Official VersaTuner Feature Request Thread

[mituc]

Originally Posted by Enki You also have to account for wind resistance...right?

Wind resistance, rolling resistance, and a lot of other crap which is not linearly proportional to the numbers.

Originally Posted by xfeejayx The math behind it is simple, 20 minutes tops in excel. You know your Time between readings (which is all the resolution any software like VD has), you can pick up Speed, ask the user to put in weight(Mass) and tire size(Radius), (dV/dT)*M*R, add in some filtering, done. I completely understand implementation is more difficult than that with a GUI and all. I'm not disputing the value of raw numbers, I'm just offering a feature that people seem to care about. Load curves also do not take spark advance into account, so they really can't be used to tune that. I think I'm actually going to go do this in excel right now because VD is such a pita.

Making a file for VD is a lot less than 20 minutes, and on the other hand why implement something that is not accurate when the tool which does this properly can import almost directly the data produced and it's at everyone's reach...

[Enki]

Originally Posted by mituc Making a file for VD is a lot less than 20 minutes, and on the other hand why implement something that is not accurate when the tool which does this properly can import directly the data produced and it's at everyone's reach...

Fixed

[xfeejayx]

Are you under the impression that the accuracy of VD is worth any more than a comparison between runs on the same car on the same road?

Unless you have inputs for wind speed, tire pressure, tire width, air density, type of road in VD, gtfo with that "complex" nonsense. Don't complicate this more than it needs to be for a back to back comparison.

There was a handheld programmer and tuning softwares available before VT existed. They still made it, we still use it. Not sure where this "there's already some stuff that does that in a way" is coming from.

[Ugnius@VersaTune]

Virtual dyno functionality is something that we are looking into adding to VersaTuner after we finish the major overhaul that enables tuner editions and multi-VIN licenses.
The HP/torque numbers are only as accurate as the variables inserted into the formula (mass, tire radius, frontal area, coefficient of drag, gear ratio/final drive, etc.) and we think having all that data built in to the app will make it much easier for users to get an estimate of their tuning gains and to compare their numbers to other users running similar setups.
We have access to exactly the same formula that is used by the Virtual Dyno app and we can guarantee that our calculations will be just as accurate and will account for all the variables. The main challenge is getting an accurate acceleration value at any given time. Differentiation of speed readings over time (what Virtual Dyno app does) is rather inaccurate, because the speed value is filtered and smoothed before being reported on the CAN bus and that creates a noticeable delay compared to the rise of the RPMs. Furthermore, the reported speed is rounded off to whole numbers.
In fact, we are investigating the possibility of obtaining direct acceleration readings from the onboard accelerometer of the dynamic stability control module for even greater precision.

[xfeejayx]

Originally Posted by Ugnius@VersaTune Virtual dyno functionality is something that we are looking into adding to VersaTuner after we finish the major overhaul that enables tuner editions and multi-VIN licenses. The HP/torque numbers are only as accurate as the variables inserted into the formula (mass, tire radius, frontal area, coefficient of drag, gear ratio/final drive, etc.) and we think having all that data built in to the app will make it much easier for users to get an estimate of their tuning gains and to compare their numbers to other users running similar setups. We have access to exactly the same formula that is used by the Virtual Dyno app and we can guarantee that our calculations will be just as accurate and will account for all the variables. The main challenge is getting an accurate acceleration value at any given time. Differentiation of speed readings over time (what Virtual Dyno app does) is rather inaccurate, because the speed value is filtered and smoothed before being reported on the CAN bus and that creates a noticeable delay compared to the rise of the RPMs. Furthermore, the reported speed is rounded off to whole numbers. In fact, we are investigating the possibility of obtaining direct acceleration readings from the onboard accelerometer of the dynamic stability control module for even greater precision.

Sounds like RPM isn't as filtered, maybe use that and calculate out from there? It certainly has more resolution than speed. You already have all the gearing data you need to turn RPM into speed if a tire size is input. Slower logging rates certainly won't help here.

The accuracy will be questionable until corrections are determined, but I would really just like repeatability. Even VD has to get smoothed and corrected to some value that someone correlated to real numbers at some point.

[HeTheAethyr!]

Originally Posted by xfeejayx Sounds like RPM isn't as filtered, maybe use that and calculate out from there? It certainly has more resolution than speed. You already have all the gearing data you need to turn RPM into speed if a tire size is input. Slower logging rates certainly won't help here. The accuracy will be questionable until corrections are determined, but I would really just like repeatability. Even VD has to get smoothed and corrected to some value that someone correlated to real numbers at some point.

All speed sensors are filtered regardless, either physically or post conditioned and this is dependent on the program's power required if needed. If the intake manifold pressure is limited based to 8 bit mapping, I'm sure there isn't a lot of development applied where not needed. ABS sensors being type an inductive pickup will be inaccurate at low frequency by design emphasizing on signal, shorting the window of calculation. The diesel market has a few monitors that calculate however, have been very inaccurate. I believe VersaTuner will only produce quality.

The programming to achieve sold torque numbers is going to rely on Versatuner's ability to map the engine load, which appears as if Mazda's Mitsubishi ECM base programming doesn't apply load base speed control? I see a lot of work here..

Versatuner, am I correct? If not eh, Hello MazdaSpeedForum !

Fantastic Thread !

[xfeejayx]

Originally Posted by xfeejayx "Shift light" by gear. Be able to customize the rpm warning in dashboard by different gears. custom set points by gear>auto rpm by gear ratio.

You guys getting anywhere near doing this? This would be incredibly useful for me going forward.

[xfeejayx]

@Steve@VersaTune; Can you please check/update descriptions on boost limit tables to clarify if limits are absolute or gauge? And is it consistent across all units?

[Steve@VersaTune]

All manifold pressure references in the ECU are absolute. What specific tables are you asking about?

[troubled81]

on the 3D maps can you please add black lines to the edges? I have trouble seeing irregularities with the colours used.