Dyno testing is probably the most widely accepted and discussed form of performance measurement and comparison for high performance tuners and enthusiasts. With many of these tests, the results can be misleading, an incorrect assumption can be made from poor testing, or variables such as airflow or driving conditions have not been recreated on the dyno to obtain an accurate measurement. To avoid this, our engineers and test team at GruppeM have put together some notes which address these variables and assumptions which should be a valuable reference for the enthusiast or tuner who wants to learn more.
The notes below address performance parts testing (e.g., baseline, install, then test). While these guidelines will help get a more accurate, one-time pull for horsepower bragging-rights (like on a dyno day), you probably already know that different dynos will output different numbers in different parts of the country. Additionally, things like the differences in fuel, weather, correction factor, the shop fan’s air flow, and even differences in tie-down techniques don’t help standardize the idea of “wheel horsepower” consistently from region to region (or shop to shop) either. That’s why it is best to always use the same dyno when comparing parts, and take into account the items noted in this article.
First off, it is preferable to have the dyno in a climate-controlled room to control heat and humidity – both are power killers. Shops with a dyno inside a climate controlled room have a notable accuracy advantage, however most dyno shops do not have this capability.
The dyno weather station has (or at least should have) the ability to correct values based on the outside weather and altitude (e.g., cars in Denver actually make less horsepower on the road then cars in Newport Beach) and this helps create consistent results immensely. However, the car’s ECUchanges its fuel trim or ignition timing based on other factors like intake-air and coolant temperatures, which the dyno has no control of, and that varies from car to car as well.
On a turbocharged engine, this effect can be even greater. Imagine getting a cold baseline in the morning, where the engine physically sees more oxygen/power, and spools up the turbo quicker. Reaching peak boost sooner will usually result in a higher torque figure. The problem is, if the sun creeps in the shop later that afternoon after your exhaust install, the dyno is compensating for the additional heat in the room with perhaps a 1-2% improvement in the readout, but the car cannot influence the forces of nature, which in this example means less oxygen is going into the motor at the same time. And while you may still see an increase over your morning baseline because the factory exhaust system you replaced was that restrictive, you’re still not seeing the full benefit of what the new exhaust has on performance. Because of this, testers frequently end up seeing a loss, when they could easily see a gain by testing the next, cool morning.
Forget about doing a baseline in August, and then retesting for a new part installed in December. The best thing to do is run the baseline test the day that product will get replaced, reducing changes in weather as much as possible. It is understandable this might be impossible if a custom turbo system is fabricated or a new engine is built, and too much time happens between runs. During this time, the engine could have been run many miles, have more wear/tear in the engine or carbon build-up, be running different fuel or different weight oil, have clogged cats/fuel filter, different/worn spark plugs, etc.—all things that make a significant difference in your next run many months later.
INTAKE AIR TEMPERATURE (IAT)
The reason we want similar ambient temperatures in the dyno room is to help us keep the IAT consistent when a run is started. This is one of the parameters an engine uses to keep ignition timing and fuel trims at an optimum level for both performance and engine longevity.
After a pull, IATs will soar because, on the dyno, there is not the same airflow present as would be on the highway at 100mph. In turbo cars this is even more prominent. While a pull on the street might start at 100F and go up to 105F in a third gear pull, the same car on the dyno – which will spend much less time at wide-open-throttle in the same gear – could make the IAT soar to an excess of 135F… in one pull! And then try another pull and wonder why the power is so down.
In normally-aspirated cars, this problem is very evident as well. While testing with the hood open helps, the intake manifold still gets hot from the rising heat of the engine block. So when testing, it is always good to find a consistent IAT you can start each pull with. The only time this may be impossible is if we are testing an intake system that has much better heat-shielding capabilities, keeping the intake always cooler than the one baselined. In that case the intake should be rewarded with cooler temps, assuming all other parameters listed here are met.
When monitoring ignition timing, it should be the same value during a dyno pull for both the baseline and the test—assuming we are testing a simple modification, like an intake or an exhaust. Of course, if the ECU itself is being tested with an upgraded software upload or “chip”, then the values will be different. In that case, make sure all other temperatures are the same.
As with the IAT, some of the ECU parameters change the ignition timing values and fuel trim (meaning changes in power) when coolant temperature is not optimal. When an engine is really cold (say, sub 130F), fuel is usually rich, and one should not test a car that cold anyway. When the coolant is higher (like over 215F), not only is the heat causing the nearby intake to heat up, but once again the ECU will start to pull back timing to avoid detonation, and negatively impacts power. A properly monitored cool-down is essential between runs if the temps climb, and sometimes this only takes a few minutes.
The oil temperature can play an important role as well, but many vehicles do not provide the ability to monitor oil temperature through a scanner, so you will need a separate gauge to monitor this. The same goes for transmission and differential temperatures, which in a perfect world should be similar between tests. When the coolant is monitored properly, the oil temp will often move somewhat with coolant temp, assuming the car has an adequate cooling system and does not have an oil-cooling issue. Fortunately, some makes, like Audi and Porsche, have an oil temp gauge you can monitor to ensure it is the same prior to the next run.
Unless we are testing a car for a specific octane, the best available fuel should be used to get a proper test. If poor fuel is used, the ECU retards timing as it picks up knock noise, and many times the new performance product or upgrade being tested gets blamed. Some cars are tuned more sensitive than others for knock retardation. We have tested Porsches (turbos and NA) and BMWs that had trouble putting out consistent readings because of the ECU pulling timing back due to engine knock (and these were bone stock cars) thanks to the “not-so-good” 91 fuel grade we had when testing in California. These tests were scrapped because testing like this only addresses the ECU’s adaptive capabilities, and not the difference between a stock performance component when compared with a modification.
When testing, the car should have been recently fitted with new spark plugs. Any ignition hiccups will show up as sharp declines in the graph, giving it a choppy look that is sometimes misunderstood for knock retard. This is especially important in turbocharged engines because of increased cylinder pressure.
Always test before/after using the same gear. Most testers use third or fourth gear, with ultra-high horsepower turbocharged cars tested in fifth to get the most load and accurate torque readings. Many cars, AWD or not, will not show the same whp numbers in different gears, and the curves will differ as well. This is especially prominent in turbocharged engines because the spool-up is much earlier in the RPM band the higher the gear you go, thanks to increased load. And because different diameter tires (as in overall height) change the gearing of a car, same tire sizes should be used on the drive tires to avoid changing the gearing.
Keeping the before-and-after parameters, both in the engine and around the car, will ensure proper testing and avoid numbers that can be misconstrued as a loss, or an excessive gain, from a newly installed performance part or modification.
For example, if during a dyno test the temps are slightly off (which may result in a 5-10 whp difference on the dyno) while testing a turbo upgrade in your otherwise normally-aspirated engine, it probably will not change any perceptions about the product that just netted you 200whp. However, when testing intake and exhaust systems that are usually good for anywhere between 10-25whp, the 5-10 whp gain or drop from the aforementioned temp difference between tests significantly changes the entire perception of the value of the modification, and the tuner, tester, or owner are left with results that indicate the product is better or worse than it actually is.
The easiest way to ensure your comparisons are accurate is to take the points of this article into consideration when doing your next test, and keep a detailed journal of the environmental variables, as well as car and engine differences, settings, measures and levels, each time you test.