VCDS Logging section - How to take VCDS logs, what logs to take, and how understand the data
Understanding Logging:
VCDS is probably one of the most under-utilized tools out there. I myself was guilty of it for a long, long time. I could read fault codes and clear them, change some little behaviors with the CCM for locking/unlocking doors, disable seatbelt chime, etc. Yay, cool! Take boost logs but the only thing i really understood was whether it was spiking, overboosting or underboosting, but nothing about truly why or how, and what really to do about it other than the common (not always right) mantras of armchair technicians saying to "lengthen the actuator rod", "check for vac leaks", "hammer mod" etc and often sending people on wild goose chases or throwing wrong parts at it blah blah
So in order to use VCDS as far as understanding the engine, you need to be able to take data logs. The ecu provides a limited way to get data from it in the form of what's called "measuring blocks" - a group of three possible data values which are polled from the ecu at a fixed rate of time.
On the edc15 ecu for example, data can be polled ~3 times/second and includes a timestamp in seconds starting from when the log was begun. on a edc16 ecu (BEW for example), with TURBO mode, data can polled ~10 times/second - a huge improvement and very much satisfactory. If you have a edc16 ecu (all US PD engines), you would ALWAYS want to log in TURBO mode. Edc15 (alh) does NOT have turbo mode. In order to activate turbo mode, you MUST click the TURBO button before clicking the "Log" button. If you don't, you need to back out of measuring blocks and redo it.
It's important to understand how rough the sample rate is in order to take meaningful logs and what sort of driving is conducive to that. With VCDS standard measuring blocks you can record up to 3 measuring blocks in one log. However - only one block can be polled at any given instant, which means for each additional logged block, it takes up a polling spot. Logging 2 blocks will effectively cut your sample rate in half, and then again with 3. If you were to log 011-001-004 together, you would only get 011 (boost) data about once every second, which is almost useless except for steady state highway driving or long 5th gear pulls. Even with logging only one block, you only get data about once every third of a second, which for such a rapidly changing dynamic environment as a turbo diesel engine revving up to possibly 5000rpm, that is very poor and leaves much to be desired. For data that doesn't change rapidly, like coolant or air temps, it's not a big deal - but for boost and n75 data (as well as timing data), it can sometimes be tricky to understand or guess what might be happening between the lines (namely n75).
How to setup a log:
Key on - go to standard measuring blocks. You can start a log even before the car has started, as long as you don't turn the key off, of course. Select the first block to measure, say 011 for boost data (target boost, actual boost, n75 duty cycle). 99% of the time we are going to want to see boost in the log. About the only times you aren't going to want to see boost is if you're looking for something very specific like tracking timing values to see how well the pumps timing mechanism might be working by compared specified vs actual timing. If you select boost data (011) as the first block to measure, this will make things easier if you ever use VC-scope (see below) to replay your log data with graphs since it defaults to graphing the first measured block in the log and 99% of the time that's boost stuff. It also makes it easier for someone who might be looking at logs (*cough*) if 011 is always first for consistency when combing over CSV files. Then select whatever additional blocks you may want logged.
Then click the "Log" button. You will now see a default logfile name on the left, something like "Logs\LOG-01-011-001.CSV". The intial "Logs\" part of the name indicates the logs will be saved in the "C:\Ross-Tech\VCDS\Logs" sub-folder on your computer. If you delete the "Logs\" portion, it will save the file in "C:\Ross-Tech\VCDS" main directory. You can start the filename with a "C:" in which case it will save it in that exact location as you specify.
If a log file of that name already exists, VCDS will append new log data to the existing file!!!!! If you take a log, PLEASE (for your own sake as well) give the log file a unique name that is meaningful, such as "Logs\myname-aug17-workcommute1-011-001.CSV" ; or perhpas you adjusted the turbo actuator, "Logs\myname-aug17-adjustrod18hg-test1-011-001.CSV"
The log has not started until you click the "Start" button. As soon as you click Start, the log file will be created and data is appended to it as long as VCDS is connected. I suggest after clicking Start ,move the mouse cursor away from the button as it now is a "Stop" button and it's easy to accidentally somehow click it and stop the log, not realize it, and then you complete your drive you wanted to log and have zero data. If you do not click STOP to stop the log, you might lose the last few (or more) seconds of data - how much data lost would vary from computer to computer, so it's best to click Stop, especially if you are logging fueling and want to see what hot idle IQ is at the end of the log.
So make sure to click Start!
It's really critical to make sure each log you take has a unique name, else data gets jumbled and it's a real pain to keep track of and anaylze. It soon becomes meaningless because you can't remember the conditions for which the log was taken. Maybe you took a log, then adjusted the actuator rod length, changed IQ adaptation, fixed a boost leak, whatever. Then took another log. For that day, if that's all you care about, that's fine. But more than likely, you still have other problems to fix, or it's not quite fixed yet, or other tweaks to make, or retuning to do and you take more logs. Without keeping track of what's what, you lose a lot of the value the data has, which leads to lost time, wasted fuel, lack of understanding, confusion, miscommunicating if your working with a tuner, etc. Say you're putzing with rod length, if you don't record in the log name what you're doing and take 3 sets of logs, you won't be able to understand the true affect it's having. Sometimes, you won't see it in one session and need to go back to it and revisit it in a week, but now you can't remember which log was which....
An ideal logfile name would include at least the most basic data - like date, some unique identifier, + info about any changes made like this:
"Logs\silvergolf-Nov11-8am-commute-011-001.CSV" or maybe "Logs\silvergolf-Nov11-shortenrod1turn-011-001.CSV"
So by taking an extra 5 seconds to give the logfile a unique name, you can preserve the value of the log data and avoid a whole lot of confusion if you need to return to the log to compare it to other logs you will likely take in the future.... If you care about how well your tdi is running and want to understand the why's and how's, this will be necessary. If it's not easy to do or make sense of the logs, it will be just that much more difficult to progress and one will likely just give up and remain confused or, worse yet, make wrong assumptions and come to incorrect/flawed conclusions.
What logs to take and How to drive for meaningful logs:
This is going to be a very diverse answer depending on what you're looking for so this short section isn't going to list out all the types of logs one can take, but more to describe the thought process involved with logging. So I will ramble...
Generally, most people are only concerned about boost - but even for boost, the standard 011 "3rd gear WOT from 1500-4000rpm pull" is about one of the least informative logs to take. If it looks great, things are most likely good (but not necessarily - it's very possible for a boost log to look perfect but the car perform poorly, be slow, still smoke, etc). Usually the reason why we're taking the log is because something seems amiss. I myself have yet to see a single decently long 011 log that shows what i would call basically decent boost control across the board and no hint of troubles. The results of that one pull does very little to answer any question as to what is wrong - is it fuel, timing, boost control? Could intake air temps be too hot? When air temps are too hot, what is the tune doing about it? Is it "pulling fuel" as many people mistakenly think? Is timing behaving and is specified timing being reached or is timing lagging and retarded from spec? Is the spec timing even "good" for the fuel/boost? Is pump voltage maxxed out? A boost/fuel (011-001) log of your entire commute has far more meaning when it comes to figuring out what may be wrong with your boost or boost control.
First and foremost, you need to know if all of the sensors are working and reading correctly. Boost pressure reading incorrectly is of course bad
Everyone who has a standard MAF tune is always having problems with MAFs reading right - those generally have more tangible and visible side effects. However, if your coolant temp sensor is reading the wrong temp (seen a number of times, sometimes reading flaky for a minute, other times seeming to read normal. or maybe consistently reading low/high), the tune is going to be using the wrong timing maps for the actual conditions - which can greatly affect things. How much so would vary wildly from tune to tune. On some tunes (including factory), depending on the conditions and actual readings, might even show an improvement
but not necessarily across the board. If CTS is reading a bit low, for example, you might gain some efficiency cruising because of some added advance (results may vary if EGR is intact or deleted), but lose some torque/response in the midrange of IQ or higher up because of too much advance... Every error has compounding/cascading affects and rarely is it a simple answer.
A 011-001-004 log will log the most crucial data all at once - boost, fuel, timing. If you have logged intake air temps before (007 block contains fuel, air, and coolant temp) and know generally how well your intercooler setup performs, if you note the ambient temps in the log file name, you can have a pretty good idea of what those would be and apply that knowledge to what else you see in the log. So the sample rate is horrible with 3 blocks measured, but if you're logging a commute and it's mostly highway or at least more steady-state driving with little 1-3 gear pedal mashing, you will still get good data out of it. You'll see how coolant temps track for the type of driving you did (maybe find a flaky CTS or CTS wiring), you can spot check how well the spec/actual timing values match up, see how boost/fuel is affecting coolant temp, generally how well the boost is on target. If you do any WOT pulls, you can see what the timing maxxes out at roughly and if pump voltage is maxxing out and when.
Taking full-drive logs is the most informative - like logging a trip to the grocery store, homedepot or a commute. Start the log before leaving the driveway, and stop it when you get to your destination. You can't have too much data, and it's easy and quick to fast-forward or scroll through all the areas that are "fine and good". Sometimes you won't realize the usefulness of a log until later - perhaps after having made modifications, retuning, or simply miss something. Another reason why it's important to name logs with a unique and semi-meaningful name.
Putting the most amount of data in the log file name that is generally static, or that which can be easily deduced with a good degree of certainty, is a very good way to get better data granularity in your logs. As mentioned, IATs is one such thing. If you've changed timing or IQ adaptation and take a log for something, you don't necessarily need to include 004 (timing) in the log, just tack on "_timing32668" to the log file name to indicate you have taken off -100 for -1.0 deg of advance. Log might be "Logs\silvergolf_air75F_timing32668_nov11-011-015.CSV" for a boost log to track boost + see difference between actual delivered IQ and driver wish IQ. Knowing the air temps is really critical, because tunes can/do change timing based on air temps, and also change the target boost, and of course, air is denser when colder - so all important stuff to be aware of when watching the boost/fuel.
The Marker Button:
VCDS provides a very simplistic way to add "markers" into a log to signal some sort of event. Every time you click the marker button, it puts an ever-increasing number into the marker column at the current time in the log. So the value of the marker button isn't in itself, but what you make it to be (yes, a little philosphical lol). There are a number of ways you can use this to your advantage, especially when logging alone. If you want to use this feature, after start the log, what I do is place the mouse cursor over the marker button, so when you want to mark something you just reach over and click the mouse button and you don't even need to take your eyes off the road.
One example of using the marking button would be in efficiency testing. You're logging 011-015-004 (boost, fueling + consumption in L/hr, and timing) and want to compare efficiency between cruising @ 65mph on your tune (Factory or otherwise) with normal timing vs adding 1.5deg advance using timing adaptation of +150. You would of course need to do the logs on days with little-no wind and similar ambient temps. On these logs, every time you are approaching a flat section of road you will be maintaining 65mph, you click the marker button. If you can't watch the live data and keep it in your head, you now have it in a log to review later. You wouldn't necessarily need to log 004 - you could just log 011-015 and note in the log name the timing adaptation value to get better sample rate, but in a case like this when you're doing steady state driving you don't necessarily need the best sample rate and having the convenience of seeing the actual timing values in the log is might be worth it. You could do the same thing but with comparing different tunes or tune versions, or even across your multiple TDIs - most people have more than one
Another example would be to use the marker button to indicate moderate/heavy smoke. You do a big 5th gear pull, or maybe 4th gear moderate throttle @ 1400rpm going up a hill, you see smoke of a concerning amount, you click the log button right after the fact. Back at home, you can review what conditions were and try to figure out why. Such a log might likely be 011-001-004; or if you know the tune (so you can deduce the fueling), 011-004 if you suspect timing and you want better data granularity.
There are a lot of possibilities and by getting creative you can squeeze out a lot of information/results....
Advanced Measuring Values:
This is just a convenience function of VCDS, however, it can lead to even worse sample rate. I'm not exactly sure how this works, but my guess is the ecu can poll any number of blocks (not just 3, but VCDS must just limit it to 3) and the advanced measuring values simply polls whatever blocks it needs to to get all the data you want. So if the data you want comes from 4 measuring blocks, you're going to get even worse sample rate/data granularity. I never use this, except when using VCDS-mobile on my phone, since that's all that's available, but I'm careful to pick and choose what data I want in order to keep a better sample rate if desired. I usually use my own custom measuring blocks
(see below).
Adaptations:
On the edc15 ALH/AHU/1z, the ecu provides a simplistic way to dynamically alter some of the aspects of the tune by "Adaptations." These are identified by "channel number", VCDS might also have a drop-down menu naming them:
01 - IQ;
02 - idle speed;
03 egr duty cycle;
04 timing (SOI); there is one for glow plug times, i forget what channel. Unless it's registered, VCDS-lite won't allow adaptations to be read or altered. In order to read/write adaptations, you must click the "Login" button in engine module using the code "12233." The default value for adaptations is 32768. Adaptations work by increasing/decreasing the amount of X - eg, increase or decrease "IQ". Under the covers, it is adding/subtracting a flat amount to the corresponding map in the tune. A number >32768 will add that amount (minus 32768) to the map, a number <32768 will subtract that amount. For IQ, the map is the pump map, which maps IQ/rpm to a pump voltage (injection duration). So if you set IQ adaptation to 32268, that will -40 (-.040v) to the whole pump map, meaning injection durations will be decreased across the board by that much. This will have the effect of "raising IQ." If you set Timing (04) adaptation to 32718, that will subtract 50 (.5 deg) of timing advance across the whole SOI base map, thus retarding the entire tune by .5 deg. As VCDS warns, you should not mess with timing adaptation unless you fully understand what you are doing, and what your tune is doing
Using IQ/Timing adaptation is a simplistic but useful tool to work on fine tuning or expirementing/testing without reflashing every time you want to test something. Often times, the flat adaptation will not affect all IQ/rpm areas the same, so one should be aware that changing adaptations like this might improve performance in some areas, but hurt in others, and the results can also vary wildy from tune/tune and setup/setup. One tune/setup might benefit from -50 timing in the mid-range of fuel/rpm, but lose some edge at the high end of IQ. Most tunes are egr-delete, so egr adaptation is useless, and even for a stock car, the effect is extremely minimal.
Custom Measuring Blocks:
The measuring blocks and what they contain are defined in the "tune". One can change them completely from stock, or you can simply add new measuring blocks containing the data you want into unused slots. This is incredibly useful, as the stock measuring blocks leave a lot to be desired if you care about getting more informative logs. There is a lot of redundant information in the stock blocks (like rpm, which can be easily interpolated) and it's not even possible to get all the data you might want with just 3 blocks! As an example, one block i define is 014, which is: IQ, spec timing, intake air temp, and coolant temp - some of the most critical data aside from boost. In order to get that data, you would have to use 4 measuring blocks (011, 004, 007, 001) and sample rate would be absolutely useless for boost other than highway driving. With two blocks, the sample rate is decent enough that it's still good data for most driving. Not ideal, but usable, and a lot more informative than the usual 011-001 log since you rarely are needing to inspect pump voltage or duty cycle values for that, rpm is redundant so all you're getting from that is essentially IQ.
Some paranoid tuners delete measuring blocks by blanking them out so one can't "steal the tune" from log data. lol, that's a good one... but of course now you can't log or view ecu data even on your own car, you would have to rely completely on your own gauges....
.
Using VC-Scope:
VC-scope is an awesome little tool that allows you to watch in real-time (or fast-forward) your log data as pretty graphs. 99% of the time this is just going to be target/actual boost + n75; but with the timestamps, you can then go into the .CSV file using excel and closer examine the numbers. I believe it comes packaged with VCDS, but if not, it's a free download from ross-tech. Easy to use - simply start it up, and open up a log file and click play
If you always use 011 as the first measuring block to log, it will default to displaying the boost graphs,. So please, if you're logging 011, always select that as first block to log.
There's a couple features i wish it had that would make it immensely more helpful:
- being able to scroll through the log. as it is, you can only play it, pause it, or rewind it completely- annoying for long logs
- being able to have multiple graphing windows up at once so you can watch/pause two or more logs at once and scroll through them and compare the graphs visually at the same time
Bottom Line:
Without being able to understand how to take good meaningful logs and make sense of the data, troubleshooting or trying to improve the performance of a TDI is virtually impossible. You are blind. You can spend a lot of time and waste a bunch of $$ throwing parts at things, asking questions online that have been asked and answered (correctly or not) a million times... or log... and log the right data, and if you don't understand what you're looking at, at least you have something you can post up and hopefully get help with and have a starting point to ask other pertinent questions to narrow things down. With the right logs and intelligent thought/questions, most problems can be pinpointed and resolved very quickly with least amount of time and $$ down the drain. And you learn something.
When it comes to figuring out issues on tuned TDIs, the answer is rarely as simple as "adjusting rod length" or "hammer mod", but unfortunately, IMO, most people have been lead to believe otherwise. But this gets into a much more "subjective" topic
It would be in ones best interest to thoroughly log a tune and understand what it's doing and how well it's truly running. This can be difficult if you don't have anything to compare it to other than itself, similar tunes, what you remember a stock TDI to be like, and all the other stuff you read online and take to be true...
.
i think a ported head is very much well worth it. speaking of which, just dropped off a head i ported at the machine shop to be finished off