Difference in a Dynapak dyno vs another Dynapak
Thread Starter
Joined: Jan 2008
Posts: 1,279
Likes: 9
Difference in a Dynapak dyno vs another Dynapak
Hi guys,
was having a discussion with some friends and we noticed some variance in numbers from a dynapak at one dyno operator compared to another dynapak from another dyno operator when comparing both numbers generated by the same car with at TCF 1.0 . A check on the dynapak website says that dynapaks are self calibrating so I was just wondering how the same car could produce 2 significantly different readings on two machines which are supposed to be identical. What variables would change the readings we get? Is there some kind of internal calibration or some kind of sensor manipulation that could result in differing numbers?
hoping to understand this a bit better...
was having a discussion with some friends and we noticed some variance in numbers from a dynapak at one dyno operator compared to another dynapak from another dyno operator when comparing both numbers generated by the same car with at TCF 1.0 . A check on the dynapak website says that dynapaks are self calibrating so I was just wondering how the same car could produce 2 significantly different readings on two machines which are supposed to be identical. What variables would change the readings we get? Is there some kind of internal calibration or some kind of sensor manipulation that could result in differing numbers?
hoping to understand this a bit better...
Moderator
Joined: Jul 2004
Posts: 36,017
Likes: 226
From: Queen City, NC
Dynometers are completely unreliable across each other - they are only good for showing differences as you change a car and retest on the same machine. It doesn't really matter what the numbers are, just how they change as you retest a car.
Registered User
Joined: Oct 2009
Posts: 2,220
Likes: 0
Saki is right. Air pressure, air temperature, and humidity all play a factor in engine output. Comparing dyno runs on different days, even on the same dyno, is problematic because weather will affect output.
There are several different kinds of chassis dynos out there. Dynopack attaches a module to each hub (bolted in place of the wheel) and measures torque and rpm using magnetic fields. The more torque, the stronger the field required to oppose it. Measuring electricity tells them engine output at the hubs. Dynopack advertises "Our Dyno runs are repeatable to better than 0.3%." In reality, they are subject to the same variations as everyone else, including weather, bad gas, heatsoak, etc.
Dynojet is an industry standard, using a large roller accelerated by the wheels. The mass of the roller is a known value, so measuring the rate of acceleration will give the energy, or torque, required to accelerate the roller.
Other kinds use rollers attached to a dynopack-type device to measure torque and horsepower. Mustang dynos are an example of this type.
The Dynojet is perhaps the most reliable from operator to operator, as the power measurement is simply based on the acceleration of a roller of a given mass. Even so, all dynos are best used as a tuning tool, using repeated runs on the same day to optimize engine output. Dynopack numbers are notorious for being fudged by the operator to enhance numbers to justify tuning costs. And forget trying to match numbers from one Dynopack to another.
All of them incorporate atmospheric adjustments to try to mediate variations in atmospheric conditions. IOW, you input the raw numbers into a formula including air pressure, temperature, and humidity to arrive at a "corrected" number. Theoretically this will allow you to compare dyno runs across different days. In reality, the same car gives different numbers on different days, even after "correction" for conditions.
Mustang dynos tend to give lower numbers than Dynojet, and Dynopack numbers are all over the map, even from the same operator, but especially so when comparing numbers from different Dynopack operators. Work with a single Dynopack operator for tuning and don't pay attention to the peak numbers but watch the air-fuel ratios and exhaust gas temps.
Another variable is the "smoothing function" each dyno incorporates. Especially with four cylinder engines, each firing cylinder causes a jump in output, resulting in a jerky, zigzag graph. Smoothing averages the numbers so you get a more readable output. You can get higher peak numbers with less smoothing, or lower peak numbers but a better sense of the curve with more smoothing. The engine is outputting the same power but the graph will look different. My understanding is that you can even skew the smoothing function to be closer to the peaks than the valleys (ignore the time between cylinder firing, where no power is being made), with the idea that we don't care about the valleys so much. The same car will appear to make more power with "high smoothing" than "average smoothing."
Dynopack and their operators are incredibly defensive about the idea they inflate numbers. In their defense, no dyno number is accurate in the sense that a tape measure is accurate, so it is impossible to "inflate." On the other hand, all dyno numbers are so subject to operator input variables that one would be right to question numbers that seem high. "Self-calibrating" is meaningless in the context of power measurements from internal combustion engines operating in atmospheric conditions. Dynopack makes a big deal about eliminating the variables introduced by wheels and tires but entirely ignore the variables introduced by operators and atmospheric conditions, which can have a far greater influence. That's why Dynojet is the standard when it comes to "How much power does my car make on a dyno?"
If you want to know how much power your car makes in real-life conditions on actual pavement, get timed at a drag strip. And even then, well, driver and weather and traction and stuff
There are several different kinds of chassis dynos out there. Dynopack attaches a module to each hub (bolted in place of the wheel) and measures torque and rpm using magnetic fields. The more torque, the stronger the field required to oppose it. Measuring electricity tells them engine output at the hubs. Dynopack advertises "Our Dyno runs are repeatable to better than 0.3%." In reality, they are subject to the same variations as everyone else, including weather, bad gas, heatsoak, etc.
Dynojet is an industry standard, using a large roller accelerated by the wheels. The mass of the roller is a known value, so measuring the rate of acceleration will give the energy, or torque, required to accelerate the roller.
Other kinds use rollers attached to a dynopack-type device to measure torque and horsepower. Mustang dynos are an example of this type.
The Dynojet is perhaps the most reliable from operator to operator, as the power measurement is simply based on the acceleration of a roller of a given mass. Even so, all dynos are best used as a tuning tool, using repeated runs on the same day to optimize engine output. Dynopack numbers are notorious for being fudged by the operator to enhance numbers to justify tuning costs. And forget trying to match numbers from one Dynopack to another.
All of them incorporate atmospheric adjustments to try to mediate variations in atmospheric conditions. IOW, you input the raw numbers into a formula including air pressure, temperature, and humidity to arrive at a "corrected" number. Theoretically this will allow you to compare dyno runs across different days. In reality, the same car gives different numbers on different days, even after "correction" for conditions.
Mustang dynos tend to give lower numbers than Dynojet, and Dynopack numbers are all over the map, even from the same operator, but especially so when comparing numbers from different Dynopack operators. Work with a single Dynopack operator for tuning and don't pay attention to the peak numbers but watch the air-fuel ratios and exhaust gas temps.
Another variable is the "smoothing function" each dyno incorporates. Especially with four cylinder engines, each firing cylinder causes a jump in output, resulting in a jerky, zigzag graph. Smoothing averages the numbers so you get a more readable output. You can get higher peak numbers with less smoothing, or lower peak numbers but a better sense of the curve with more smoothing. The engine is outputting the same power but the graph will look different. My understanding is that you can even skew the smoothing function to be closer to the peaks than the valleys (ignore the time between cylinder firing, where no power is being made), with the idea that we don't care about the valleys so much. The same car will appear to make more power with "high smoothing" than "average smoothing."
Dynopack and their operators are incredibly defensive about the idea they inflate numbers. In their defense, no dyno number is accurate in the sense that a tape measure is accurate, so it is impossible to "inflate." On the other hand, all dyno numbers are so subject to operator input variables that one would be right to question numbers that seem high. "Self-calibrating" is meaningless in the context of power measurements from internal combustion engines operating in atmospheric conditions. Dynopack makes a big deal about eliminating the variables introduced by wheels and tires but entirely ignore the variables introduced by operators and atmospheric conditions, which can have a far greater influence. That's why Dynojet is the standard when it comes to "How much power does my car make on a dyno?"
If you want to know how much power your car makes in real-life conditions on actual pavement, get timed at a drag strip. And even then, well, driver and weather and traction and stuff
Thread Starter
Joined: Jan 2008
Posts: 1,279
Likes: 9
Saki is right. Air pressure, air temperature, and humidity all play a factor in engine output. Comparing dyno runs on different days, even on the same dyno, is problematic because weather will affect output.
There are several different kinds of chassis dynos out there. Dynopack attaches a module to each hub (bolted in place of the wheel) and measures torque and rpm using magnetic fields. The more torque, the stronger the field required to oppose it. Measuring electricity tells them engine output at the hubs. Dynopack advertises "Our Dyno runs are repeatable to better than 0.3%." In reality, they are subject to the same variations as everyone else, including weather, bad gas, heatsoak, etc.
Dynojet is an industry standard, using a large roller accelerated by the wheels. The mass of the roller is a known value, so measuring the rate of acceleration will give the energy, or torque, required to accelerate the roller.
Other kinds use rollers attached to a dynopack-type device to measure torque and horsepower. Mustang dynos are an example of this type.
The Dynojet is perhaps the most reliable from operator to operator, as the power measurement is simply based on the acceleration of a roller of a given mass. Even so, all dynos are best used as a tuning tool, using repeated runs on the same day to optimize engine output. Dynopack numbers are notorious for being fudged by the operator to enhance numbers to justify tuning costs. And forget trying to match numbers from one Dynopack to another.
All of them incorporate atmospheric adjustments to try to mediate variations in atmospheric conditions. IOW, you input the raw numbers into a formula including air pressure, temperature, and humidity to arrive at a "corrected" number. Theoretically this will allow you to compare dyno runs across different days. In reality, the same car gives different numbers on different days, even after "correction" for conditions.
Mustang dynos tend to give lower numbers than Dynojet, and Dynopack numbers are all over the map, even from the same operator, but especially so when comparing numbers from different Dynopack operators. Work with a single dynopack operator for tuning and don't pay attention to the peak numbers but watch the air-fuel ratios and exhaust gas temps.
Another variable is the "smoothing function" each dyno incorporates. Espcially with four cylinder engines, each firing cylinder causes a jump in output, resulting in a jerky, zigzag graph. Smoothing averages the numbers so you get a more readable output. You can get higher peak numbers with less smoothing, or lower peak numbers but a better sense of the curve with more smoothing. The engine is outputting the same power but the graph will look different. My understanding is that you can even skew the smoothing function to be closer to the peaks than the valleys (ignore the time between cylinder firing, where no power is being made), with the idea that we don't care about the valleys so much. The same car will appear to make more power with "high smoothing" than "average smoothing."
Dynopack and their operators are incredibly defensive about the idea they inflate numbers. In their defense, no dyno number is accurate in the sense that a tape measure is accurate, so it is impossible to "inflate." On the other hand, all dyno numbers are so subject to operator input variables that one would be right to question numbers that seem high. "Self-calibrating" is meaningless in the context of power measurements from internal combustion engines operating in atmospheric conditions. Dynopack makes a big deal about eliminating the variables introduced by wheels and tires but entirely ignore the variables introduced by operators and atmospheric conditions, which can have a far greater influence. That's why Dynojet is the standard when it comes to "How much power does my car make on a dyno?"
If you want to know how much power your car makes in real-life conditions on actual pavement, get timed at a drag strip. And even then, well, driver and weather and traction and stuff
There are several different kinds of chassis dynos out there. Dynopack attaches a module to each hub (bolted in place of the wheel) and measures torque and rpm using magnetic fields. The more torque, the stronger the field required to oppose it. Measuring electricity tells them engine output at the hubs. Dynopack advertises "Our Dyno runs are repeatable to better than 0.3%." In reality, they are subject to the same variations as everyone else, including weather, bad gas, heatsoak, etc.
Dynojet is an industry standard, using a large roller accelerated by the wheels. The mass of the roller is a known value, so measuring the rate of acceleration will give the energy, or torque, required to accelerate the roller.
Other kinds use rollers attached to a dynopack-type device to measure torque and horsepower. Mustang dynos are an example of this type.
The Dynojet is perhaps the most reliable from operator to operator, as the power measurement is simply based on the acceleration of a roller of a given mass. Even so, all dynos are best used as a tuning tool, using repeated runs on the same day to optimize engine output. Dynopack numbers are notorious for being fudged by the operator to enhance numbers to justify tuning costs. And forget trying to match numbers from one Dynopack to another.
All of them incorporate atmospheric adjustments to try to mediate variations in atmospheric conditions. IOW, you input the raw numbers into a formula including air pressure, temperature, and humidity to arrive at a "corrected" number. Theoretically this will allow you to compare dyno runs across different days. In reality, the same car gives different numbers on different days, even after "correction" for conditions.
Mustang dynos tend to give lower numbers than Dynojet, and Dynopack numbers are all over the map, even from the same operator, but especially so when comparing numbers from different Dynopack operators. Work with a single dynopack operator for tuning and don't pay attention to the peak numbers but watch the air-fuel ratios and exhaust gas temps.
Another variable is the "smoothing function" each dyno incorporates. Espcially with four cylinder engines, each firing cylinder causes a jump in output, resulting in a jerky, zigzag graph. Smoothing averages the numbers so you get a more readable output. You can get higher peak numbers with less smoothing, or lower peak numbers but a better sense of the curve with more smoothing. The engine is outputting the same power but the graph will look different. My understanding is that you can even skew the smoothing function to be closer to the peaks than the valleys (ignore the time between cylinder firing, where no power is being made), with the idea that we don't care about the valleys so much. The same car will appear to make more power with "high smoothing" than "average smoothing."
Dynopack and their operators are incredibly defensive about the idea they inflate numbers. In their defense, no dyno number is accurate in the sense that a tape measure is accurate, so it is impossible to "inflate." On the other hand, all dyno numbers are so subject to operator input variables that one would be right to question numbers that seem high. "Self-calibrating" is meaningless in the context of power measurements from internal combustion engines operating in atmospheric conditions. Dynopack makes a big deal about eliminating the variables introduced by wheels and tires but entirely ignore the variables introduced by operators and atmospheric conditions, which can have a far greater influence. That's why Dynojet is the standard when it comes to "How much power does my car make on a dyno?"
If you want to know how much power your car makes in real-life conditions on actual pavement, get timed at a drag strip. And even then, well, driver and weather and traction and stuff
Exactly what I needed to know. Weather wise, as I am in Asia, the weather is fairly consistent. The operator input variables are the exact part I am concerned about, guess I will have a get someone to dyno their stock S together with mine to get a sense of how much gains I actually have. '
Thanks very much.
Trending Topics
Joined: May 2011
Posts: 474
Likes: 4
This is very good reading. Info prepared by Sean Church. Funny coincidence is I was reading this today for no particular reason.
http://home.earthlink.net/~spchurch/id12.html
http://home.earthlink.net/~spchurch/id12.html
Joined: Jan 2010
Posts: 23,911
Likes: 4
From: Benicia, CA
This is very good reading. Info prepared by Sean Church. Funny coincidence is I was reading this today for no particular reason.
http://home.earthlink.net/~spchurch/id12.html
http://home.earthlink.net/~spchurch/id12.html