Self Built RC DYNO

[henzy190]

What's "tut"?

short for tutorial.

to be honest, what you've done there is a cracker. I could do with a dyno for an upcoming "project"

and the software is cool too

[joeqsmith49]

It's a bit premature for a tutorial.

For anyone wanting to build something like this, here is a list of things I have learned.

1) Don't rush it! I built this thing in short order, not really thinking about the problems I would run into. Start with a list of goals. What your wanting it to do and why. If the goal is just to add a small load for motor breakin, that's easier than trying to accuratly measure the power curve for these small motors.

2) Pick a better absorber, or make one from scratch. The rear support is just not stable enough. I had some ideas on what I wanted to measure so I sorted motors to find something in this range. The motor's bearings, the fact it has brushes, balance, etc all make this a bad idea. Designing something to do the job right would be better time spent.

3) Use better construction practices. While I knew the aluminum would warp during welding, I had left the bearing holes 10 under. After welding, I then borded them with a reamer. This is fine, but the small amount of warp in the base causes drag when it is clamped to a flat surface. IMO, don't weld. Machine the bearing support from one piece and then bolt it to the base.

4) Use C -clips to hold the bearings. The setscrews cause the bearings to distort and add drag. It does not take much.

5) Make sure that the bearings you use are up to the task. This dyno was made from scraps. This included the bearings that came from a small stepper motor I had tore apart. This was a poor choice.

6) With a 6" torque arm, make sure the shafts are near perfect. Any errors here are just multiplied by the arm length.

7) Figure out a way to balance the rotating parts. I have some ideas, but nothing proven.

8) I had tried to keep the rotating weight down. A flywheel may help to reduce vibrations.

9) The coil type of coupler I used may not be the best choice. Any missalinment adds drag.

10) Design a better mount for the test motor. The block is OK, but imagine drilling holes for all your motors. Maybe something with micrometers to fine tune the XY&Z.

11) Think about adding a closed loop control system. This system would have it, but I think it has enough problems that it does not warrent any more time spent on it.

For those building your own, have fun and good luck.

[joeqsmith49]

After some searching, I found this website:Depac Dynamometer Info

They have some very good ideas I had not considered when building mine.

Some wood + glue + screws + nuts + bolts + washers + paint = new base. Beats the brick. A hole was drilled in the end to allow use of a smaller C clamp to hold it to the table. There are nuts sunk in to the block, so there are two studs sticking out. Then two nuts hold the dyno to the base.

The torque arm are was turned around so the harnesses don't sit inside the scale.

Here is the last data I took with the 6V VCR tape drive motor running at 6, 12,18 & 24V. I had started with 6 and adjusted the absorber for each run. This is why the data doesn't overlay very well.

Now that the system can make a pull using the computer, the green motor was installed. It smoked last time with too much load for too long a time. This time the the power supply was adjusted to 3V and just enough load was added to find the peak power.

There are a few more problems with the setup that need to be corrected. The power supplies compensation lines need to be connected to the motor or the cable sizes would have to be increased. At 20 Watts out, I need to use a larger heat sink for the absorber's FET drive. It may not hurt to come up with some sort of air duct system for the motor and absorber as well. The pulls are fairly quick, but I have to let the green motor cool between pulls. At least no smoke this time..

[joeqsmith49]

This is the new DC-DC converter used to drive the motor as well as the current control for the absorber. Both are controlled by the computer.

[joeqsmith49]

I now have the dedicated electronics controls setup. It's fast enough now that I started to work out some of the control software.

Here the green motor is being speed controlled. So, you set some voltage to the motor and give it a speed and the controls will set the load to reach that speed. The system uses a simple PID control algorithm. The following shows the RPM as I step to different speeds. The settling time is changing as I tune the controls.

[joeqsmith49]

Work on the dyno has been slow. I have added torque control along with the speed control. It's turning into a decent little system. The software is finished for the most part. You set the mode you want to run. Give it a start, stop and number of steps for the motor voltage. Next set the start, stop and number of steps for the torque (example for closed loop torque control) and press go. The system starts running all the sweeps while you go have a coffee.

A new DC-DC needs to be built next. I picked up a new core to build a higher power system and now have a 1KW supply to drive it with. Obviously, the dyno will not be good for motors this large (unless they are very inefficient), but I should be able to do a good job mapping the green hobby motor once this is done.

I wrote Land & Sea after seeing that they had a small S type load cell listed. They are asking $900 for the sensor. Way outside my budget for this toy dyno so I'll stay with using the balance for now.

[joeqsmith49]

Work on the dyno has been slow. I have added torque control along with the speed control. It's turning into a decent little system. The software is finished for the most part. You set the mode you want to run. Give it a start, stop and number of steps for the motor voltage. Next set the start, stop and number of steps for the torque (example for closed loop torque control) and press go. The system starts running all the sweeps while you go have a coffee.

A new DC-DC needs to be built next. I picked up a new core to build a higher power system and now have a 1KW supply to drive it with. Obviously, the dyno will not be good for motors this large (unless they are very inefficient), but I should be able to do a good job mapping the green hobby motor once this is done.

I wrote Land & Sea after seeing that they had a small S type load cell listed. They are asking $900 for the sensor. Way outside my budget for this toy dyno so I'll stay with using the balance for now.

[joeqsmith49]

I ended up with a VICOR PFC Mini chassis with a 500W module that can be programmed directly. So the new DC-DC was scrapped.

On top is the new load for the absorber. This is a linear MOSFET design like the previous, but can handle much higher currents.

Below it is the VICOR power supply. It's wired with a couple of feet of 12, using the sense wires. 40A short runs are no problem with this setup. I had assumed that the motor's maximum voltage was near 5 volts. They spec 50% eff @ 14K RPM and 67 Watts at 10,700 RPM. Roughly guessing then the motor needs around 27 Amps.

On the bottom is the main computer and electronics.

To the right there is now a blower for the test motor and much larger clip leads to connect power to the motor.

*

The first pull with the new setup, the Green motor put out about 30 Watts. I let it cool down and ran another sweep with more aggressive settings. This time the motor put out about 40 Watts. The dyno seemed to be holding the added power just fine so I increased the supply voltage to the motor to and made one last pull just to see if the Green Monster really would put out 67 Watts.

The pull started out normal, as the load was being added I watched as the power went over 50, then 60, then BANG and I threw the breaker. The data later showed it maxing out at 63.52 Watts.

The following shows what's left of the generator...

[joeqsmith49]

Anyone in the states know a low cost R4 ceramic? Price looks about $60/ea.

[joeqsmith49]

3-phase absorber.

Thompson mounts with hybrid dry ceramic bearings.

[joeqsmith49]

I machined up the new motor mount. It's much thicker than the original and has ears (looks like a top hat) so it will bolt to the plate from the top rather than the bottom like the original unit.

The torque arm is mounted to the absorber with 4 bolts rather than 2.

There are no shims in this design. The end of the coupler was cut to form the shim and the outside edge was remove to allow clearance in the Thompson block. Again, magnets were mounted inside the coupler on the output side. There is only one drive to connect the test motor to the coupler. Made from brass, you can see it attached to the test motor. Hoping less parts = better balance and lower run out.

The other bracket is for the speed sensor.

I need to make up a new arm stop. This will be the same as the old setup (about the only thing that worked).

The base will be made from 1/2" steel this time to prevent warping when bolting down. The original dyno would bind (you could hear the pitch change) as the plate was bolted to the block. The Green motor puts out enough power that these errors are very small. With the VCR motor, all the small things come into play.

I used a synthetic lube for the bearings. Just put them into a canning jar with the lube and pulled it down with a pump. The lube is pretty thick, but the total drag even with 5 bearings is still less than the original dyno.

I put 4 amps through all the windings for a couple of minutes and the heat rise was very little. For short pulls, a fan should be good enough to cool it.

Connected the VCR motor to the absorber and it seems to work well while just holding the parts in my hand and shorting the windings. I tried the Green Machine motor and guessing I could pull it down to 100 RPM. Well into the smoke test area.

A ground shaft will be inserted into the three blocks to align the parts on the base plate before mounting the absorber and test motor.

[joeqsmith49]

It's starting to look like something. Here is a picture of the base bade from 5/8" thick cold steel. Top and bottom surfaces are ground to get the flatness. This unit will be

The ceramic bearings do not allow any sort of path for static buildup and a high strand wire was added for this.

I have a friend who does a lot of woodworking who wants to make a new wooden base for it.

I have been thinking of making a little video (strange this topic was started under the video section) of it but not sure what I could show that would be helpful to anyone wanting to build such a thing. It's sort of a bore to watch an electric motor spin. I saw a few videos of people making similar systems but they don't directly read torque for some reason (cost??), so I am not sure what they are doing. The absorber, like the motor has it's own curves. The closed loop controls are nice for this..

Next picture, it should be finished.

[joeqsmith49]

Mounted all the parts then tried it out with the VCR tape motor and found the speed pickup needs to move. I made a few pulls under a couple of Watts and it seems very smooth. The ceramic bearings have next to no drag so I can read much lower power levels now.

I installed the Green Machine and made a few manual pulls under 10 Watts. It's was very smooth even at 18,000 RPM. It was working so good I decided to see what I could get out of the little motor. Last time the old dyno was damaged at 63 Watts. This time I had it to 86 Watts and just held it there until the Green motor started to overheat even with the blower right on it. The absorber stayed nice and cold.

I'll make up a new mount for the speed sensor and a friend who does a lot of woodworking offered to make a nice base for it. By next week I'll finish it up and post some sort of video of it.

[joeqsmith49]

Here is my final version of the eddy current dynamometer. I used a chemical gun blue on the base plate to help prevent oxidation. I did not want to screw up the grinding with paint. The base has brass inserts fitted and studs that are bolted from the bottom into the base.

I made a little video of the Green Machine being tested: http://www.youtube.com/watch?v=aWUfDp5x7ek

Good luck with your dyno projects.

[joeqsmith49]

A second video showing the response time while modulating the absorbers output power.

http://www.youtube.com/watch?v=muZUoeE88jw&feature=related

Edited April 30, 2011 by joeqsmith49

[joeqsmith49]

Added small compartment to hold a few tools at the ground shaft used to align the unit.

[DamoRC]

Joe - have to say, excellent job! It's a really nice looking piece of equipment.

Do you mind me asking what kind of environment you built this in , industrial, an academic lab perhaps? It seems that you have access to really high end machining and electronics capabilities (not to mention a license for the NL Labview software).

DamoRC

[joeqsmith49]

Joe - have to say, excellent job! It's a really nice looking piece of equipment.

Do you mind me asking what kind of environment you built this in , industrial, an academic lab perhaps? It seems that you have access to really high end machining and electronics capabilities (not to mention a license for the NL Labview software).

DamoRC

Thanks. This is just a little home project.

A few years back I helped a friend with a small water brake he was building. This was a 75KW or so, direct drive setup, mostly used for small car motor break in. Places I have worked at have always used direct drive systems. Mainly these have been eddy current types and most could do some basic controls. Some were motoring type. These were rated for much higher power levels. I have been thinking about building my own large dyno for some time.

I would like to do more with the model dyno but I don't like risking the balance. My plan is to replace it with an LVDT or linear pot. Below is my very low cost torque sensor. Linear pot with a spring. 100 grams is about half travel. So 200 grams on a 6" arm. Absorber won't handle that sort of power. I have not had time to try it out yet.

[joeqsmith49]

Using high grade scale to make up some calibration weights for the new torque sensor.

Edited March 5, 2012 by joeqsmith49

[joeqsmith49]

Custom clevis, base, spring, sensor and calibration weights.

[joeqsmith49]

The assembled unit. It's not going to be near as good as the balance for accuracy but it does not cost much and can be read much faster. I plan to use the weights to do a 4 point curve fit. The spring should be fairly linear over the short range.

[joeqsmith49]

I have a few different springs that I can use to set the torque range. The small VCR motor will be a good test to see how sensitive it is. My plan is to benchmark it against the balance.

The new software can run the dyno in speed, torque and voltage control modes. There are seperate tuning parameters for each mode. There are various ways to display the data now as well as some simple post processing.

[joeqsmith49]

Made some progess with the linear pot. I wrote some software for it and tried it out. It's about as good as you would expect for the price.

I need to get a better test motor that can make about 100 Watts of continious power. The little Green Machine has been over heated a few times now.

[joeqsmith49]

Showing the entire toy dyno with calibration standards and test motors.

[joeqsmith49]

This short video shows how to calibrate the dyno. Basically remove the linear pot and start putting the weights on. The software walks you through the procedure. You can use as many weights as you want.

http://www.youtube.com/watch?v=Fm8MrdZ8cUQ

Edited March 7, 2012 by joeqsmith49