Interesting design but I would question the use of dyno with regards to an accelerometer. Since this only can measure the spin up characteristics of a fixed weight with zero load being applied to the motor being tested. How is this a dynamometer? Back in the days of brushed motors many companies tried to sell accelerometers as dyno's but there was only one true mfg of a dyno, Competition Electronics Turbo Dyno. Just curious if I missed something in the videos that shows how there is a load being applied beyond the weight of the metal disc?

Yes this is a flywheel dyno, similar to the Fantom Facts Machine Dyno.
It calculates the energy ie power required to accelerate the mass.

There are other methods to apply a load to a motor ie a fan, slave motor or a brake.

Each have there pros and cons.

No prob more a case of being a moderator we're meant to enforce the no paypal gift thing on the for sale forums here, so it's a rule I need to stick to as well.

For those interested in putting the MD2 dyno results good use here are a couple videos showing what you can do using RC Crew Chief.

Flywheel Dyno Analysis and Brushless Motor Modelling

Motor Timing and Gearing Analysis

Sorry for the long post to follow, but I have heard this too many time before - people thinking the CE turbo dyno was the only real dyno out there.

Absorbing dynos, like the CE, typically measure torque and rpm and calculate power.

Inertial dynos measure angular acceleration of a mass over time, and calculate power using Newtons second law, as applied for rotation. The "fixed weight" is the load in an inertial dynamometer.

One is easier to make than the other and therefore attractive from a cost perspective, but this is not the only reason to pick one over the other - they both have their places. In order for an absorbing dyno to measure the peak torque output of an electric motor, you would need to stall the drive. Unless you are using a prony brake, simply adding a "load", like a slave motor in the CE case, will not allow you to measure the full operating range, so you are still estimating both the upper and lower ends of motor performance.

Specifically, the CE turbo dyno slowly runs up the motor to the preset load steps, and measures rpm, power and torque for each load step. The slave motor is simply used to generate the load steps, and the computer keeps feeding current to the drive motor until the preset load step is reached.

As an electric motor generates peak torque at zero rpm, this information, in my opinion, is useful to be measured as opposed to approximated and is measured easiest by an inertial setup.

Choosing the mass and sampling frequency for the inertial dyno is an entirely new discussion.

Perfect explanation hprt. People don't realize that most of the full scale car chassis dynos are Inertia (Flywheel) dynos. The drum the tires spin up is the inertia wheel.

Regarding your quote above I'm having 50mm and 60mm dia flywheels machined for the MD2. The inertia will be 2.5x and 5.5x greater than the 40mm flywheel that is provided. I have found that the spin up times are so fast you get very few data points at lower RPM with the 50hz sample rate. The larger flywheels will help this issue somewhat. I would be interested in your thoughts.

BobW - What are your thoughts about running 1 7.4 battery pack vs 2 7.4 battery packs?

Running dual batts does two things.

First it splits the motor load so you won't blow as many fuses when testing 13.5 motors. So if the peak motor draw is 80 amps you will only draw 40 amps per side. Since the fuses are only 30-35amp per side it is easy to blow them using a single batt.

The other thing it does is reduce the voltage droop you will see for the same reason. Now you have two batteries in parallel which reduces the effective Battery internal resistance the system sees. You will get different results testing the same motor using single versus dual batteries so don't try and compare them directly.

Hi Bob,

Using a flywheel with a larger moment of inertia will increase the number of readings per rotation as you already know. According to Newton, doubling the inertia will reduce by the same factor the angular acceleration, assuming the torque applied by the motor under test is the same for both cases. I quickly calculated the number of samples recorded during the same angular displacement.

For a flywheel with 2.5x the inertia of the OE unit, you would get about 148% more samples across the run. For the 5.5x inertia flywheel, you would end up with 234% more samples. The sampling frequency hasn't changed but the flywheel is taking longer to spool up and, as a result, we have more samples collected over the time period. If the total run time is 2 seconds with the OE flywheel, expect the new spool time to be about 1.48x (for the 2.5x flywheel) and 2.34x (for the 5.5x flywheel) longer, or 2.96 and 4.68 seconds respectively.

This increased number of samples per run will give more accurate results when plotting the motor performance, and based on the above, I would focus efforts on the 5.5x inertia flywheel. The 2.5x will not give significantly better resolution during the spool time.

This will put more load on the drive circuit. I don't think you will risk popping a 35 amp ATO fuse with a 21.5, but you might with a 17.5 or 13.5. I would run two batteries using the 5.5x inertia flywheel. Back to back dyno runs might actually see some effect from motor heat...

Let me know how you make out on this. I might have to get one made out here.

Totally agree. I ran some simulations before I picked the Flywheel sizes. Comparision between the spin-up times for a 40,50 and 60mm dia flywheel are in the images below. Larger Flywheels will definitely make a big difference in the samples at lower RPM. Worried about blowing fuses though. If that becomes a problem I may try a different DAQ system.

I can post the drawings I'm using to have the parts fab'd if anyone is interested.

I have attached an ATO fuse time-current characteristic chart from Littlefuse. For the 35 amp fuse, it can handle about 200% of its rated current for at least 150ms. This means two fuses in parallel could safely handle 140 amps for about 7 consecutive dyno readings. I have never seen current spikes that high with any of the 17.5 or higher turn motors I have tested. I haven't run any 13.5's as I don't run any spec classes with that turn motor.

I would love a copy of your plans for the 60mm flywheel.

My dyno came with 35amp fuses but the rating says 30 amp. Not sure what's the deal with that. Haven't blown any fuses lately but I will have a good supply on hand when I get the new Flywheels.

Flywheel drawings are attached. The shroud details are just a suggestion. I'm letting the fabricator figure out something easy. Drawing will plot to scale on 18"x24" paper

thanks Bob. Nice work.

Civil Eng.?
I'm a Mech.

I'm Mech as well.:nod:

hprt that is excellent information great explanation thank you.

BobW thanks for posting the drawing I will have to try one.

A further note on the dyno and consistency.

When running in brushed mode the consistency is very high but in brushless mode the consistency varies depending on the ESC used.

The best ESC I have used with the dyno is the LRP sphere. With this ESC I was able to make small adjustments to timing and rotor position and see the change in power output.

With the x car ESC the output varied a lot from run to run which whilst it was a cheap ESC I did not find the ESC helpful in tuning the motor. (perhaps the x car ESC with HW software loaded will work better?)

By finding the timing sweat spot and optimizing the rotor and sensor position I managed to gain 5% more power with several motors which has been noticeable on the track.