2004 to 2020 Mazda 3 Forum and Mazdaspeed 3 Forums banner

1 - 20 of 21 Posts

·
Registered
Joined
·
246 Posts
Discussion Starter · #1 ·
I found a site, linked below, that seems to give a lot of good info about cars. I was wondering if anyone could confirm the gearing as accurate though? The reason I ask is twofold.

First, I wanted to make sure that I had entered accurate specs into Vdyno as I had to create the vehicle from scratch (they dont have the new Mazda 3 entered already). I have been getting some low power readings on my first few files and want to make sure that I am not causing an error.

Second, if it is accurate than we have a good source for vehicle details.

As a side question- I am on my third tune file from OVTune and seem to be maybe as powerful as untuned. The torque curve is flatter but I dont feel any additional power and Vdyno seems to show the same or less power than stock. Was this the same for everyone else? I was hoping by file 3 to see a difference. Did anyone else get a drop in power early on?

2015 Mazda MAZDA3 vs 2015 Toyota Corolla - The Car Connection
 

Attachments

  • Like
Reactions: M3Dude

·
Registered
Joined
·
606 Posts
The gear ratios seem about right for a 3rd gen 2.0 manual US spec. I remember using a little more for frontal area; around 24 ft sq. Weight seems a little too low, as it should be the cars weight (~2865 lbs for the sedan) + weight of passenger.

Before you compare Vdyno's, consider that they are less accurate unless you are recording WOT on a flat road (and the exact same stretch of it), temperature is roughly the same between days, there is hardly head/tail wind, and you went WOT at about the same rpm. All these factors skew the curves and the power & torque produced. Intake temperature is particularly huge in determining peak tq/hp, but the software does not compensate very well. For example, you loose around 10 whp going from 0 oC (32 oF) to 30 oC (86 oF) intake temperatures, and it is getting warmer out there... Another big factor is the gear they are done in... 2nd gear is too sloppy to get consistent results on the Vdyno, as you will find large variations on the exact same tune file. 3rd gear makes them more realistic, and useful for comparisons.
 
  • Like
Reactions: M3Dude

·
Registered
Joined
·
246 Posts
Discussion Starter · #3 ·
The gear ratios seem about right for a 3rd gen 2.0 manual US spec. I remember using a little more for frontal area; around 24 ft sq. Weight seems a little too low, as it should be the cars weight (~2865 lbs for the sedan) + weight of passenger.

Before you compare Vdyno's, consider that they are less accurate unless you are recording WOT on a flat road (and the exact same stretch of it), temperature is roughly the same between days, there is hardly head/tail wind, and you went WOT at about the same rpm. All these factors skew the curves and the power & torque produced. Intake temperature is particularly huge in determining peak tq/hp, but the software does not compensate very well. For example, you loose around 10 whp going from 0 oC (32 oF) to 30 oC (86 oF) intake temperatures, and it is getting warmer out there... Another big factor is the gear they are done in... 2nd gear is too sloppy to get consistent results on the Vdyno, as you will find large variations on the exact same tune file. 3rd gear makes them more realistic, and useful for comparisons.
Thank you for the assist. I would write off my power loss as environmental changes, however it is to the tune of 20hp. I just got file 4 so I hope to see them return.

Stock was about 155hp and torque.
Tune 1 was about 143hp and torque.
Tune 2 was about 153hp and torque.
Tune 3 was about 135hp and 140 torque.

I am sure I will get the power I have seen from others but was trying to see if my situation was unique. Could VDyno create that much disparity?
 

·
Registered
Joined
·
188 Posts
Thank you for the assist. I would write off my power loss as environmental changes, however it is to the tune of 20hp. I just got file 4 so I hope to see them return.

Stock was about 155hp and torque.
Tune 1 was about 143hp and torque.
Tune 2 was about 153hp and torque.
Tune 3 was about 135hp and 140 torque.

I am sure I will get the power I have seen from others but was trying to see if my situation was unique. Could VDyno create that much disparity?
How do you lose HP on tunes? :surprise:
 

·
Registered
Joined
·
606 Posts
Thank you for the assist. I would write off my power loss as environmental changes, however it is to the tune of 20hp. I just got file 4 so I hope to see them return.

Stock was about 155hp and torque.
Tune 1 was about 143hp and torque.
Tune 2 was about 153hp and torque.
Tune 3 was about 135hp and 140 torque.

I am sure I will get the power I have seen from others but was trying to see if my situation was unique. Could VDyno create that much disparity?
Are you recording in the exact same location? Is it a flat road? What are the intake temps in Tune 1 vs. Tune 3 according to the log file? 3rd gear only? The combination of a number of these factors can skew the results by 20 hp. If none of those are an issue and you are still low in power by Tune 4, there may be other factors involved that we can discuss. I can already tell that 155 whp for the stock tune is way too high, implying that something is off with the quality of that log or the software settings. I would expect around 140 hp (like tune 1) on the Vdyno with some bolt-ons in moderate temperatures.
 

·
Registered
Joined
·
246 Posts
Discussion Starter · #6 ·
How do you lose HP on tunes? :surprise:
I dont think I am. Im just trying to get more accurate readings so that I can post a before and after VDyno and real world dyno. Matt is not done with me tune yet and there are a lot of things to sort out that could cause fluctuations. Since I am new to running the dyno software I wanted to see how much of this was variation from my errors or environmental issues vs minor power losses during the tuning process. I do believe that I will see a solid gain in the end.
 

·
Registered
Joined
·
246 Posts
Discussion Starter · #7 ·
Are you recording in the exact same location? Is it a flat road? What are the intake temps in Tune 1 vs. Tune 3 according to the log file? 3rd gear only? The combination of a number of these factors can skew the results by 20 hp. If none of those are an issue and you are still low in power by Tune 4, there may be other factors involved that we can discuss. I can already tell that 155 whp for the stock tune is way too high, implying that something is off with the quality of that log or the software settings. I would expect around 140 hp (like tune 1) on the Vdyno with some bolt-ons in moderate temperatures.
I am going to try a new location and see if it helps. I was using a canyon road, since it is isolated, but there could be a grade that I dont notice. I will try a different stretch near home and rerun stock pulls as well.

Thank you again for the info. I assumed that there could be some fluctuation from temp and user error but 20hp seemed a bit high. Then again, that is why I wanted to verify the stats I entered for the car as well. Small changes can have large results.
 

·
Registered
Joined
·
188 Posts
Could a tune not done correctly cause heavy gas usage?

New to this whole process. Never thought about how you do logs could cause issues and all.
 

·
Registered
Joined
·
246 Posts
Discussion Starter · #9 ·
Well, file 4 is loaded and "feels" much better. I will drive the requisite 60+ miles and then do some pulls tomorrow. Fingers crossed.


On the plus side, it looks like that website is useful. It has weights, gearing, drag, etc.
 

·
Registered
Joined
·
246 Posts
Discussion Starter · #10 ·
Could a tune not done correctly cause heavy gas usage?

New to this whole process. Never thought about how you do logs could cause issues and all.
A bad, or even just an "off" tune could have all kinds of consequences. The reason we all look to Matt and OVTune is that there had been issues with really bad tunes before by other companies.

I knew that changes in temp, altitude, grade, etc. could effect hp ratings but wasnt sure it could be this great. I have always taken my car to get dyno'ed before. This is my first time doing it myself. I want to get good at it because I find it fascinating.
 

·
Registered
Joined
·
606 Posts
I am going to try a new location and see if it helps. I was using a canyon road, since it is isolated, but there could be a grade that I dont notice. I will try a different stretch near home and rerun stock pulls as well.

Thank you again for the info. I assumed that there could be some fluctuation from temp and user error but 20hp seemed a bit high. Then again, that is why I wanted to verify the stats I entered for the car as well. Small changes can have large results.
Your welcome. Let us know how the new location turns out.

Just to give everyone an idea of what hills can do, here is a fairly small down-hill that came up near the end of this pull. You can see how the torque and hp curve is shifted upwards, as if the car is pulling harder in the top-end:

Hill near end.jpg

Now from an OV tuned dyno of the 3rd gen Mazda 3 2.0 tuned, you can see torque falls off hard in the top-end, typically after 5500 rpm. This was confirmed in other skyactiv 2.0 dynos as well. The cams just don't allow the car to breathe enough in order to produce that much torque in the top-end.

Final Tuned- run 11.jpg


So my point is that if your Vdyno graph has uncharacteristic curves in it, they are likely caused by factors outside the tune (eg. hills, etc.). Another example I don't have time to show right now is comparing 2nd vs. 3rd gear Vdynos in the same location. They can vary by 5-10 whp/tq on the same tune file...
 

·
Registered
Joined
·
606 Posts
Could a tune not done correctly cause heavy gas usage?

New to this whole process. Never thought about how you do logs could cause issues and all.
I'm assuming you mean that the logs are not done correctly causing the tuner to not dial-in the settings ideally rather than a tune from a bad tuner which is a whole other story hehe. For the former, it is possible to have worse fuel economy mainly through badly calibrated fuel trims. If you basically flash, then do a log 30s later and not allow the fuel trims to settle, you could be getting ultra lean unsettled fuel trims in the logs. So if your car showed +15 LTFTs (very lean) in certain fuel economy sensitive areas at the time of the logs, but then it settles to 0% after 60 miles, the tuner will add 15% more fuel and you will then be 15% richer after you drive the next tune file over 60 miles. Although the ECU corrects most of the excess fuel (feedback from O2 sensor), you will still be burning more gas over the long run...
 
  • Like
Reactions: M3Dude

·
Registered
Joined
·
188 Posts
I'm assuming you mean that the logs are not done correctly causing the tuner to not dial-in the settings ideally rather than a tune from a bad tuner which is a whole other story hehe. For the former, it is possible to have worse fuel economy mainly through badly calibrated fuel trims. If you basically flash, then do a log 30s later and not allow the fuel trims to settle, you could be getting ultra lean unsettled fuel trims in the logs. So if your car showed +15 LTFTs (very lean) in certain fuel economy sensitive areas at the time of the logs, but then it settles to 0% after 60 miles, the tuner will add 15% more fuel and you will then be 15% richer after you drive the next tune file over 60 miles. Although the ECU corrects most of the excess fuel (feedback from O2 sensor), you will still be burning more gas over the long run...
So ideally, you really need to drive 60 miles straight then pull logs?

Ohhh another question!

When doing 10-20 minute "town" logs. Is there an ideal setup/format/run/ect.... to do logs that help the tuner the best? Should you cruise control, do lots of stop and gos or?????
 

·
Registered
Joined
·
246 Posts
Discussion Starter · #14 ·
So ideally, you really need to drive 60 miles straight then pull logs?

Ohhh another question!

When doing 10-20 minute "town" logs. Is there an ideal setup/format/run/ect.... to do logs that help the tuner the best? Should you cruise control, do lots of stop and gos or?????
The guideline is to drive 60 miles or so after a file is loaded or the ECU is reset so that the computer can relearn the fuel trim based on the new parameters. After that "break in" period it is okay to log your data.

I believe that varying conditions are best for the "around town' log. It allows the tuner to see a wide variety of conditions. Stop and go. Uphill and downhill.
 

·
Registered
Joined
·
606 Posts
So ideally, you really need to drive 60 miles straight then pull logs?

Ohhh another question!

When doing 10-20 minute "town" logs. Is there an ideal setup/format/run/ect.... to do logs that help the tuner the best? Should you cruise control, do lots of stop and gos or?????
Yes, at least 60 miles will help. Better not to rush the logs as they affect what you can get out of a tune.

You want to expose the car to a wide range of engine loads and rpms while doing the around town logs. You want to have a lot of stop-and-go's rather than prolonged cruising. Cruise control doesn't help achieve this much.
 
  • Like
Reactions: M3Dude

·
Registered
Joined
·
17 Posts
@skyactiver
I see you have a ton of experience and good suggestions on that :)

May I ask you what are your settings in the MazdaEdit Vdyno section at:

Left side setting
1) "Log Parameters" --> "Use RPM as possible" --> checked or unchecked? (what does it actually mean?)
2) "Dyno Correction"
3) "Environment" --> "SAE" --> checked or unchecked? (what does it actually mean?)

Bottom of the chart
4) "Smoothed" --> checked or unchecked? (what does it actually mean?)
5) v1, v2, v3
6) Factor

I guess after this I will have my most perfect settings done and can go for best comparison logs.

Thank you in advance :)
 

·
Registered
Joined
·
606 Posts
@skyactiver
I see you have a ton of experience and good suggestions on that :)

May I ask you what are your settings in the MazdaEdit Vdyno section at:

Left side setting
1) "Log Parameters" --> "Use RPM as possible" --> checked or unchecked? (what does it actually mean?)
2) "Dyno Correction"
3) "Environment" --> "SAE" --> checked or unchecked? (what does it actually mean?)

Bottom of the chart
4) "Smoothed" --> checked or unchecked? (what does it actually mean?)
5) v1, v2, v3
6) Factor

I guess after this I will have my most perfect settings done and can go for best comparison logs.

Thank you in advance :)
No prob, here you go:

1) "Log Parameters" --> "Use RPM as possible" --> checked or unchecked? (what does it actually mean?)

Have it checked to get the right values. Sorry, not sure what that does.

2) "Dyno Correction"

Put no correction and a factor of 1 to get true power at the wheels. You can correct to show power at the crank (eg. 1.10-1.15; compensates for 10-15% drivetrain losses)

3) "Environment" --> "SAE" --> checked or unchecked? (what does it actually mean?)

If you check it off, it will automatically bump temperature to 25oC and At. press. to 0.99 Bar which is the standard temperature/pressure used by automotive companies when measuring their engines hp and tq. I leave it unchecked, and put in the intake temperature in the middle of the run, leave pressure at 0.99 Bar. Note that it does not correct hp/tq very well if you are comparing 2 logs with a larger variation in temperature (around >10 oC). Haven't tested pressure, but I do my runs close to sea level, and I don't think daily atmospheric changes in pressure do very much.


Bottom of the chart
4) "Smoothed" --> checked or unchecked? (what does it actually mean?)
5) v1, v2, v3
6) Factor

This is a smoothing feature which irons out some of the spikes in data. I run smoothing v2, but don't raise the factor past 6. If you go above 6-7, it will smooth it too much, and create an artificial curve which may create false losses/gains.

Here is a little useful feature... When you have 2 separate Vdyno's loaded in 2 windows, you can paste one curve into another for comparison. Click the copy icon below the Vdyno graph (2 sheets paper), then in the other graph click the paste icon (line graph) -> from clipboard. You can save it as a pic on to your computer with the save icon. Remember that the runs have to be done in the same conditions (see page 1 of this thread), and both Vdyno's should have the same settings for a decent comparison.
 
  • Like
Reactions: Lazzaro and M3Dude

·
Registered
Joined
·
17 Posts
Thx skyactiver!

My idea now is to get to know engine power on crank after the final tune only using vdyno. I choose power on crank because I know the car manufacturers specifications and can take them as a base for calculating the final result. As in vdynos the most precise data I can get from are deltas between runs I will take them to calculate the gain by adding 15 % drivetrain loss. The Procedure should be done at one single day with same weather, road and car conditions (except the tune).

Power on Crank in 10 Steps:

1. final tune is flashed and long fuel terms are set (min. 60 miles with mix of city, highway, WOT)
2. refuel the tank
3. find the most perfect road for logging in 3rd and 4th gear and log min 2 runs each (3rd and 4th gear)
4. flash stock ecu file
5. drive min 60 miles with a mix of city and highway with WOT
6. refuel the tank (same level like before)
7. take logs in the same way on the completely same road like before (weather conditions should be the same -> wind, temperature)
8. compare the logs in vdyno with equal settings (each WOT to the other) and calculate deltas
9. take the average of all deltas and add about 15 % drivetrain loss
10. add the result to your car manufactures crank power specifications and get your cars power on crank

Would this be a good way by vdyno and what do you think how precise the result would be?

I am looking forward for your feedback.

Best,
Lazzo
 

·
Registered
Joined
·
606 Posts
Thx skyactiver!

My idea now is to get to know engine power on crank after the final tune only using vdyno. I choose power on crank because I know the car manufacturers specifications and can take them as a base for calculating the final result. As in vdynos the most precise data I can get from are deltas between runs I will take them to calculate the gain by adding 15 % drivetrain loss. The Procedure should be done at one single day with same weather, road and car conditions (except the tune).

Power on Crank in 10 Steps:

1. final tune is flashed and long fuel terms are set (min. 60 miles with mix of city, highway, WOT)
2. refuel the tank
3. find the most perfect road for logging in 3rd and 4th gear and log min 2 runs each (3rd and 4th gear)
4. flash stock ecu file
5. drive min 60 miles with a mix of city and highway with WOT
6. refuel the tank (same level like before)
7. take logs in the same way on the completely same road like before (weather conditions should be the same -> wind, temperature)
8. compare the logs in vdyno with equal settings (each WOT to the other) and calculate deltas
9. take the average of all deltas and add about 15 % drivetrain loss
10. add the result to your car manufactures crank power specifications and get your cars power on crank

Would this be a good way by vdyno and what do you think how precise the result would be?
This is very much like the way I do things when I tune my own car for comparison of changes. You are planning to put a good amount of miles to let the fuel trims settle, match weather conditions, and you even went to the extra step of making sure the fuel is at the same level which compensates for fuel weight differences. You are also doing an average of 2 runs which adds more accuracy :) The only other advice I have is to make sure that you are using the exact same fuel from the same pump location in both runs to avoid fuel quality issues. I would also like to clarify that finding flat roads is very important, but even more important is that you are going WOT in the exact same spot. So it helps to find 2 spots along a straight road (spot 1 & 2), and go WOT at the same rpm when you hit a certain marker in each spot.

Below I posted examples of 3d gear Vdyno results from 2 separate flat road spots on the same road. The runs were done minutes apart on the same tune file. Notice how it does not completely replicate the power curves in spot 1 vs. 2 on a road that appears very flat. In one day, you can actually see a rather large difference between runs, and that is because they were done in opposite directions on a fairly windy day (some effect of tail/head winds) in addition to the surface variations in the actual roads and that engine output fluctuates with every WOT run. Overall, its good to compare spot 1 stock vs. spot 1 tuned, spot 2 stock vs. spot 2 tuned... so that you are averaging the peak power differences (tune-stock hp/tq) in spot 1 & 2. With this method, and what you have so far, the Vdyno should be within around +/-5 hp/tq accurate based on variations created in the runs.

Spot 1 vs. 2- Day 1.jpg

Spot 1 vs. 2- Day 2.jpg

Another great tool to use is the time to complete runs. It gives you an rpm breakdown of where your car is actually faster in real world driving conditions. Just click on the "Speed & Distance" tab in the larger Dyno tab. This will be a more accurate tool than just peak numbers when comparing the gains from stock. Here is a sample; note that my 3000-4000 times are going to be typically slower than 3d gen Mazda 3s because of my crappy stock 4-1 header...

Speed & Distance.jpg
 
  • Like
Reactions: Lazzaro
1 - 20 of 21 Posts
Top