire circumference alot giving the car more assumed HP

 

I don't recall.. Tire size is taken into account, but growth is not.. but even with low tire pressure, the tires grow to their correct circumference for the most part so I doubt its as big a factor as we might think.

I do run tires that are 1/2 taller than stock for a reason though..

 

hey malibujack... where am i suppose to put that file ? thanks for your help again