Cheap DIY Brushed Motor "Dyno"

[Arise0185]

Background: Trying to eek out every last bit of performance from a stock Traxxas Slash (brushed 2wd) for Spec Slash racing.

I've been mulling over how to determine which motor is the "best" one to run out of a batch of motors that I have on hand. For context, we have to run the stock Traxxas 12T Titan motor, which is a sealed unit. There are some things you can do to break them in, lube them, etc, but for the most part you just buy one and hope for the best.

I know there are some expensive motor dynos that are available that utilize a heavy flywheel to act as a load, and you can measure how quickly the flywheel spools up, etc. But those cost hundreds of dollars and probably many hours of tinkering/learning to understand how to even utilize/interpret the data.

I was wondering if it might be useful to create a much simpler, cheaper "dyno" that consists of an RPM meter ($20 on amazon for an optical tachometer). My thought is that I could run a motor with no load at various fixed throttle inputs (e.g. one "low" throttle input with the throttle trim turned all the way up, and one "max" throttle input with the trigger pulled all the way in). The more efficient the motor, the higher its RPM would be at each of the above conditions. And hopefully that would translate over to the motor's performance under load. Maybe I could measure current draw as well.

Not sure how much validity there is in my reasoning here, but the tachometer is cheap enough I might just go for it anyway

 

[Tex Koder]

Background: Trying to eek out every last bit of performance from a stock Traxxas Slash (brushed 2wd) for Spec Slash racing.

I've been mulling over how to determine which motor is the "best" one to run out of a batch of motors that I have on hand. For context, we have to run the stock Traxxas 12T Titan motor, which is a sealed unit. There are some things you can do to break them in, lube them, etc, but for the most part you just buy one and hope for the best.

I know there are some expensive motor dynos that are available that utilize a heavy flywheel to act as a load, and you can measure how quickly the flywheel spools up, etc. But those cost hundreds of dollars and probably many hours of tinkering/learning to understand how to even utilize/interpret the data.

I was wondering if it might be useful to create a much simpler, cheaper "dyno" that consists of an RPM meter ($20 on amazon for an optical tachometer). My thought is that I could run a motor with no load at various fixed throttle inputs (e.g. one "low" throttle input with the throttle trim turned all the way up, and one "max" throttle input with the trigger pulled all the way in). The more efficient the motor, the higher its RPM would be at each of the above conditions. And hopefully that would translate over to the motor's performance under load. Maybe I could measure current draw as well.

Not sure how much validity there is in my reasoning here, but the tachometer is cheap enough I might just go for it anyway

You need to take into consideration the friction of a "simple" transmission and at least one wheel with some sort of load -to accurately gauge the performance of a motor..
Having electric motors spin freely.. will not give a realistic reading.

 

[Hector_Fisher]

Agree with Koder, you'd need a consistent load of some sort to accurately portray the right conditions and draw any meaningful conclusion. I've seen folks use a prop or some sort of prop assembly to provide constant force (add prop guards if you do!).

To remove voltage fluctuations as a variable, you'd do well to use a fixed power supply (many chargers have this as a mode) rather than a lipo to eliminate any variablility in lipo charge and voltage drop from pack to pack, or for one pack across time.

But overall, yeah I think the system could work as a rudimentary motor dyno. Under the same load, i'd be looking for motors that run higher RPM than the others and/or run cooler when you shoot it with an IR gun. Those would be my indicators for a more efficient motor.

Regarding your "hours of tinkering" comment to interpret data, I think you'll still be faced with that, regardless of the price of your dyno, though

 

[Arise0185]

You need to take into consideration the friction of a "simple" transmission and at least one wheel with some sort of load -to accurately gauge the performance of a motor..
Having electric motors spin freely.. will not give a realistic reading.

Agree with Koder, you'd need a consistent load of some sort to accurately portray the right conditions and draw any meaningful conclusion. I've seen folks use a prop or some sort of prop assembly to provide constant force (add prop guards if you do!).

To remove voltage fluctuations as a variable, you'd do well to use a fixed power supply (many chargers have this as a mode) rather than a lipo to eliminate any variablility in lipo charge and voltage drop from pack to pack, or for one pack across time.

But overall, yeah I think the system could work as a rudimentary motor dyno. Under the same load, i'd be looking for motors that run higher RPM than the others and/or run cooler when you shoot it with an IR gun. Those would be my indicators for a more efficient motor.

Regarding your "hours of tinkering" comment to interpret data, I think you'll still be faced with that, regardless of the price of your dyno, though

I had a similar thought regarding the DC power supply to eliminate variations in battery voltage. And my thoughts exactly regarding using RPM as the indicator of efficiency. I hadn't thought to check temp but that's a good idea as well.


Prop seems like a good/simple idea to add a constant load. I guess there would be some variability in the load depending on air temp/density, but probably not enough to battery given the climate controlled interior of our house.

That did get me to thinking about some kind of submerged water wheel type device, however. That might enable an even greater load with more(?) consistency, albeit with more design work on my end.

 

[Arise0185]

"Water brake" is the term I was looking for in my previous post. Assuming I can design and 3D print a simple turbine/water wheel thing, I think I could submerge it in a glass jar filled with water. Then, as long as the water level and turbine design is kept consistent, it should provide a consistent load.

Regarding your "hours of tinkering" comment to interpret data, I think you'll still be faced with that, regardless of the price of your dyno, though

Specifically responding to this comment... yes this will devolve into something requiring many more man-hours than simply buying one off the shelf, I'm sure But hey, that's part of the hobby, right?

As a simple unloaded test, I checked an unprepped motor against a "prepped" motor (comm cleaner/polisher). The latter has an audibly higher unloaded RPM at max throttle. We'll see this Friday if that makes a noticeable difference on the track.

 

[Arise0185]

Quick water brake dyno I modeled up. The idea is that I would bolt the motor to the top plate, use a shaft coupler to connect the motor to the shaft-driven "turbine," then apply varying voltages to the motor and record the RPM and current draw.

The shaft is 5mm in diameter and is a few bucks from McMaster. The bearings are Traxxas 5116 (front wheel bearings on 2WD Slash). The rest of the parts would be 3D printed. Not sure how water tight the 3D printed can will be, but I'm sure there are surface treatments I can apply if needed.

There are various knobs I could turn to tune how much resistance this thing provides, like turbine diameter, depth of blades, fluid viscosity, etc. I could also have added gearing but wanted to keep it simple to start.

Now what this WON'T tell me is actual power output of the motor. To calculate power you need RPM and torque, but I don't know of a cheap/easy method of measuring the torque at the output shaft. So this will just be a way to compare different motors against each other, NOT to provide absolute data.

 

[rgfast]

I think you would end up with to much cavitation with that design, to create load a pump impeller with inlet/outlet allowing water to flow would serve you better
Bobby Allison's first engine dyno was loaded by using a variable pitch airplane propeller

 

[Arise0185]

I think you would end up with to much cavitation with that design, to create load a pump impeller with inlet/outlet allowing water to flow would serve you better
Bobby Allison's first engine dyno was loaded by using a variable pitch airplane propeller

I was concerned about that... but also I was hoping to reduce the need for continuously flowing water.

I wonder about adding gearing to slow down the turbine RPM, as well as using a more viscous fluid, like maybe mineral oil? I liked the idea of tap water because its viscosity is relatively repeatable/consistent. Something like mineral oil could change in viscosity depending on what brand you buy, etc. Maybe something like gear oil and sticking with the same MFG and weight could do the trick.

 

[Hector_Fisher]

I was concerned about that... but also I was hoping to reduce the need for continuously flowing water.

I wonder about adding gearing to slow down the turbine RPM, as well as using a more viscous fluid, like maybe mineral oil? I liked the idea of tap water because its viscosity is relatively repeatable/consistent. Something like mineral oil could change in viscosity depending on what brand you buy, etc. Maybe something like gear oil and sticking with the same MFG and weight could do the trick.

Problem is, as @rgfast alluded to, power is only shown when work is accomplished. Meaning, if you aren't moving fluid, you aren't really demonstrating power. As an extreme example, if the fluid is so viscous that it sticks to your impellers or paddles and doesn't really move, then all you'll be doing is creating a merry-go-round for syrup. This is why air is such a good candidate because it's viscosity and density are so low that due to turbulence it will very quickly lose it's momentum, requiring alot of power input to keep it moving.

If, for example, you used something with the physical properties of maple syrup (thick but not too much) and geared the motor way down, you could exploit gravity pulling the syrup down quickly enough and "stir" the syrup. Sounds like an interesting fluid dynamics problem for sure!