JohnBradley

iTrader: (30)

Okay, so after talking about this alot in the other thread and a fair amount of testing the beta version of the xml is ready to be made public. First I am gonna say that I have purposely induced knock and the airflow (which is power) drops off in direct proportion to the amount of timing being pulled.

THIS IS ONLY A TOOL FOR SHOWING CHANGE it is not 100% accurate and relies on personal experience with watching the logs closely to understand when peak power has been "achieved". By releasing this as a beta I hope as many people as possible can take a look, test and post improvements. There arent really any questions to be answered as it will be laid out as well as possible. Please use PM's for questions that arent specifically covered and try to use this thread as a datasource instead.

Requirements-

You must be able to log 2 byte load and 2 byte airflow in order to use this. Datalogging with only the bare essentials, log to screen off, and logging superspeed will make it more accurate.

here is the xml for Evoscan for those-

<DataListItem DataLog="Y" Color="" Display="2 Byte Load" LogReference="Load" RequestID="00" RequestID2="01" Eval="0.3125x" Unit="units" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="500" ChartMin="0" ChartMax="500" ScalingFactor="1" Notes=""/>

<DataListItem DataLog="N" Color="" Display="2 Byte Airflow" LogReference="Hz" RequestID="02" RequestID2="03" Eval="6.29x/64" Unit="Hz" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="3200" ChartMin="0" ChartMax="3200" ScalingFactor="0.1" Notes=""/>


Here is the xml for logging g/rev, Lbs/min, WHP and Torque. The data is presented as it would be on a Mustang.

<DataListItem DataLog="Y" Color="" Display="g/rev" LogReference="gRev" RequestID="00" Eval="[Load]/95" Unit="g/Rev" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="12" ChartMin="0" ChartMax="12" ScalingFactor="1" Notes=""/>

<DataListItem DataLog="Y" Color="" Display="Lbs/min" LogReference="LbsMin" RequestID="00" Eval="([gRev]*[rpm])/454" Unit="lbs" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="65" ChartMin="0" ChartMax="65" ScalingFactor="1" Notes=""/>

<DataListItem DataLog="Y" Color="" Display="HP" LogReference="WHP" RequestID="00" Eval="[LbsMin]*9.15" Unit="whp" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="1000" ChartMin="0" ChartMax="1000" ScalingFactor="1" Notes=""/>

<DataListItem DataLog="Y" Color="" Display="Torque" LogReference="Lbsft" RequestID="00" Eval="([WHP]*5252/[rpm])" Unit="lbsft" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="0" GaugeMax="1000" ChartMin="0" ChartMax="1000" ScalingFactor="1" Notes=""/>

Now here is where you need some skill, and thats interpreting the data honestly and correctly. The point at where peak power occurs is right when 2 byte airflow starts to drop off and gain slowly. It will look something like this for instance:

1516.184844
1525.521563
1740.266094
1852.896406
1898.204063
1863.609063
1862.822813
1919.531094
1949.9
1973.4875
1983.610469
2011.129219---this is point where the 20-30hz jumps STOP (happens to be 6625)
2004.347813
1989.310781
2013.389688
2021.350469
2029.999219
2032.259688

This happens to be my car this morning in 55-60* temps. Later in the log on he next shift though the hz has dropped some the plateau happens at the same rpm more or less (6718) so to me this shows power is starting to drop off. Not scientific but helpful for quantifying changes. I have a few logs showing that more boost at higher rpm will stave off the plateau in hz (meaning I would theorize that the power is at the higher rpm) until about 7000. That with a Green and S1's.

I would like to give a big thanks to Eric (L2r99gst), Charlie (Kreionic), Jason (Bigric09), and Jeff (Jeff_Jeske) for their help in bouncing around ideas and logging to get some idea of what was going on. Kreionic (Charlie) has a log of his 351/351 run on a Mustang and with the adapted numbers and proper review of the hz, the car made peak power at exactly where hz starts to drop off (and boost for that matter). I had used a dynojet esque correction of 10 and when I changed it 9.15 I got numbers that matched the actual dynosheet almost exactly. The torque was 4 ft lbs lower by this method but Hp was the same based on airflow.

JohnBradley

iTrader: (30)

http://spreadsheets.google.com/ccc?k...f4-QQ&hl=en_US

http://spreadsheets.google.com/ccc?k...RRNQAwRQ&hl=en

Some logs from earlier today showing it in operation, the effect of knock on airflow and hp, etc.

recompile

iTrader: (38)

Very nice JB! Great work.

Here's my results, showing the same log viewed in DLL and in JB's formula.

If anyone wants to review the log, it is here: Shameless-3rdGear-10-05-07.csv

recompile

iTrader: (38)

I have a queston though. You say that you must be able to log 2 Byte Airflow to use this, however your calculations for g/rev, WHP, and LB-FT do not use Hz, only Load, so how is Hz used?

Also, you're saying this is Mustang Dyno-type data. So I'm pretty happy with my result if it is correct.

Here is the same exact log, with a scale of 10, as you stated is similar to a DynoJet:

JohnBradley

iTrader: (30)

The reason I use 2 byte airflow is only to see where the airflow plateaus. You are correct that its not used in the actual calculations. Its still not 100% refined, but as you can see if nothing else it gives you an idea of how close you are to maxing your turbo at any given rpm.

For instance in your log (the DJ one) I can look and see that you are moving around 37#'s of a total 42 possible at around 7200. In my logs you'll see that right now my car is only moving around 40 of 49.

JohnBradley

iTrader: (30)

Looking at your log in detail it looks like power actually starts to drop off at 6500 (333hp) and torque is at 3375 (318 lb ft.) That would actually even be a little closer to your mods. I suggest that you do the 2 byte airflow mod and when you are logging unclipped airflow your power peak (regardless of airflow continuing to increase) will be a little easier to spot.

If I knew what I was doing, writing an excel script to compile the whp, torque, against airflow would probably make it easier to spot.

roger smith

iTrader: (4)

This is really cool stuff. Thanks for posting.

I don't know though. According to this my car made 300 hp at 6000 rpm. Load of 235. evo 8s don't make that much power with just a turbo-back and boost turned up 19psi).

Do mustang dynos read high?

JohnBradley

iTrader: (30)

Mustangs read about 15-20% lower or more than a dynojet. Can you post a log or portion of it? I would like to look and see if I need to make a revision here or there and see whats going on. I am thinking that since you have ECU+ you are already logging total hz and not limited by the clip?

235/95=2.47368 g/rev * 6000/454= 32.7 lbs/min * 9.15 = 299whp

Ever drag raced and seen what the car will do with your present tune and mods?

l2r99gst

iTrader: (2)

You guys have to understand that this is just estimating HP from airflow. It's not reallt meant for absolute numbers. One airflow can produce several different HP from one tune to the next, depening on AFR, timing, etc.

This is just yet another approximation of HP. Just another tool to help out.


Eric

roger smith

iTrader: (4)

Here's the one I did the calculations I spoke of. It's a csv renamed to txt.
Sorry I don't have airflow in there. My ECU+ log says airflow is 1640 at 6000
At 3468 my airflow is approx 945 and boost about 16 (not at full boost yet).

I'll make some better logs in ECU+ this weekend as it can log 2-byte load now.

And no, I havent' dragged it. I'm still in the process of tuning.

roger smith

iTrader: (4)

One thing I was thinking, won't adding these things in slow down the logging? These calculations can be done afterwards for those that don't have fast computers.

EvoBroMA

wow thats crazy. using airflow to estimate whp. My problem with this, is it's very car specific.

i would rather use rpm/time and weight, personally - because it applies to more cars than just the evo.



sorry i missed the point.

The point is: it's useful for measuring the differences in power - while your tuning.

thanks JB!!!

i like how its just a few xml definitions, and doesnt require mad calculations. very nice.

kreionic

iTrader: (14)

JB make it a point to tell people even if there using 2byte load already to use the one you posted. Because of the calculation g.rev uses. Its taking the name "[Load]" for the calculation and the one in evoscan already is is named "Load2Byte". It will not log unless you rename 2 byte load to "Load" or use the one you posted

JohnBradley

iTrader: (30)

Eric thank you for making sure to point that out. THESE ARE NOT CONCRETE NUMBERS. Time/weight/accel is the only true way to test hp without a dyno. AGAIN THIS IS ONLY A TOOL, I dont want to be accused of selling snake oil or making false hopes to anyone.

Timing directly affects the amount of airflow through the engine from what I have logged and linked. The rest of what Eric pointed out is very valid as there are TOO MANY other variables that can affect the reading (such as a poorly scaled MAF/filter combo). It is quite possible to be reading false values if the load is not accurate to begin with.

travman

iTrader: (26)

does this work with 2 byte rpm?