-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
GoFundMe to Dyno the AWD (non-Performance)
- Thread starter Wildbill
- Start date
HeatproofGenie
Member
Yes that's RWD. Which the whole point of the thread is that there aren't good AWD dynos out yet!
You missed the point of my pose completely. A dyno pull isn't a reliable or meaningful number in absolute terms. It is a tool best used relatively when measuring power improvements on the same vehicle, on the same dynamometer, with upgraded parts. In other words, our two exact same cars on different rollers will have different numbers, which means the measurement is of no use to compare our two exact same cars. This is extremely widely known in the tuning industry and covered at great length on tuner forums.
Totally agree! The exact same car measured in different type of dyno can show vastly different horsepower and torque numbers.
I would just advice the people who really want to know what kind of performance they are getting relative to competition such as BMW M3, Mercedes C class AMG etc. They really should just look at 1/4 drag time in terms of seconds and trap speed in mph. It is a much much better measure of real world performance comparison between vehicles. 1/4 mile performance takes account into the weight of the vehicle, traction, transmission efficiency and performance etc, while doing dyno is just about the performance of the engine/motor and nothing else. 1/4 mile is very much as real world as you can get.
I know you are looking for non-performance model 3 numbers. But just as an example, here is the comparison of Model 3 performance vs BMW m3.
Last edited:
spesler
Member
I’m not sure whether you are implying “AWD” non -P, but this video is also a RWD car. At one point the rollers are linked to roll the front wheels.
Last edited:
Actually, doing the way you suggested is not taking advantage of time and science. It's better to start at a low SOC, dyno test at the low SOC, go charge up to mid SOC, head back to dyno to test, and repeat the process one more time at high SOC.
Doing it the opposite way of how I pointed out will require the owner of the Model 3 to discharge a large amount of energy down to the next testing level of SOC and that's a waste on time and energy.
I understand that different Dynos will have slightly different results, in the same way that track results will vary based on wind, tire wear, temp, track conditions, etc.
What is the standard deviation on dyno results. Are we talking under 10% tolerance? I don’t think anyone here cares if it might vary 10-30HP... most are just looking for some kind of quantified ballpark. 300 to 350HP? 350 to 400? 400 to 425hp? Ect.
I do agree that getting one other M3 ether performance or RWD for a relativity baseline would be more beneficial then different SOC baselines.
I’m also interested in looking for any nuisances in the power curves that might indicate software related throttling. Some of the draggy results showed choppy acceleration pauses on the AWD, which struck me as strange given there is no gear shifting.
What is the standard deviation on dyno results. Are we talking under 10% tolerance? I don’t think anyone here cares if it might vary 10-30HP... most are just looking for some kind of quantified ballpark. 300 to 350HP? 350 to 400? 400 to 425hp? Ect.
I do agree that getting one other M3 ether performance or RWD for a relativity baseline would be more beneficial then different SOC baselines.
I’m also interested in looking for any nuisances in the power curves that might indicate software related throttling. Some of the draggy results showed choppy acceleration pauses on the AWD, which struck me as strange given there is no gear shifting.
HeatproofGenie
Member
Thanks for explaining that to me I really needed that
The point or evolved point of the thread is to ideally get a AWD and P car on the same dyno on the same day to compare.
Not sure why it affects you that people want to do this.
I'll post here to keep the general information on this board so that anybody can come across some details for their own research until there's an agreed arrangement that can be made.
So, going by that address, I'm looking at 1.5-2 hours drive, not including traffic. For round trip realistic timing, it'll probably be 5 hour ride with traffic. Throw in time for dyno and charging, I would try to have the dyno start some time between 12-1 PM if I manage to leave by 10 AM, so that I could be back home before 4 PM or so.
For anybody who is actually closer by in Richmond who has a Performance, it's probably better to volunteer for better time management flexibility and for better appointment time.
Somewhat of a convenience is a Tesla Supercharger at about 6 miles from the dyno shop:
9950 Brook Rd, Glen Allen, VA 23059
In order to cut down on the time to dyno the car(s), SOC of the car should start low and progress higher for the next dyno run of 3 runs. But if the dyno shop is limiting the 3 runs to just back to back, then SOC of 90%-100% is beneficial to do all 3 runs.
Last edited:
Except that dynos have most of those problems plus their own inconsistencies.
There isn't one, that's the problem.
No. I have seen two of the same brand dynamometers show wildly inconsistent results. Two places I brought my own Subaru to gave results that varied by 100 HP. Slightly more, but close enough.
No, 10-30 probably wouldn't be a big deal. But that's not what we're talking about.
And I'm trying to tell you that this is not at all what a dyno is, what it does, or why you use one.
If they really had a variation of 100hp it would render the machines completely pointless and they would not exist. Perhaps you had a poor experience or someone that didn’t know what they were doing.
I appreciate your concerns but still exctied and interested in seeing the results.
I appreciate your concerns but still exctied and interested in seeing the results.
spesler
Member
In fairness, he is alluding to an accuracy vs precision issue. If two machines are wildly innacurate (as he is suggesting) but precise, yielding repeatable results, they can still be useful as a tool for comparing relative values. I.e. did this exhaust swap add HP or sacrifice torque? Assuming both runs are done on the same machine under the same conditions, it doesn’t matter that the values are off dramatically as compared to what another machine would report.
I am, however, suspicious of the reported 100hp discrepancy between machines. If true, that sounds like one of the shops is doing something wrong, whether it be in the machine’s calibration or the test setup. Of course, dynos are not usually run by scientists, and are primarily used for relative comparisons, so ::shrug::
I still want to see the results!
@spesler - completely agree and understand the accuracy vs precision tradeoff. I'd argue though that even from an accuracy standpoint the variation between different machines shouldn't be that large. I did some quick searching and the best quantified examples I found were maybe 10-20 HP differences between dyna machines.
It also appears that some of the variations between machines may be minimized by the fact these are EV's. Altitude, gas quality, air intake, and other factors that might create "some" variation in accuracy between machines will likely be further minimized. Yes, things like how well was the vehicle strapped down could still have impacts, but likely negligible, certainly for this exercise.
On that note, have we had any luck pinning down a RWD and AWD+ local in the area?
It also appears that some of the variations between machines may be minimized by the fact these are EV's. Altitude, gas quality, air intake, and other factors that might create "some" variation in accuracy between machines will likely be further minimized. Yes, things like how well was the vehicle strapped down could still have impacts, but likely negligible, certainly for this exercise.
On that note, have we had any luck pinning down a RWD and AWD+ local in the area?
They wouldn't be pointless, because your car tested on the same dyno before and after changes are made to it would result in a consistent measurement giving you a relative change. Which is the point of them. Again, the absolute number isn't the purpose of a dyno, the relative change is. If you'd like to see a tuning forum thread about some of the differences, Almighty debate:What dyno reads higher!!!! - NASIOC
This article describes exactly what has been said here- A drag strip is where you measure power. And it also highlights that some basic tricks are worth almost the 10HP figure you're suggesting. Dyno Accuracy Testing - Popular Hot Rodding Magazine
I have a decent relationship with my tuner, he uses a mustang dyno, and he is very up front about not relying on the number you see on the sheet but on the difference between before and after pulls. That is what these rollers are designed for. Even engine dynos are subject to problems, and SAE had to re-evaluate their techniques for manufacturers because the system was being gamed and car makers were saying their engines were producing power numbers that couldn't be independently reproduced.
Improper correction factors are the bread and butter of tuning shops that want to rip people off- The sell you parts and a tune, put your car on the rollers, and tell you that your 300HP stock car is now making 500HP without showing you the actual relative change, and without showing you their correction and calibration data. The industry is rampant with this.
Ok, finally. I thought I was doing a really bad job of explaining what I was trying to convey. An individual dyno (which is how you should always test your tune) on the same day should produce close numbers with the same car under the same conditions. That is the precision side of things. But an accurate measure of the absolute WHP/WTQ numbers simply isn't going to happen. The best option for that is the calculate the weight of the vehicle, drive it down a measured strip at full power, and calculate the acceleration over time to figure out HP numbers.
Either way, the 100HP number is direct personal experience. I switched shops that I get work done at many years ago because the shop I used in the Connecticut area was exposed several times for their inflated numbers. They used it as a selling technique to customers that didn't understand how a dyno worked or what it was for, and only cared about the peak numbers. Basically the exact conversation happening in this thread right now. I switched to a tuner in almost Central Mass., and have had fantastic results over almost 10 years of working with him and his crew. He will be the very first to tell anybody that the same dyno on different days will produce different numbers, and he has a massive collection of sheets that show some very unscrupulous behavior of other shops in the area. Again, this is the bread and butter of the tuning industry.
I'm not sure you fully understand how a dyno works, then, or how the numbers it outputs are produced. When the DragTimes guy did the RWD Model 3, it gave some insane torque numbers. Something like 700 or 1000 ft lbs if I recall? The car doesn't produce that. But because the data entered into the machine was wrong, and the RPM monitoring wasn't enabled, it just output garbage data. Fortunately, he knows how these things work, and he knew the numbers were bogus.
I provided a link above. I'd love to see some of your links. 10-20HP is basically an acceptable error margin between pulls on the same machine on the same day with the same car due to heat soak alone.
Strapping down, as described in a link I provided above, is good enough for almost your 10HP number alone. I'd agree that 10HP would be a negligible difference between cars, but I'm not talking in increments of 10HP. Also, the other factors you described are literally the factors in ICE cars. That's not a discrepancy in the dyno, that's a discrepancy in the vehicle and its possible performance. My car could lose 10WHP based on humidity for crying out loud, and it's turbo charged.
I think what we're seeing here is that some people have had experience with tuning and measuring vehicle performance, and some people are just breaking into it. Which I think is great, BTW. But those of us with experience know what to look out for. And like I said when the first DragTimes numbers came off their rolling dyno setup, "I'll believe it when I get the brand, correction factor, entered vehicle data, temperature, and humidity.". Without that data, it is impossible to trust any numbers that come from a dyno. And even with that data, it's only reliable to use it for relative change on the same vehicle on the same day. So, measuring performance difference vs. state of charge would be a great use of dyno time. But absolute power numbers, not so much.
If there is no correlation to actual HP coming out of Dyno machines and it is purely relational, then why wouldn't they just use a generic stepping factor of 10... remove HP out of it all together.
I'd be interested in knowing more about the math behind these machines. Below appears to be the equation for HP based on a quick google search. To calculate HP you need just two inputs RPM (should be 100% accurate from a Dyno machine) and Torque, I'm guessing the element that injects some potential error margin. Solving for Torque requires distance (I'm guessing shaft to tire?), which should be measurable, and force, which I'm guessing is the element that creates room for interpretation.
Assuming the force put onto the vehicle can be relatively accurately measured, then you should receive a fairly accurate output of HP.
Your point around the "gaming for profitability of selling products" is very interesting, and I don't doubt that for a moment. It doesn't prove though that the machines are incapable of a calculating a proper measurement when used responsibly. What do auto manufacturers that publish their HP and Torque specs utilize for testing?
(RPM * T) / 5252=HP
T= torque
Solve for torque
Where
T = torque
F = force
D = distance
I'd be interested in knowing more about the math behind these machines. Below appears to be the equation for HP based on a quick google search. To calculate HP you need just two inputs RPM (should be 100% accurate from a Dyno machine) and Torque, I'm guessing the element that injects some potential error margin. Solving for Torque requires distance (I'm guessing shaft to tire?), which should be measurable, and force, which I'm guessing is the element that creates room for interpretation.
Assuming the force put onto the vehicle can be relatively accurately measured, then you should receive a fairly accurate output of HP.
Your point around the "gaming for profitability of selling products" is very interesting, and I don't doubt that for a moment. It doesn't prove though that the machines are incapable of a calculating a proper measurement when used responsibly. What do auto manufacturers that publish their HP and Torque specs utilize for testing?
(RPM * T) / 5252=HP
T= torque
Solve for torqueWhere
T = torque
F = force
D = distance
Yes, I have the same question. What do the auto manufacturers do to test how much hp their engines generate?
And what can we use to accurately test our cars to see how much hp our engines are putting out so we can check up on the manufacturers to make sure they are not making wild claims?
Several auto manufacturers changed their horsepower ratings in 2005 or 2006, I believe, to be more in line with changed SAE power measurement standards. Basically the trick was to strip every accessory off an engine, remove the exhaust and intake that would normally be fitted, and run the engine in a stand to measure torque and horsepower. This produced inflated numbers, and the varied widely from manufacturer to manufacturer depending on how each tested their equipment.
The newer SAE standard mandated things like airflow, required equipment, fueling, engine load, and the test setup. If your car comes with an SAE horsepower/torque number and it's from after 2005 or so, then it uses this updated standard. The standard itself is how we can compare across manufacturers with some level or reliability. The dynamometer stand is an actual scientific device, though, that connects directly to the crankshaft of the engine. Manufacturers still mess with ratings, as you'd expect, but that's the general gist of it. Unfortunately, this still doesn't give the consumer a good understanding of how the vehicle will compare when driven, because generic values are assumed for drivetrain losses.
Anyway, dyno testing is a whole wacky thing on its own.
The newer SAE standard mandated things like airflow, required equipment, fueling, engine load, and the test setup. If your car comes with an SAE horsepower/torque number and it's from after 2005 or so, then it uses this updated standard. The standard itself is how we can compare across manufacturers with some level or reliability. The dynamometer stand is an actual scientific device, though, that connects directly to the crankshaft of the engine. Manufacturers still mess with ratings, as you'd expect, but that's the general gist of it. Unfortunately, this still doesn't give the consumer a good understanding of how the vehicle will compare when driven, because generic values are assumed for drivetrain losses.
Anyway, dyno testing is a whole wacky thing on its own.
Similar threads
- Article