Tuning: Dyno vs Datalog
#1
Out Of Control!!
Thread Starter
Join Date: Sep 2004
Location: San Diego
Posts: 13,394
Likes: 0
Received 4 Likes
on
4 Posts
2015 S212
Tuning: Dyno vs Datalog
Alright, so since this dyno tuning idea has caught on here and we are walking from canned tunes, I'm curious if anyone sees more in using real life data vs generated on a dyno to get the tunes tight?
I know the Gold Standard will be the dyno, since not all cars are prepped to collect data. But, for those of us (and we are obviously growing) that are getting into dataloggers, etc.
I honestly don't see a reason to put my car on the dyno again. I don't want a WOT tune anyways. i want it to be all around and real life, especially intake and cooling. I'm hoping to use a datalogger to track parameters and get tweaks and have that correlate to track times, etc.
Feedback, please. ECU tuning is very new to me so maybe I a missing many points.
I know the Gold Standard will be the dyno, since not all cars are prepped to collect data. But, for those of us (and we are obviously growing) that are getting into dataloggers, etc.
I honestly don't see a reason to put my car on the dyno again. I don't want a WOT tune anyways. i want it to be all around and real life, especially intake and cooling. I'm hoping to use a datalogger to track parameters and get tweaks and have that correlate to track times, etc.
Feedback, please. ECU tuning is very new to me so maybe I a missing many points.
#2
MBWorld Fanatic!
Join Date: Jul 2008
Location: Philly
Posts: 2,171
Likes: 0
Received 3 Likes
on
3 Posts
997TT, R8 V10+, G550, Plaid S
I think data logging and then changing ECU map is the best way to go. Thats what I used to do back in my Honda days. Had a Zeitronix wideband, logged wideband A/F, boost, EGT is an important one, IAT would be another one. I did 3rd gear WOT pulls from 2k to red line, and after all said and done you have a nice line chart of parameters vs RPM. Cool the car down a bit, make some changes in the ECU map and repeat. I miss open source
![Frown](https://mbworld.org/forums/images/smilies/frown.gif)
#4
my datalogger allows me to monitor IATs, Boost, EGT, A/Fs (except I don't have a wideband O2 yet)......in addition, my datalogger can accurately dyno my car.....so i would prefer to tune on the street since it is more accurate with regards to intake temps, boost, etc.....
like suggested, the datalogs recorded on the street could be used in tuning the car at a shop or on a dyno
like suggested, the datalogs recorded on the street could be used in tuning the car at a shop or on a dyno
#5
MBWorld Fanatic!
with a load bearing dyno you can do everything you would do on the street (maybe with the exception of replicating air pressure in front of the s/c) and do it in a fraction of the time. you'd be better off dyno tuning then maybe doing a final tweak on the street if you're paying for someone's time to tune your car.
#6
MBWorld Fanatic!
jangy,
Dyno tuning is as much about convenience as it is about accuracy. The reality is, finding open roads to do multiple WOT runs is difficult for many of us (and would likely land us in jail) so having a nice load-bearing dyno to do this on is the next best thing. As some other members have pointed out "dyno-tuning" for MBs is a complete joke because we have little to no access to onboard sensors (and please, don't get me started on OBDII crap). We are, essentially, tuning in the dark with maybe 1 candle partially lit (wideband O2 in tailpipe)
The real way to do it is having access to the onboard sensors in real-time, logging them, and looking over every finite data point. This notion of dyno-tuning to a tail-pipe O2 sensor does yield results but it's funny how big tuners laugh when I tell them that's how we are tuning our cars.
Someone out there needs to make a CAN-bus or MOST-bus based datalogger that can pull data from the onboard sensors in real time. That, to me, is one of the holy grails for ALL the MB platforms since that is one things that binds them all together - lack of quality datalogging.
If you have any friends in the area with an LSx series engine, or Honda, or Supra, please take a look at some of the awesome datalogging capabilities these platforms have. They are simply amazing... especially the LSx guys. The LSx guys have gotten to the point where the datalogger and tuner/tuning softare are one in the same... essentially the datalogger populates maps and you go in with the maps taken from the datalogger overlayed with the ECU set targets and make changes directly to them. That is how it should be done, but for us is pretty much a pipe dream for now.
-m
Dyno tuning is as much about convenience as it is about accuracy. The reality is, finding open roads to do multiple WOT runs is difficult for many of us (and would likely land us in jail) so having a nice load-bearing dyno to do this on is the next best thing. As some other members have pointed out "dyno-tuning" for MBs is a complete joke because we have little to no access to onboard sensors (and please, don't get me started on OBDII crap). We are, essentially, tuning in the dark with maybe 1 candle partially lit (wideband O2 in tailpipe)
The real way to do it is having access to the onboard sensors in real-time, logging them, and looking over every finite data point. This notion of dyno-tuning to a tail-pipe O2 sensor does yield results but it's funny how big tuners laugh when I tell them that's how we are tuning our cars.
Someone out there needs to make a CAN-bus or MOST-bus based datalogger that can pull data from the onboard sensors in real time. That, to me, is one of the holy grails for ALL the MB platforms since that is one things that binds them all together - lack of quality datalogging.
If you have any friends in the area with an LSx series engine, or Honda, or Supra, please take a look at some of the awesome datalogging capabilities these platforms have. They are simply amazing... especially the LSx guys. The LSx guys have gotten to the point where the datalogger and tuner/tuning softare are one in the same... essentially the datalogger populates maps and you go in with the maps taken from the datalogger overlayed with the ECU set targets and make changes directly to them. That is how it should be done, but for us is pretty much a pipe dream for now.
-m
#7
Someone out there needs to make a CAN-bus or MOST-bus based datalogger that can pull data from the onboard sensors in real time. That, to me, is one of the holy grails for ALL the MB platforms since that is one things that binds them all together - lack of quality datalogging.
Trending Topics
#10
with a load bearing dyno you can do everything you would do on the street (maybe with the exception of replicating air pressure in front of the s/c) and do it in a fraction of the time. you'd be better off dyno tuning then maybe doing a final tweak on the street if you're paying for someone's time to tune your car.
#12
Out Of Control!!
Thread Starter
Join Date: Sep 2004
Location: San Diego
Posts: 13,394
Likes: 0
Received 4 Likes
on
4 Posts
2015 S212
Thanks for all the responses, guys and especially the details M.
I was thinking that I would datalog, so that wouldn't "cost" and I would take my time.
Back when Vadim (and apparently not PC) was tuning my car, I often enjoyed the datalogged tweaks more than the dyno results. I just wanted to make sure that datalogging wouldn't "miss" something that dyno tuning would get.
Even in Alan's case, why not collect the datalogged data (assuming our cars were setup) and share it with PC prior to getting on the dyno?
Anyhow, again thanks for the replies. I am very intrigued on using the factory sensors. Worst case, why not literally splice into them?
I was thinking that I would datalog, so that wouldn't "cost" and I would take my time.
Back when Vadim (and apparently not PC) was tuning my car, I often enjoyed the datalogged tweaks more than the dyno results. I just wanted to make sure that datalogging wouldn't "miss" something that dyno tuning would get.
Even in Alan's case, why not collect the datalogged data (assuming our cars were setup) and share it with PC prior to getting on the dyno?
Anyhow, again thanks for the replies. I am very intrigued on using the factory sensors. Worst case, why not literally splice into them?
#13
Out Of Control!!
Thread Starter
Join Date: Sep 2004
Location: San Diego
Posts: 13,394
Likes: 0
Received 4 Likes
on
4 Posts
2015 S212
So you don't care how much power it makes? Not one bit. It is a number. I care more about how refined the tune is. Vadim pays special attention to that. I never cared about peak HP. I am into driveability and how consistently I can beat which cars.
Usually a person would get a dyno tune and then final tweaks on the road
Usually a person would get a dyno tune and then final tweaks on the road
#14
Out Of Control!!
Thread Starter
Join Date: Sep 2004
Location: San Diego
Posts: 13,394
Likes: 0
Received 4 Likes
on
4 Posts
2015 S212
![Stick Out Tongue](https://mbworld.org/forums/images/smilies/tongue.gif)
I respect the opinion though and would like to sometime get more insight on why "all the way".
#15
MBWorld Fanatic!
The data sampling rates you get from our OBDII ports is nowhere near sufficient to tune from. Someone needs to step up and develop something that can either pull directly from the ECU, or off the can/MOST-bus.
-m
#16
Hey marcus, check out my dashdaq thread. I'm doing exactly what you mentioned. Parts are on order and should be here next week.
#17
MBWorld Fanatic!
Thanks for all the responses, guys and especially the details M.
I was thinking that I would datalog, so that wouldn't "cost" and I would take my time.
Back when Vadim (and apparently not PC) was tuning my car, I often enjoyed the datalogged tweaks more than the dyno results. I just wanted to make sure that datalogging wouldn't "miss" something that dyno tuning would get.
Even in Alan's case, why not collect the datalogged data (assuming our cars were setup) and share it with PC prior to getting on the dyno?
Anyhow, again thanks for the replies. I am very intrigued on using the factory sensors. Worst case, why not literally splice into them?
I was thinking that I would datalog, so that wouldn't "cost" and I would take my time.
Back when Vadim (and apparently not PC) was tuning my car, I often enjoyed the datalogged tweaks more than the dyno results. I just wanted to make sure that datalogging wouldn't "miss" something that dyno tuning would get.
Even in Alan's case, why not collect the datalogged data (assuming our cars were setup) and share it with PC prior to getting on the dyno?
Anyhow, again thanks for the replies. I am very intrigued on using the factory sensors. Worst case, why not literally splice into them?
You are touching on a real thorny issue for tuners here. Tuning should not be some secretive, mail order bull**** marketplace. Access to our cars maps is expensive, but the truth is almost everyone uses the same software. You can buy handheld tuners for LSx series engines for $500, which give you FULL access to the factory ECU, and do your own tuning all day. MB owners are reading that and laughing, cause it sounds impossible. It is not. I still have such a culture shock dealing with MB sometimes it makes me want to rip my hair out. This platform seriously is in such an infancy when it comes to our understanding and capabilities with it, it is downright staggering. We have not even come close to understanding all the nuances when it comes do the different 55k motors, (g55 vs sl55 vs e55 etc) not to mention all the M113 motors and how parts can interchange between them/etc. We also are totally in the dark when it comes to tuning and datalogging.
I could go on and on and on but I think I've made my point. To think that after 7 years of having these cars on the marketplace we are just NOW beginning to put cars on the dyno and tune them, even if that is the most archaic way of doing so, to me is difficult to swallow. It's amazing how much we are willing to put up with for the love of these cars.
-m
#18
MBWorld Fanatic!
jangy,
You are touching on a real thorny issue for tuners here. Tuning should not be some secretive, mail order bull**** marketplace. Access to our cars maps is expensive, but the truth is almost everyone uses the same software. You can buy handheld tuners for LSx series engines for $500, which give you FULL access to the factory ECU, and do your own tuning all day. MB owners are reading that and laughing, cause it sounds impossible. It is not. I still have such a culture shock dealing with MB sometimes it makes me want to rip my hair out. This platform seriously is in such an infancy when it comes to our understanding and capabilities with it, it is downright staggering. We have not even come close to understanding all the nuances when it comes do the different 55k motors, (g55 vs sl55 vs e55 etc) not to mention all the M113 motors and how parts can interchange between them/etc. We also are totally in the dark when it comes to tuning and datalogging.
I could go on and on and on but I think I've made my point. To think that after 7 years of having these cars on the marketplace we are just NOW beginning to put cars on the dyno and tune them, even if that is the most archaic way of doing so, to me is difficult to swallow. It's amazing how much we are willing to put up with for the love of these cars.
-m
You are touching on a real thorny issue for tuners here. Tuning should not be some secretive, mail order bull**** marketplace. Access to our cars maps is expensive, but the truth is almost everyone uses the same software. You can buy handheld tuners for LSx series engines for $500, which give you FULL access to the factory ECU, and do your own tuning all day. MB owners are reading that and laughing, cause it sounds impossible. It is not. I still have such a culture shock dealing with MB sometimes it makes me want to rip my hair out. This platform seriously is in such an infancy when it comes to our understanding and capabilities with it, it is downright staggering. We have not even come close to understanding all the nuances when it comes do the different 55k motors, (g55 vs sl55 vs e55 etc) not to mention all the M113 motors and how parts can interchange between them/etc. We also are totally in the dark when it comes to tuning and datalogging.
I could go on and on and on but I think I've made my point. To think that after 7 years of having these cars on the marketplace we are just NOW beginning to put cars on the dyno and tune them, even if that is the most archaic way of doing so, to me is difficult to swallow. It's amazing how much we are willing to put up with for the love of these cars.
-m
![crazy](https://mbworld.org/forums/images/smilies/crazy.gif)
#19
Super Member
The real "gold standared for true tuning is the street. There are plenty of times the dyno will not show a problem but will be an issue on the road. If your tuning a race car maybe not as much but daily drivers see many varibles that you dont simulate on the dyno. I'm not knocking dyno tuning just saying that having a way to data log on the street is a great saftey meassure.
#20
Former Vendor of MBWorld
Join Date: Jan 2007
Location: Glendale Arizona
Posts: 3,193
Likes: 0
Received 2 Likes
on
2 Posts
C55,SL55,C63
140mph from my car is no problem too. The thing is there is just no safe way to do it. Do you make a pull down the road then stop, wait a while when a new map is made or tweaked? Then wait for the file to upload to the ECU? Its just a total pain in the but.
If something does wrong then what? I've seen it a bunch of times having stop the dyno and check out whats going on under the hood. Or smoke starts coming from some place. It would be hard to see a sensor wire of some sort getting burnt up on the exhaust with the hood down on a public road.
Another reason, If the car starts pulling timing on the dyno the operator can see it right away and abort the run. How do you do that from an on street data log? You either have one guy driving and one watching the computer. Or one guy watching both the computer and the street.
Alot of time people data log runs then check out the data after. What good does that do you if something abnormal already happened? The dyno can see that in real time because your eyes are always fixed on the dyno's monitor. Thats worst case stuff though.
I can see why people like to data log runs at the track. There is a chance you get some variations from true wind speed and hood down rather than up on a dyno. Then send that info back to the tuner for some real world adjustments.
I myself have many data log runs around our Dyno-Comp make shift race track. Most of the time I'm looking for knock and things like that. However any adjustments that need to be made are done on the dyno. In fact its common place at the shop to data log pulls on the dyno. That way we can make adjustment on the spot.
For example we had a car in the shop getting a 91oct custom dyno tune. After data logging the run after the tune knock was detected early in the first pull. The run was aborted on the spot. Turns out the car had Arizona's crappy 87oct summer blend gas in the car and caused it to knock...Point is you could see that from the data logger, but also hear it with the hood open. Might be something you couldn't really hear with the hood down and blazing down the street.
So yeah there is some pros and cons about ways to choose how you tune. Always remember: what do you do if something happens and your on a long road in Mexico?
On another note:
I do my TCU tuning on the road. Thats the only place your going to feel how your car shifts. I did some testing for C55 TCU tuning (basically non existent up until I tested it.) I've been sent down the road with only 2 working gears lol. The loss of comfort mode and gained Winter mode lol. I've be sent down the road with only 2nd gear working lol. After working out the bugs on the street we came to a really good end result for my car. The road proved to be a better testing site than the dyno in this case.
Last edited by hooleyboy; 09-10-2009 at 02:51 AM.
#21
MBWorld Fanatic!
I do datalogging running in 3rd. The values are the same for the load conditions under WOT in 3rd or anything higher and the speeds are more reasonable with traffic all around.
#22
Out Of Control!!
Thread Starter
Join Date: Sep 2004
Location: San Diego
Posts: 13,394
Likes: 0
Received 4 Likes
on
4 Posts
2015 S212
Thanks for the reply, Hooley and I see your point. I was planning to just datalog everyday driving for a few weeks and use that data. I wasn't thinking about making a quick run. But, I see your point that I would need a tune FIRST before I can safely drive around if I have made a hardware change. I will still datalog prior to any hardware changes. Then do the hardware and dyno tune, then datalog again to see how I like things. I think that is basically what you were saying up front.