There are several other great Durametric threads around that several members have put together over the years. But it still seems there are a lot of questions or uncertainty about how or what to do that pops up pretty frequently. So this thread is meant to help compile some of these threads and to help some people understand what to do with their logs after they're taken, rather than just send off to their tuner and not understand what really is going on.

Basics:

What is a durametric Cable:
A Durametric Cable is the cable Porsche uses to interface with the cars computer. With this cable you can check & clear fault codes, data log the car, reset the service indicator, view over-revs on a car and much more. For this thread we're interested in the datalogging portion. See their FAQ page for more info: http://www.durametric.com/faq.aspx


Where do I get a durametric Cable:
From the manufacture is your best bet. http://www.durametric.com/buyus.aspx
They sell a "enthusiast kit" which lowest cost vs their professional kit. In it they have an option 1 & option 2.
-Option 1 is the absolute cheapest, but works on a smaller range of cars (Fortunately, our 996's are within this limited group)
-Option 2 allows for it to be used on more Porsches and has an adapter piece to fit older diagnostic ports for slight price increase.

One key factor to note, as soon as the cable is used with a car, it is locked to the VIN of that car. The enthusiast versions of the Durametric cable only allow for each cable to be used with 3 VINs. However, with remaining VINs left, it is a pretty easy item to sell on the forums later to recoup your initial investment since it still has 2 available vins left.


What do I need to run the Durametric Data Logging software and where do I get it?
The software can be downloaded from their website at no cost. However it will need an official Durametric Cable in order to be used. If you're cable was not made in the past couple of years, it may not work with the most current software, but will still be usable with older versions of the software.

The software requires a PC to be run. However it will work in Parrallels and VMware. I use Parallels with a windows server 2008 framework installed to do all my logging and I have not had an issue with Durametric or Ross-Tech's cable for VAG cars.



If anyone would like to send me their logs, I'm happy to edit some down and help you see the differences as time and mods go on. Ideally it'd be nice to get some rough baseline logs from different hardware and tuning setups to compare across the board. Just PM me for more info or help. I'm certainly no guru or wizard, but I may be able to help some of those who haven't done much editing to logs before.



Some other thread Durametric logging information threads to use:
https://www.6speedonline.com/forums/...ic-primer.html <- one of the best and everyone should read through
https://www.6speedonline.com/forums/...se-advise.html
https://www.6speedonline.com/forums/...tric-logs.html
https://www.6speedonline.com/forums/...urametric.html

 

Lets get Started!
Now that you've got the cable, program installed, and in your car, the following will walk you through how to log. FYI, the car has to at least be in the accessory position for Durametrics to work. If you're doing a data log, I prefer to wait to hook up the cable and program until after I'm at my logging location and the car is still running.

Step 1
Plug your cable into the OBD port and your computer's USB Port.


Step 2
When you first open the program, this is what you will see.

Select 996 for our generation Turbo




Step 3
This is what you should now see if it auto-detected your car
(If not, please Click the following)




Step 4
Follow the Clicks below




Step 5 a - Recall Logging Settings
Either use the logging file attached in this thread (or at this dropbox account). Or use your own custom settings.




Step 5 b - Open previous setting
Locate and open



Step 6
Click Stop (it automatically started when you opened "actual values")
Click Clear




Step 7
When you're ready, Click Start. Ideally this would be moments before you're 3rd gear log. It's going to start logging many data points very quickly, so you don't want to wait too long or your log will be buried under a multitude of data.




Step 8
Click "Export" prior to clicking ANYTHING else! Make sure you click "Export" and NOT "Save"




Step 9
Don't forget where you save them. I usually save to desktop for ease of access




Step 10
Make sure to click "Clear" prior to starting your next log

 

Key Factors on How to do the Log

1)
You want a very smooth road with a minimal grade to it. It's best to find a location you can use over and over so you can more accurately compare your logs as you change hardware/tune/gas/etc.

2)
Ideally you want a passenger to do the computer work for you. Either way, you want to setup the parameters for your log prior to getting to your logging location as it may take a minute or two to select everything.
If you don't have a passenger, you can click start much earlier prior to you log and then click stop afterwards and just edit out the extra data.

3)
You want to do a 3rd gear pull, starting from below 2000 RPM to redline (kinda like dyno runs, but on the street). The idea is that when ready at ~1800RPM, you will click start on your log, depress the throttle fully, and take the car to redline without shifting or letting off the gas.

You are NOT wanting to do a 1-2-3-4 gear pull. It provides too small of a window to view and not enough data points. You want one long 3rd gear pull from 2k WOT to 7k WOT. Just like how a Dyno would be used.

It's ideal to shift to another gear (5th or 6th) while you brake to slow down, so that you keep oil pressure up

4)
I prefer to save each run/log as it's own file rather than try to do multiple runs on one file

5)
When you get back home make sure to make a txt file with the Car's condition (amount of gas, type of gas, hardware changes, passenger, etc) Weather (temp/barometric pressure/relative humidity/dew-point/elevation) and I like to not my roads beginning elevation and ending elevation (can be found via google maps. To simplify the weather, I simply calculate the DA (density altitude) and put it with the run as well as the temp.

A month from now, much less a year from now, it'll be much more useful to compare when you look back if you save this data with it

6)
I prefer to copy all the individual spreadsheets into one workbook with individual worksheets. This makes it easier to share and reference later.
Then make one folder with the date or reference and put all the files into one folder. To clean things up even more, I compress the multiple data logs into one zipped file inside the folder and delete the individual files. This way you end up with a folder that has: combined log, conditions, zipped individual logs



Quick Tips

-F6 is the start/stop toggle button. This can be used instead of clicking the button which can help multitask. However, be careful not to stop a log when you think you are in fact starting it.

-Shutting off the "Graph Function" *and selecting "digital display only" can help reduce CPU workload, help battery life, and help overall log accuracy.

-You can also adjust the refresh rate to create more, or less, data points as well.

-If you're having issues with the data logging values, try turning off your antivirus as it may be creating a conflicting issue.




How do I know what parameters to Log?

Well, here are some suggested parameter to use.

Majority of these suggested parameters for each logging group was suggested by Todd at Protomotive for my initial tuning. However, I've seen similar requests by Markski and other tuners on 6speed.

Some basic suggested parameters are:
> RPM
> LOAD
> IGNITION ANGLE
> OXYGEN SENSING BANK 1
> ADAPTION RANGE 2 FRA BANK 1
> OXYGEN SENSING ACTUAL LAMBDA VALUE BANK 1
> OXYGEN SENSING BANK 2
> ADAPTION RANGE 2 FRA BANK 2
> OXYGEN SENSING ACTUAL LAMBDA VALUE BANK 2
> BOOST PRESSURE OF SENSOR

To look at the Injector Duty Cycle Add: INJECTOR TIME
To look at the IC cooling efficiency Add: IAT
To look at the MAF readings, Add: Hot Film MAF, and/or Mass Air Flow

To look at Boost Control or Knock, use the following:
> RPM
> Engine load
> Ignition angle
> Mass air flow
> Actual value throttle
> Oxygen sensing bank 1 Lambda Value
> Lambda setpoint B1
> Boost pressure of sensor
> Setpoint boost pressure
> Boost pressure control P/D factor
> Corr. factor for BPC with charge air temp.
> Corr. factor for BPC with knock control

Using the latter set of parameters, we can use these values to see the duty cycle of the N75 valve, the boost setpoints, etc. This will tell you what the boost control is doing through the RPM range.


The preset version of each of these can be downloaded via dropbox, or as an attachment to the thread.
Here is the dropbox link for preset logging recall file

 

Okay, I did all the work, I have piles of Data,
but what do I do with it and how do I understand it?

Alright, here's where the computer editing / geek side comes out. I've made a spreadsheet through excel that you can use to view the formulas and what each parameter is used for, or how to convert it to a more usable measurement.

First, you'll most likely notice a small green triangle in the corner portion of the box for a majority of the cells. You can highlight all of these and then click the small error box that pop up and select "convert to number". This is where the datapoints of the .csv file were saved as txt, but excel recognized them as numbers, but needs your confirmation before using them as such.

The formulas spreadsheet can be downloaded at the same dropbox link or via the thread attachment.

Here is a screenshot of the First part of the formulas I've put together:


Here is a screenshot of the Second part of the formulas I've put together:





After all the editing, here is a log of my car in stock form:




Here is a another after editing. This was after my initial base flash:

 

FAQ

How many data points can I log at once?

This has been debated some. I've found with the newest version Durametric program that it can log up to ~10 points at once decently well. (this is also somewhat computer dependent as well) The less points used, the more accurate most likely, however, you want to see how the different points interacting together. Logging only one or two parameters at a time isn't an accurate example if used to compare across logs as they didn't happen at the same time.



But I want to do a 1-2-3-4 gear pull, just like I would use on a drag strip or at a race track.

Cool. . . good for you. You've now got a thousand data points that are basically useless to compare / evaluate.

We're wanting to evaluate tuning and conditions of the car, not look at the shift points and acceleration from 1st to 5th. One long 3rd gear pull is best. (again, just like a dyno would be used) Using multiple gears brings in driver error, additional vibration from shifts, and unneeded variables. We want to see the car at full throttle STARTING at 1800-2000 and continue full throttle in the one gear all the way to redline.



What gear to use? 3rd vs 4th gear. Dynos usually use 4th

We use 3rd gear as you can get almost equally valuable reading in 3rd with a much greater ease of attaining the log due to speeds required, distance covered, and quality of road needed. It is true, that we are basically wanting to look at the motor under load, and 4th gear offers a greater span of data with load on the engine and greater load sooner, but the comparable numbers are equal enough that there is little reason to use 4th. Therefor 3rd gear is more than fine.

Also, dynos typically use 4th gear as it is typically closer to a 1:1 ratio for converting power output to hp, which isn't the direct goal of datalogging. Starting to compare to dynos opens up a whole box of issues, so lets ignore the idea of dynos vs datalogging.



Timing, and comparing timing from one car to the next:
Timing Advance. Timing is logged via "Ignition Angle". It is also referred to as being "Advanced" or "retarded" depending on the shift.

For general knowledge to help understand how timing is used in our car and how it is used, the following is a brief explanation: The timing is inversely proportional to the load. So, as the load drops off dramatically due to the cams not retarding on a 996tt, the timing can go up to compensate and help the power out.



Boost Set Point vs Boost Pressure

Boost set point is the ecu requested boost pressure. Whereas the boost pressure the sensor's reading.

Also, the stock boost gauge on the dash maxes out at 1.54 bar.



I want to log my MAF. What can I use?

Mass Air Flow (HFM)
Mass Air Flow (KG/H)
Hot Film MAF

The first two are a measurement of volume of air, the third is the voltage the MAF measures. The little know fact is that when the stock maf hits 5v, it is pegged to it's maximum reading. This occurs at varying power levels, but I've see this is right near the maximum power of a flashed car with exhaust, and on a k16hybrid car is pegged out by 4500 rpm. The options are switching to a MAF that can read higher and tune for such, tune for a mafless car, or use a tune that utilizes the maf up to that point, and then acts as a semi-mafless car beyond it's reading capacity. See this thread for more information https://www.6speedonline.com/forums/...urametric.html



What about AFR
The quick and simply calculation typically used is 14.7 x O2 Lambda

However, this is a very interesting and valid point



What is IAT

IAT is basically a look at the efficiency of the intercoolers. The more air is heated from the initial temperature readings, the less efficient the IC are working. An increase in pressure causes an increase in temperature, but this is where we use intercoolers to try and combat this temperature increase because hotter air means less power.

It's been mentioned that after 50 deg C the ECU will begin to pull timing to compensate for the poor air conditions and protect the motor. This is an important safeguard and shows the direct effect of quality IC.


Knock - The culprit, result, and measurement?

Also known as pinging, detonation, knocking and a few other names. It often is pretty distinctly audible and noticeable. It is typically more common in higher RPM's in taller gears where the car is under full load and at higher pressure/RPM's/load/etc

The car is always attempting to detect knock, but when it does, we see the resulting effect of the ECU attempting to pull things back (timing/boost/enriching AFR) to create a safer environment, not the direct values from the cars sensors. There isn't a direct formula to show it from values we obtain through data logging (i.e. you can't combine X + Y to get Z), and I don't have any samples logs at this time to show it. However, the key factor to remember, knock is usually the end result of other parts not performing as they should.

Knock is a fairly common culprit term used these days, but it's the end result typically, not the start. There usually are other signs of parts/data-values which are not ideal prior to knock occurring. And it still seems the audible sound is one of the best ways to detect knock as it is difficult to show the exact knock point with measuring sensors.

Without any tuning changes, we can try to diagnose/improve it by trying to change the fuel (try some 100 octane) usually helps to see if things improve in the right direction as they should with the boosted octane (no retune needed, just wanting to see if the quality of fuel improves the cars desire to run smoothly). General maintenance such as spark plugs etc is also a good direction to go to help the car run better and see if it improves the issue. Also, if you're running a EBC/MBC lowering the boost to see if it's occurring at a certain threshold could be beneficial. Also the efficiency of IC and the temperature of the air entering the motor is something to look at.

If the issue still occurs, tuning might be the next step by reducing timing, adding fuel to increase the AFR, decreasing boost, partially closing the throttle plate, adding a methanol system, and other techniques.

Knock in its simplest form is often the result of very high cylinder temps. So it's something not seen as often in colder days/climates and as such. Knock is not to be confused with pre-ignition. They can share some of the same culprits, but they are not the same thing.



What is Engine Load and Volumetric Efficiency


For more information about the following, please see the following link as it goes into a great detail about a lot of good topics:
https://www.6speedonline.com/forums/...ic-primer.html

Thanks to 6speed members who posted valuable information and discussion which I referenced in this thread:
bbywu
Markski
FAST FWD
And many others

Also thanks to Todd at Protomotive as he helped walk me through the initial parameters of what to log and general customer support with data logging.


If anyone has any suggestions, or notices anything incorrect, please let me know as I don't want to misinform anyone and will gladly make changes or additions.

 

If anyone would like to send me their logs, I'm happy to edit some down and help you see the differences as time and mods go on.

Ideally I think it'd be nice to get some rough baseline logs from different hardware and tuning setups to compare across the board. Just PM me for more info or help. I'm certainly no guru or wizard, but I may be able to help some of those who haven't done much editing to logs before.

 

Reserved for future use

 

GREAT thread my friend, rep points sent your way.

 

I'm sticky'ing this.

 

amazing thread. learned more in the last 5 minutes than i did the entire time i have been using the durametric. lunch on me and rep to you!

 

Best Thread ever!!

Thanks this is great info, thanks for doing this.

 

Rep for you! What a wealth of information, thanks so much for posting.

 

Hi,

many thanks for your work. Although I already spent some time with this logging stuff, I learned a thing a two

I can confirm the above. Having the graph running slows down the sample rate. I am actually logging 18 values with a resolution of 10 samples per second. More should be possible with a faster notebook (currently 2.13 GHz Core2Duo).

Having an SSD harddisk is quite helpful. Durametric didn't spent time on perfomance of that logging stuff. You have to tweak the system in order to get what you want.

What I learned:

- adding RPM/s to my graphs. Very helpful information

- Logging knock:

That's what I missed and I wonder what these values will show if knock occurs. Do you have samples when that happens and probably explanations of the values and what they really represent?

In understand that these are not the knock sensors, but a reaction on knock, right?

Andreas

 

One more thing: I strongly suggest disabling your virus scanner during that logging. Mine had a serious impact on the sample rate (like only 50% compared to without virus scanner)

 

Great thread. Subscribed!Thank you for all the screen shots, cartoon bubbles for sequences and tips.