How to use Virtual Dyno Properly
 

How to use Virtual Dyno Properly – by Boosted Tuning & TSCOMPTUNED

**This how-to is a work in process and will be added to and/or altered**

Virtual Dyno (VD) is Brad Barnhill’s standalone program for plotting horsepower and torque. It was developed to give enthusiast a way to know their power without having to go to the dyno. VD is mainly based on and tested against a Dynojet Dyno in ‘Uncorrected” mode. It uses .csv files to operate, which most loggers by default already export their file format data as.

Virtual Dyno has been compared to many Dynojet dynos and is very accurate, usually reading within 0-3% of a REAL Dynojet dyno. Here are some comparison tests I have conducted between VD and Real Dynojet 424 Dyno to show VDs accuracy: http://www.boostedtuning.com/index.php?route=product/product&path=160&product_id=237

Sometimes people fall victim of unknowingly posting inflated graphs on the forum without realizing it themselves. In this How-To, we are going to explain how to use VD properly and how to detect an incorrect, inflated graph vs. a correct graph.

The Basics

VD uses log files with data from your car along information about your car to produce a dyno chart. The information about your car that you have to enter is very important to the accuracy of VD, so its essential to make sure this information is 100% correct.

The first part is selecting your vehicle make and model and transmission. Since transmission/final drive gearing highly affects the numbers, make sure you have the correct model and transmission selected. If you have a transmission & final drive from a different model, then you would select that model instead of your actual model. Basically you are selecting your transmission and final drive ratios when you select a car. If you have custom gear/final drive rations, then you will need to use the “Car Editor” to edit your car.

Next is the car weight. When you select a car, VD enters the STOCK CURB WEIGHT for that vehicle. The STOCK CURB WEIGHT is the vehicle in 100% stock configuration, with all the fluids needed for the car to operate at the full level. Modifying a car in anyway will change the weight. Even something as simple as a turboback exhaust or a low tank of gas will take off 30-40lbs. In reality, the best way to ensure you are entering the correct weight is to have your actual car weighed.

Now you have to enter the occupant (Occ.) weight. This is the weight of the driver and any passengers in the car. Also included is the weigh of any random items you may have in the vehicle ie tools, stereo system, etc.

VD also uses tire height as crucial information for calculation power. To calculate tire height, you need the complete tire size. For the stock EVO 8/9 this was a 235/45/17, which equals 25.33 inches. If you are running a different tire size then stock, you will have to change this value. There is a tire size calculator including in VD to help you figure out the tire height.

Finding the Proper Road

When using VD, the roads on which the logs are taken from is very important. If the road has any bumps, dips, or uneven road surfaces in general, it could throw your readings off. It is also important to note that the computer you’re using needs to be able to log quick enough and gather enough data for the software to utilize and show an accurate power estimate. When using a slower machine, we recommend unchecking “Log Data to Screen” under the “Logging" drop down menu in EvoScan.

Also, some newer computers also have SSD & HDD sensor protection built into them. If this safety feature is enabled, it will halt the read/write process of your hard drive / Solid State Drive during logging, and make your logs pretty much useless. Please make sure all these features are disabled.

A great, simple way to check the flatness of a road, is to take two logs, one going one way and the other going the opposite direct on the same stretch of road, while making sure AFR, timing and boost are identical during both runs. If the logs have the same power output, you have a great road. If you have a large difference in the two logs, then you should average the numbers or better yet, find a new, better road.

The next example is a more in depth way of finding a suitable road to use VD on (and also a great way to find errors in logs).

The ‘0” smoothing setting in VD can be used to determine if the road you’re using to do pulls on is sufficient enough for accurate data simulation of the DJ. The green graph represents what a good log from an acceptable road should look like, with smoothing at 0:


https://cimg7.ibsrv.net/gimg/www.evo...4ea199a2b2.jpg


The yellow graph below represents a bad road/log that is going to inflate the number quite a bit:

https://cimg4.ibsrv.net/gimg/www.evo...08fec8719b.jpg



How to Correctly Use Smoothing (Under Smoothing / Over Smoothing)

Misuse of the smoothing feature can mask incorrect or inflated runs. The first graph is smoothed correctly and is acceptable and should be considered accurate, while the second graph is under smoothed and is considered “inflated”.

https://cimg0.ibsrv.net/gimg/www.evo...44afe5e3bd.jpg



https://cimg9.ibsrv.net/gimg/www.evo...34297bab51.jpg



While the second graph may look “Okay” to the average viewer, it is under smoothed, which is making a bump of power at 6500rpm, that will lead to inflated numbers. This run needs to be smoothed more, but for now, lets have a closer look at this graph with smoothing 1:

https://cimg7.ibsrv.net/gimg/www.evo...f00cbf550c.jpg


As you can see, with smoothing 1, its very apparent that the bump is unnatural and inflating the numbers. And although we can smooth this graph to the point it almost looks accurate, as seen below, the real solution would be to discard this log and get a better log. If you have to use max smoothing to make the chart look right, most likely the log is incorrect and shoudnt be used. This is over smoothing and shouldn’t be done.

https://cimg9.ibsrv.net/gimg/www.evo...c040db4055.jpg



Using Smoothing to Verify the Accuracy of the Logs

The smoothing function in VD can be used to determine how accurate your power output is based on the consistency of power from Smoothing 1 to Max. A good rule of thumb is, if your power drops less then 12-15WHP from smoothing 1 to Max, then the run is accurate and will most likely make very similar numbers on a real Dynojet dyno, usually with 0-3%. If your power drops off 20-25WHP or more from smoothing 1 to Max, you should be aware that the VD chart may not be accurate and your power could be off 5%, if not more, from a real Dynojet dyno.

Below is an example of a graph that would be considered an accurate chart. As you can see, there is next to no power droppage from 1 to Max.

https://cimg6.ibsrv.net/gimg/www.evo...1b7559aaee.jpg


Now lets view that graph with Max smoothing.

https://cimg2.ibsrv.net/gimg/www.evo...d359a116a3.jpg


As seen above, there is a difference of 9WHP from smoothing 1 to Max. This car will most likely back its numbers up 100% on a real Dynojet.

** "Max" smoothing for versions 1.1.8 and below is "10". The new version 1.1.9 has a max smoothing of "6". **


When to use the “RPM Trim” Feature

Sometimes at the end of a dyno plot, there will be a spike. These spikes will lead to incorrect, inflated numbers. These spike come from the fact that VD uses a continuous averaging algorithm and sometimes near the end of the data, there is less points to average and it can cause these spikes. Most users will over smooth this log, to get rid of the spike at the end. This is incorrect and should not be done. The correct way is to use the “RPM Trim” feature like shown below.

The first graph below shows such a spike.

https://cimg1.ibsrv.net/gimg/www.evo...5d9659041d.jpg



In the next graph, the “RPM Trim” feature has been used to fix the spike.

https://cimg2.ibsrv.net/gimg/www.evo...ed24fb1c0a.jpg

Use multiple runs to show power is not a fluke

Another thing me and TSCOMP have been doing, is using multiple runs when showing results. VD (& all road dyno software) can throw out fluke numbers from time to time. So, using multiple runs to show power is consistent and not a 1 time fluke, is a good practice that will ensure more accuracy in your charts.

I encourage everyone to read this. Its a very thorough and detailed how to regarding VD written by me and Boosted tuning!

It will be added onto & edited periodically as mentioned by Jay.


If anything is left out that you feel others could benefit from, please let us know what you're thinking!

One thing I will add is that I will do a run each way on the same road and look at the average. If it is exactly the same you have a great road. If you have a large difference in the two ways then you should average the numbers or find a new road.

Nice thread guys. Ill have to check mine per these suggestions

I'm having a hard time getting the latest version of evo scan to log with my tephra 7 :(

boosted tuning ftw
jay's great guy and has helped me a bunch.

great info on the vd most def.

dave

Added {thumbup}

Thanks. TSCOMPUSA also contributed a good amount of info and charts to help. {thumbup}

now thats thinking with your dipstick! good info Adam. :)

:beer: to TSCOMPUSA also :D

This is a great usefull and great 3d!

Thanks a lot guy!

Awesome info Jay!

I am glad that you put together this information to help people use the program more reliably. {thumbup}

Thanks for your time to put this together guys - a great resource. {thumbup} Please do continue to update and add to it.

fantastic start guys

A little off topic but how do you change the color of the hp/tq lines? Also in my hp graph with 0 smoothing i avg around 600whp and I have a small spike to almost 700.