I thought it would be a good idea to write up instructions for dynoing the M3 motor on a hub-attached dyno. Hub-attached dyno's (such as the Dynapack, and Rototest) are going to offer the most accurate results -- far more accurate than any type of roller, or chassis dyno. Hub-attached dynos are more accurate because they operate on the same principles as a real motor dyno. Motor dyno's work by running the motor at wide-open-throttle, then applying a hydraulic load on the motor until it bogs down to a specified RPM. Once you arrive at your desired RPM, a torque sensors measures the amount of torque at the crankshaft.
Hub-attached dyno's operate on exactly the same principle. The wheels are removed from the vehicle, and attached to the dyno. The car is put in gear, and run at wide-open-throttle. The dyno applies a hydraulic load to bring the RPM down to a specified point, and torque measurements are taken. This is by far the most accurate type of dyno, and is typically accurate to 1/10th of a ft/lb of torque.
Comparing to RRI dyno results
I've seen the RRI dyno results. The RRI dyno and Dynapack 4000 are
essentially the same thing. Regardless, if you compare my results to theirs, you
will see they are apples-to-oranges. 1) They dyno'd over an extended period of
time and took the best numbers they could come up with. You can see this by
downloading the data and seeing the temperature change by 6 degrees and the
humidity change by 6% during these dyno runs. 2) They used a steady-state dyno
run instead of a ramp dyno run. In a steady-state run, they hold the car at a
specific RPM for 5 seconds, then move to the next RPM point and measure again.
3) They are using 98 RON octane, and I am using 91 (R+M/2) octane -- a
difference of about 4 points. 4) They used 4th gear (you can tell by analyzing
the raw numbers), whereas we used 3rd gear.
How to dyno on a 6MT:
- Make sure you have enough air flow. The M3 motor will substantially reduce horsepower output if it detects there is not enough air flow. Horsepower can be reduced by as much as 30-50 horsepower. Ideally, you would want one big 70MPH fan that will blanket the entire front of the car. But if this is not possible, multiple high-output smaller fans may be used. I recommend at least three high-velocity smaller fans, pointed at the drive's side air intake (in the lower air duct), one fan at the lower radiator, and one fan at the upper radiator (pointing into the upper air intake duct). The transmission cooler is behind the passenger side lower air duct, and doesn't need to be used.
- Do not place the ambient temperature sensor in the air intake. The Dynapack implements an ambient air temperature sensor on a long cable. Do not place this sensor in your air intake. The temperature reading on this air sensor is intended to measure ambient temperature, not intake temperature. The values reported by this sensor directly affect the SAE 1349 horsepower correction formulas. By placing the sensor in your air intake, it will read too hot, and will substantially increase the horsepower readings -- sometimes by as much as 8-9 percent. It's best to leave this sensor hanging by the dyno, in the shade, away from direct sunlight and heat.
- Use 3rd gear. If you want the greatest accuracy, chose the gear with the greatest mechanical efficiency. Forget the myth that you should always dyno on a 1:1 gear; This theory is based on the mythical belief that the 1:1 gear does not suffer any drivetrain losses. Like I said...it's just a myth. Every gear, whether it's 1:1 or not is comprised of two helical gear-halves which mesh together. The output gear drives the output shaft in the transmission, which in turn drives the final drive gears. So regardless, you're always going through the same sets of gears, and always through the final drive ratio. Besides, if you dyno in 5th gear (1:1), you'll hit the speed limiter anyways. Instead, you should dyno in the gear with the greatest mechanical efficiency (or least mechanical efficiency losses). The ME of the gears on the 6MT are approximately as follows: 1st) 0.92, 2nd) 0.52, 3rd) 0.42, 4th) 0.49, 5th) 0.64, 6th) 0.62. So you can see that not only does 3rd gear have the greatest mechanical efficiency, but 5th gear has the second-worst.
- Do not let the dyno operator "guess" about your gear ratio. The dynapack can guess your gear ratio by holding the motor at a fixed RPM and measuing the hub speed. If you do not use the exact gear ratio, your results will not be accurate.
- Input 6.084 as the gear ratio. Instead of guessing your gear ratio, input the value of 6.084. 6.084 = 3rd gear * Final Drive Ratio: 6.084 = 1.5818 * 3.8462.
- Use a RPM ramp testing, not stead-state testing. Steady-State testing is performed by holding the motor at a fixed RPM and measuring the torque output. Steady-state testing is the most accurate and will yield the highest output. Auto manufacturers all use steady-state testing on their motor dynos. However, steady-state testing also takes a long time because you need to hold the motor at each test RPM for 4-5 seconds; it also might put more wear on your motor. Instead, use a ramp that sweeps through the RPM range. Ramp testing will yield lower results, but it simulates driving up an infinite length hill. With a sufficiently long RPM ramp time, you can yield results that are nearly identical to state-state testing.
- Dyno from 2000-8400 RPMs. You really want to measure as much RPM range as possible. Start at 2000 RPMs and end at 8400 RPMs. You may never reach 8400 RPMs due to the load placed on the motor by the dyno; but set the upper limit to 8400 RPMs.
- Set a 16-second ramp time. It will take 16 seconds to measure the results from 2000-8400 RPMs. The default ramp time is 5 or 6 seconds -- which is FAR too low and leads to lower results. You want to place as much load on the motor as possible for as long as possible to simulate steady-state testing. This can only be done by selecting a sufficiently long ramp time. In this case, 16 seconds is perfect, and will yield nearly idential results to steady-state testing.
- Set a 2-second settling time. The dyno will pre-load your motor for two seconds before releasing it and starting the RPM ramp. The settling time is needed so the motor will not immediately overshoot and produce oscilating torque readings. Using a 2-4 second settling time will minimize the overshoot and oscilating torque readings.
- Turn off your traction control. Press and hold the DSC button until traction control is disabled. If traction control is enabled, the dyno will not work correctly.
- Turn off your air conditioning. I found the only way to turn off the air conditioning completely was to manually turn the fan off. This disables the air conditioning compressor.
- Turn off all of your electronics. Make sure your stereo, headlights, driving lights, running lights are all turned off. Some of these may only be turned off through the iDrive menus.
- Take your USB flash drive with you. The Dynapack can store your results to a USB flash drive. Be sure to save your results, and take them home. After all, you paid for them. When you get home, download the Dynapack viewer, load your files, and look at your results on your own computer.
- Report results as "SAE Adapted." The most modern horsepower correction formulas use the "SAE Adapted" formulas. It doesn't matter what settings you used when you dyno tested because the Dynapack will store the raw data into your flash drive. After you get the results home, you can use the Dynapack viewer and view, print, output graphs using any one of five different horsepower correction formulas. When reporting to the forum, use the "SAE Adapted" output.
Thanks to m3post user 's78' for providing the calculations for gear ratio Mechanical Efficiency.
Anybody considering chassis dynoing their new M3 should read the following
Dynamometer Testing and the Modern BMW Engine
It's a real eye-opener to explain all the possible reasons why we've seen such different results to the chassis dyno testing -- even with the same dyno technology. After reading it, I realized that my own runs were not done to these standards -- and now I'm contemplating re-running them (again).
If you are considering dyno testing and posting the results here, you might want to contact dynapack (www.dynapackusa.com) at (559) 292-3800. Their dyno technology is far superior to roller dyno's -- like dynojet, etc. -- by mimicing the way a real motor dyno works. They have some links at their web site of dyno locations throughout the country, but I found by calling them, that there were numerous other installations in my own area.