Royal Purple + other fluids/mods
04-12-2011, 12:35 PM
#33
If Mike is convinced his one (or even other G owners) dyno pull showing a 2.63% increase in TQ and 3.58% increase in HP (which, by the way, falls within the +/- 4-5% error range of even the best (most accurate) chassis dynos) represents a noticeable (or even statistically significant) increase in performance, then he is entitled to his opinion, and I'll leave it at that
My baselines are run until the power stabilizes. There is no 4-5% error range on this dyno. Each baseline was run till the power stabilized. Published error range for a dynapack is 0.3% That would equate to a less than 1 hp variance.
I'm sorry, but you are sadly mistaken. Please don't believe all the hearsay on the internet.
Originally Posted by www.dynapackusa.com
Repeatability:
Our Dyno runs are repeatable to better than 0.3%. Other dyno manufacturers claim to be repeatable, but no other chassis manufacturer is even close to the level of repeatability we achieve. One large reason for this because we have eliminated the largest variable of all - the tire to roller interface. Rubber tires don't hold traction against a steel roller very well. Add a year or so of use, and the rollers become polished by the tires and traction decreases further. Some companies charge extra for special coatings on the rollers - which quickly wear off. When you have this variable link in your data chain, you cannot have guaranteed repeatability - PERIOD. Sure a roller dyno itself may be repeatable, but as soon as you put a car on it, all bets are off. Many people think that this slippage only occurs in high power situations, but we've seen it happen with 250HP Hondas - ask some of the import tuners who have had guys sitting on the hood and fenders trying to get the tires to hook up. With the Dynapack, we use a direct mechanical coupling to make absolutely sure that there is no loss, no slippage, and no inconsistencies in this area. We have virtually no inertia to mask small details and we use hydraulics for the ultimate in sensitivity and precision. The Dynapack is absolutely the most consistent and repeatable chassis dyno in the world.
Sensitivity:
We can reliably measure minute differences not seen on other machines.
Some examples include:
.010" change in spark plug gap
Differences between various lubricants
The alternator load when the headlights are turned on (in real time as well)
A single step fuel jet change
Different spark plugs
Our Dyno runs are repeatable to better than 0.3%. Other dyno manufacturers claim to be repeatable, but no other chassis manufacturer is even close to the level of repeatability we achieve. One large reason for this because we have eliminated the largest variable of all - the tire to roller interface. Rubber tires don't hold traction against a steel roller very well. Add a year or so of use, and the rollers become polished by the tires and traction decreases further. Some companies charge extra for special coatings on the rollers - which quickly wear off. When you have this variable link in your data chain, you cannot have guaranteed repeatability - PERIOD. Sure a roller dyno itself may be repeatable, but as soon as you put a car on it, all bets are off. Many people think that this slippage only occurs in high power situations, but we've seen it happen with 250HP Hondas - ask some of the import tuners who have had guys sitting on the hood and fenders trying to get the tires to hook up. With the Dynapack, we use a direct mechanical coupling to make absolutely sure that there is no loss, no slippage, and no inconsistencies in this area. We have virtually no inertia to mask small details and we use hydraulics for the ultimate in sensitivity and precision. The Dynapack is absolutely the most consistent and repeatable chassis dyno in the world.
Sensitivity:
We can reliably measure minute differences not seen on other machines.
Some examples include:
.010" change in spark plug gap
Differences between various lubricants
The alternator load when the headlights are turned on (in real time as well)
A single step fuel jet change
Different spark plugs
04-12-2011, 01:25 PM
#34
The dead pedal has 4 flathead (flush) screws and just goes over the exisiting dead pedal.
For the brake pedal, you remove the OE rubber pedal cover then drill and screw the new pedal into the metal pedal base.
In most cars there is no removable pedal cover for the gas, so you need to drill into the exisiting gas pedal and use 4 screws to attach the new one.
Here's a pic from my '04 Solara.
Last edited by DocJohn; 04-12-2011 at 01:32 PM.
04-12-2011, 01:43 PM
#35
I had a set of pedals (gas, brake & dead pedal) made by Ultimate Pedals for my '04 Solara and '07 Maxima. They do custom work and you can send them tracings of the shape of the pedals you want, and you can choose from lots of options in how they fabricate them. SEE: ULTIMATE PEDALS
The dead pedal has 4 flathead (flush) screws and just goes over the exisiting dead pedal.
For the brake pedal, you remove the OE rubber pedal cover then drill and screw the new pedal into the metal pedal base.
In most cars there is no removable pedal cover for the gas, so you need to drill into the exisiting gas pedal and use 4 screws to attach the new one.
Here's a pic from my '04 Solara.
The dead pedal has 4 flathead (flush) screws and just goes over the exisiting dead pedal.
For the brake pedal, you remove the OE rubber pedal cover then drill and screw the new pedal into the metal pedal base.
In most cars there is no removable pedal cover for the gas, so you need to drill into the exisiting gas pedal and use 4 screws to attach the new one.
Here's a pic from my '04 Solara.
04-12-2011, 01:50 PM
#36
Doc, I just show 5 pulls on that, but like I said, my dyno files are available to anyone who asks.
My baselines are run until the power stabilizes. There is no 4-5% error range on this dyno. Each baseline was run till the power stabilized. Published error range for a dynapack is 0.3% That would equate to a less than 1 hp variance.
I'm sorry, but you are sadly mistaken. Please don't believe all the hearsay on the internet.
My baselines are run until the power stabilizes. There is no 4-5% error range on this dyno. Each baseline was run till the power stabilized. Published error range for a dynapack is 0.3% That would equate to a less than 1 hp variance.
I'm sorry, but you are sadly mistaken. Please don't believe all the hearsay on the internet.
Even with accepting your dyno results as 100% accurate, please explain how a 2.63% increase in TQ and 3.58% increase in HP represents a, “significant, noticeable increase in engine output or performance” as I originally stated. (That is to say you can actually feel a difference when you drive the car).
04-12-2011, 02:01 PM
#37
These are pretty nice pedals. The ones for my '07 Maxima were even nicer IMO and had the flat domed rubber inserts like this.
That style pedal wasn't available yet in '04 when I had the set made for my Solara.
04-12-2011, 02:15 PM
#38
Mike: Sorry, I didn't see this was the average of 5 runs. However, I would like to see the published study that established a +/- 0.3% error over the range of measurement for the Dynapack. I bet it was published by Dynapack without any independent testing or standardization.
Even with accepting your dyno results as 100% accurate, please explain how a 2.63% increase in TQ and 3.58% increase in HP represents a, “significant, noticeable increase in engine output or performance” as I originally stated. (That is to say you can actually feel a difference when you drive the car).
Even with accepting your dyno results as 100% accurate, please explain how a 2.63% increase in TQ and 3.58% increase in HP represents a, “significant, noticeable increase in engine output or performance” as I originally stated. (That is to say you can actually feel a difference when you drive the car).
I said the gains were statistically significant beyond the error range of the dyno.
By your reasoning, all intakes, exhausts, test pipes, headers, and tunes do nothing for the VQ37, because those gains all "fall within the error range of the dyno".
If you don't like my evidence, you don't have to accept it.
The first baseline was roughly a dozen pulls. I simply hid all the pulls except for a few. Same with the 2nd. All pulls were back to back in an attempt to measure the effect of heat soak as well, which apparantly was not a factor; IAT did not vary by more than a few degrees.
I have over 100 fully documented pulls on the dynapack, and I absolutely guarantee that my results are 100% repeatable. This dyno has ZERO possibility of slip, zero variables that involve the tires (condition, alignment, slip, tire pressure, rolling resistance, etc), and zero variables when it comes to inertia, since the dyno doesn't use inertia to measure torque.
How many other dynos are there on the market that will allow you to hold a given RPM under variable throttle conditions for partial throttle tuning?
04-12-2011, 03:32 PM
#39
Please don't misquote me. I said significant, not noticeable.
I said the gains were statistically significant beyond the error range of the dyno.
By your reasoning, all intakes, exhausts, test pipes, headers, and tunes do nothing for the VQ37, because those gains all "fall within the error range of the dyno".
If you don't like my evidence, you don't have to accept it.
The first baseline was roughly a dozen pulls. I simply hid all the pulls except for a few. Same with the 2nd. All pulls were back to back in an attempt to measure the effect of heat soak as well, which apparantly was not a factor; IAT did not vary by more than a few degrees.
I have over 100 fully documented pulls on the dynapack, and I absolutely guarantee that my results are 100% repeatable. This dyno has ZERO possibility of slip, zero variables that involve the tires (condition, alignment, slip, tire pressure, rolling resistance, etc), and zero variables when it comes to inertia, since the dyno doesn't use inertia to measure torque.
How many other dynos are there on the market that will allow you to hold a given RPM under variable throttle conditions for partial throttle tuning?
I said the gains were statistically significant beyond the error range of the dyno.
By your reasoning, all intakes, exhausts, test pipes, headers, and tunes do nothing for the VQ37, because those gains all "fall within the error range of the dyno".
If you don't like my evidence, you don't have to accept it.
The first baseline was roughly a dozen pulls. I simply hid all the pulls except for a few. Same with the 2nd. All pulls were back to back in an attempt to measure the effect of heat soak as well, which apparantly was not a factor; IAT did not vary by more than a few degrees.
I have over 100 fully documented pulls on the dynapack, and I absolutely guarantee that my results are 100% repeatable. This dyno has ZERO possibility of slip, zero variables that involve the tires (condition, alignment, slip, tire pressure, rolling resistance, etc), and zero variables when it comes to inertia, since the dyno doesn't use inertia to measure torque.
How many other dynos are there on the market that will allow you to hold a given RPM under variable throttle conditions for partial throttle tuning?
Also, just because your dyno results fall within the manufacturer’s stated (amazing) +/- 0.3% range of error does not make them statistically significant.
Thus, the lesson begins... To test the hypothesis that Motul V300 makes a statistically significant difference in engine power (HP and/or TQ) as compared to your "control" oil, one would need to see your dyno run data for peak HP and TQ for each of your dyno runs with your “control” oil, and then for each of your dyno runs using Motul, and then perform a paired t-test for the data sets.
You can do this yourself here: GraphPad QuickCalcs: t test calculator
“Group one” would be the HP for each of your runs with the control oil, and “Group two” would be the HP with each run using Motul. You would then do a second t-test for your TQ numbers. This test would indicate if there is as statistically significant difference in HP and/or TQ for these 2 data sets.
There are certainly some confounding variables and bias that are not accounted for in this statistic, and we are assuming 100% accuracy in the measurement of TQ and HP, but its still a big step in the right direction in establishing statistically significant proof in this situation.
04-12-2011, 03:36 PM
#40
I am familiar with how statistics work. The variance between different pulls from the highest to lowest, on both days, is less than 0.5 hp. We're talking highest pull of the day vs lowest pull of the day within both the control set and the 300V set.
We don't need to run a t-test to see that the difference is significant. The variance within each set of samples is about 6% of the difference between the two data sets.
Again, I offer to send you my dyno data, if you want to run some statistical analysis yourself. I'm happy with my results, and am convinced beyond any doubt that the gains are significant. You're welcome to disagree. I'll continue to buy my $18/qt oil.
The gains from switching from conventional to 300V is more than the gains I had from adding a headerback, which is about the same gain as I got from getting a tune.
Are you now telling me, that the test pipe's gains are not statistically significant? What about the gains from the tune? If a dyno's error range was 4-5%, then every single addition to my car, individually, would fall within that range. The sum, however, would not be (albeit, barely).
I also have lots of on-track data to back up the individual gains, in the form of velocity increases on straights (albeit with a lot of confounding variables; this is somewhat mitigated by the sheer number of samples from other drivers). If corner exit speed is the same as before, but velocity at the end of the straight is higher, then either the car got lighter, or the car is putting down more power. My car didn't get any lighter from switching oils
Of course, better oil could mean that the engine is running a little cooler....
We don't need to run a t-test to see that the difference is significant. The variance within each set of samples is about 6% of the difference between the two data sets.
Again, I offer to send you my dyno data, if you want to run some statistical analysis yourself. I'm happy with my results, and am convinced beyond any doubt that the gains are significant. You're welcome to disagree. I'll continue to buy my $18/qt oil.
The gains from switching from conventional to 300V is more than the gains I had from adding a headerback, which is about the same gain as I got from getting a tune.
Are you now telling me, that the test pipe's gains are not statistically significant? What about the gains from the tune? If a dyno's error range was 4-5%, then every single addition to my car, individually, would fall within that range. The sum, however, would not be (albeit, barely).
I also have lots of on-track data to back up the individual gains, in the form of velocity increases on straights (albeit with a lot of confounding variables; this is somewhat mitigated by the sheer number of samples from other drivers). If corner exit speed is the same as before, but velocity at the end of the straight is higher, then either the car got lighter, or the car is putting down more power. My car didn't get any lighter from switching oils
Of course, better oil could mean that the engine is running a little cooler....
Last edited by Mike; 04-12-2011 at 03:45 PM.
That's really amazing! I've never seen a car that run 5 consecutive dyno pulls with less than +/- 0.5 HP variance. How do you do this? What's your secret?
04-12-2011, 04:01 PM
#42
Dynapacks bolt to the hubs of the car; they don't use rollers. Tires are taken out of the equation. On a roller dyno like a Mustang or Dynojet, you can change the reading with something as simple as how tight the car is strapped down.
Another example, is that once the gear/axle ratios are entered, you can hold a certain RPM, while increasing load, for partial throttle tuning.
You can also measure flywheel horsepower with the correct adapter and setup.
Many criticize the Dynapack because it reads higher; this is because some of the drivetrain loss is eliminated (the tires).
The particular dyno that my baselines are on are on is rented by many so-cal based companies/distributors due to the repeatability of the results.
04-12-2011, 04:12 PM
#43
I just did some more online research on the Dynapack. It really does seem to be a huge leap forward in technological improvement over the previous generation of dynos, so maybe it is indeed possible to get this kind of accuracy and testing results.
My congrats to you Mike! You are the G-god of dyno testing here and appear to really know what you are doing. I'm sorry I doubted the accuracy of your findings on Motul oil.
My congrats to you Mike! You are the G-god of dyno testing here and appear to really know what you are doing. I'm sorry I doubted the accuracy of your findings on Motul oil.
04-12-2011, 04:47 PM
#45
Registered User
Thread Starter
Join Date: Mar 2011
Location: NYC
Posts: 193
Likes: 0
Received 0 Likes
on
0 Posts
Those pedals look so sweet!!! I really like there look but on there website it seems that they only offer them for the G35...does anyone else have any other companies that make pedals like that which would fit on a 2011 G37x? Thanks.
By the way I am LOVING the G37...this is my first car ever (one that I can call my own and make the payments for ect) and it truly is a pleasure to drive. I'm happy I made the right choice!
By the way I am LOVING the G37...this is my first car ever (one that I can call my own and make the payments for ect) and it truly is a pleasure to drive. I'm happy I made the right choice!