Needed to figure out why my EDF fan and motor maxed out at 20,000 RPM. I built a flywheel dynamometer with RPM sensor, from which angular acceleration, torque and power can be extracted. The plan is to measure voltage/current at the battery to determine motor efficiency. I’m using my Arduino Uno for datalogging.
In the album are some results from my first run. The motor really does max out at 20,000 rpm (~2000 rad/s). The max mechanical power was 50W. These are nowhere near HobbyKing specs. I’m new to motors, anyone have ideas? I’ve got a second HK2836, I’ll test it shortly.
they has miss labeled the “output power” as 520W is the input power at their specs (14.8v * 35A) and a 10% efficiency is lower than i would have thought but not unbelievable.
your disk would also be causing alot of load on the motor (due to the turbine effect, have a look at a tesla turbine and you will see what i mean) you want the smallest, lightest and thinnest disk you can
also enclose the disk in a smooth round box that has the smallest air gap to the disk you can, this will allow the air inside the box to spin at close to the speed of the disk (this is also why a vacuum cleaner goes faster when you block the hose as the air gets trapped and can spin faster there by reducing the load on the motor allowing it to go faster).
im sure you can feel the air being thrown by the disk?
I’m sorry I miss read your post… I assume your using the area of the disk and the drag it creates to calculate the power output? I thought you were just using it to calculate the max RPM. If so then ignore most of my post lol
Hey Cain, I’m only measuring RPM right now. From time stamped RPM I can determine acceleration. Since I know what the mass and inertia of the disc is, with angular acceleration, I can determine torque (torque = inertia * accel). Finally, power = torque * vel.
I’m not sure what the efficiency is yet, but I’ll find out soon. Just waiting on a hall effect current sensor.
This is great work. Do you have any idea what the losses due to air resistance on your disk are? Some years ago I witnessed an engineer make a test rig for this same purpose. He used an electromagnetic brake that was connected to a torque meter. In your rig, if the body of the motor was free to rotate, and then held from rotating by a lever that contained a strain-gauge, you could measure the torque directly. You could then add fins to the disk to alter the air load.
Please keep us informed, this is really interesting.
Edit: It was an Eddy Current Brake, not electromagnetic.
It looks like I my ESC default settings may not be suitable for the motor. Maybe this explains the underperformance. I’m going to change some settings and run another test.
Changed ESC settings, no real change. Used a 4 cell battery, achieved almost 90 Watts shaft power, maxed out the RPM at 25k.
I’m out of ideas… A 14.7V power supply driving a motor to a peak of 25k RPM means a ~1800KV motor, far from the 3800KV motor spec… Beware of some manufacturer’s specs I guess?
I think you are underestimating the air losses. And what’s you input wattage? Do you have any way of testing the rpm completely unloaded? And have you tested it with only one volt to see if you get the 3800 rpm?
You are right! Took out my Fluid Mechanics text and did a few back of the envelope calculations. It is significant.
If the disc is assumed smooth, and the flow fully turbulent, the drag loads could be near 45 W (@ 25k RPM).
So, you think building a tight enclosure would reduce the viscous drag loads? I’ll try that. I’m waiting on a current sensor so I can log current/power during the tests.
I think it will help reduce the drag but not stop it. It will allow the air to match the speed of the disk. Rather than be thrown out and then have to accelerate the replacement air. Again think of a vacuum cleaner when you block the hose the motor speeds up. Have you tried just adding a mark on the shaft and testing the RPM with that and a laser tacho?
Yup, I used a laser tachometer and my measurements match what I’m logging which is good. Might also help to eventually find a smaller, denser flywheel (steel?) to reduce the drag effect more. For example, halving its diameter drops the drag loads to near insignificance.
