Please post your photos and thoughts on motor test stands and ship test stands.

Views: 1960

Reply to This

Replies to This Discussion

This is the current form of the test stand used for motor tests.  The components are:

o a frame for holding the thrust post (in this case extruded aluminum machine frame)

o a thrust post that translates rotor thrust to the scale and is at least 2x prop diameter long to avoid ground effect bias  (something light so the scale can tare with room for thrust)

o a soft connection device for lightly holding the thrust post to the frame (once the motor starts to spin, the rotor and thrust post will hold itself so there is a 6mm gap between the post and structure loosely held by an elastomeric speaker connector with a 4mm thick sorbothane washer between).

o a platform at the top to bolt down the motor mount (at least three places)

o a universal motor mount so fit to the base on any motor and made of highly thermal conductive material (in this case, aluminum)

o a thermocouple mounted to the motor mount to measure the temperature of the motor.

o a thermocouple display to show the temperature

o a vibration sensor connected to the top of the thrust post to ensure that the props are adequately balanced for the test.

o a vibration meter (in this case an Extech SDL800)

o velcro at the bottom of the thrust post connected to the top of the scale to counter rotor torque

o a gram scale accurate to 1 grams that can be tared (in this case a Polder 5kg x 1g KSC-350-90)

o a hard base for the scale (on carpet so needed a nylon slab)

o scale position markings so the thrust post stays perpendicular

o a sound level meter that calculates dba or dbc always placed 1m from the rotor (in this case a Radio Shack hand held) 

o a hand held laser thermometer for sensing the temperatures inside the motor or on the housing

o a watt meter afixed close to the scale reading for near simultaneous visual reading (in this case a Hobby King HK-010)

o power cord adapter going from the watt meter to the motor

o a DC power supply to precisely control voltage (motor efficiency varies significantly with voltage)

o a power cord going from the power supply to the watt meter

o a switch on the power cord capable of handling the amperage

o a transmitter 

o a receiver & power supply for connecting the ESC to throttle (this varies and is a topic in itself)

o a computer for recording the data

o a plexiglass plate separating the the computer from the test area

o a soft cover for the edge of the plexiglass plate (in this case foam pipe insulation)

o safety glasses hanging next to the computer for ease of access

o a first aid kit with anticeptic, large gauze and tape nearby

o a phone within reach for calling for help

o tools nearby for changing motors and props

What is lacking the most in this setup? A computer interface that reads in real time, the thrust, watts, dB, and temperature into Excel (I'd settle for the first two).  When testing a lot, this would save so much time and be a bit more accurate.

The test procedure is as important as the test equipment.

When taking energy (watts or watts, amps, and voltage) and thrust data:

o To compare the results of different rotors (motor + prop), calculate net thrust per watt.  For copters, the rotor has to lift itself first.  Net thrust (I have an excel worksheet that does this) is thrust read on the tared scale less:

  - motor weight

  - prop weight

  - prop mount & screws

  - motor mount & screws

  - structure

o To compare the results of different props diameters, structure needs to be taken into account.  I use a structure penalty of 2 grams per inch over 10 inches (so an 8" prop would receive a -4g structure weight and a 17" prop would receive a 14g structure weight)

o Record the voltage and keep it constant between tests (voltage matters and can bias results by more than 3%)

o Measure energy at various thrust levels as rotor efficiency varies (there is a curved peak in the middle of the thrust efficiency curve; it does not peak at maximum rpm as thought by many)

  - estimated hover thrust per motor at AUW

  - 80% hover

  - 1.2x of hover

  - 1.5x hover

  - 2x hover

  - max thrust

o balance props axial and in trace before beginning and check the result with a vibration meter before starting

o tare the scale, ramp up to full thrust for a few seconds and return to zero.  Check tare as the wires will settle.

o when testing a new prop, record noise level at hover and elsewhere as needed.

o record temperature

o keep you eye on the thermometer (I burned up a Tiger Motor that belonged to Hugues by not paying close enough attention)

o when comparing different motors, always use the same ESC.  If you change ESC settings, note those changes.

o label and record a prop number.  CW & CCW props will perform typically around 2 - 7% different.  Different props off of the same mold will perform typically up to 4% different or worse.

o Oh ... per Hugues recommendation, I added a fire extinguisher near by.  

This the the ship test stand in its current form.  Shown are two ship and the two flight locations.

o It can be wood 2x4s (also not as hard as aluminum).

o It doubles as the structure to support the motor test stand on the right

o Place above for flight tests.

o Place below for initial tests.

o Structure line with pool noodles so props don't break

o Ship loosely held in place with thin bungee cords that can be affixed to multiple ship points for initial tests or at dead center for PID tuning (so it flies well enough for auto-tune and for fine tuning after auto-tune). 

o Try to keep the upper platform easily to modify if you plan to use/build different shape ships.

o Think where the props will go as the ship rolls, pitches, yaws and climbs.

o Make it oversized

Most important on use.  Always ramp up slowly.  The copter can get out of control quickly.  When held by bungees the spring affect can make out-of-control worse. At a minimum:

o Always wear protective eye wear.

o Wear thick leather gloves when in range of possible copter paths.

o Wear thick protective clothing.

o Wear full coverage shoes

o Always have your hand on the throttle

o Always have a dead-man switch in reach as throttle does not always remove power after a crash.

o Never get close to an armed copter.

This is a great test bench. I wish I could steal a room in my house to setup something like that (with a fire extinguisher of course).

Were you able to test a XOAR propeller ? These are apparently excellent wood props, prebalanced out of factory and as light as CF props!

I must replace the current APC 15x4 props I installed on the newly assembled X8 octo with MT3515-400KV motors because I get awful vibrations, even after balancing the props. I get twice as much vibrations on the IMU of APM (in the range of +3/-3 while I was at +1.5/-1.5 with the previous MN3110-700KV and 12inch props). Using the same APM vibraiton isolation (Kyosho zeal equivalent from 3M).

I'm pretty sure this is caused by these APM props. I see with naked eye the non perfectly flat surface of the moulded plastic. These props have also a very very thick hub ring that I can't seem to balance. I will just throw them away.

Obviously it is also not abnormal to get higer vibrations with bigger props, but I did not think the increase from 12inch to 15 inch would double my vibrations on APM !

 

That's a lot of interesting stuff.

Given perfectly balanced props, does a larger prop increase vibration?  I've never done that study.

 - they turn slower so that could mean less vibration

XROAR Props - I've only tested two:  the 11x5 and the 12x6.  They are nice, but not great.  As a comparison:

o 6.27 XROR 12x6

o 6.53 GWS EP HD

o 6.38 GemFan ECF Carbon

o 6.36 APC 12x3.8 APC Slowfly

The other problem with the thick hub ring is that it requires a heavier prop mount.  So far I love the DJI carbon props as they are quite durable and inexpensive.  But you could also go with the Tiger 15" or 16" or 17" as they are also excellent but extremely expensive.

There is also the question of vibration frequency: it seems like I get lower frequency vibrations with a combination of lower KV motors and bigger propellers.

This 3M self adhesive moon gel like tape I am using now does not block low frequency vibrations very well. I should try the real kyosho zeal tape.

You have inspired me to setup a small testbed to measure vibrations generated by different propellers.

I am thinking of measuring like so:

-I will use a well balanced reserve MT3515 motor I have on stock

-it will be fixed on a aluminum arm. It will be long enough to position a smartphone on the other end so that the propeller does not blow on it.

-I will run a seismic application on the smartphone that gives a frequency spectrum analysis on three axis.

-the measurements will be influenced by the supporting base of the arm. I might try to use Velcro beneath the arm to hold the arm without dampening the vibrations to measure.

-I will take different throttle measurement points for each propeller. Each point will be identified with a power meter. 

-result: a main identified vibration frequency for each type of propeller, at different RPM

Does it look to be a correct measurement procedure to you?

Yes

o mass is the primary dampener so a light Al arm (rectangular or square)

o power isn't the equalizer if you are testing different props.  thrust is.  So test at different net thrusts.  Define different net thrusts (e.g., 200, 400, 800, 1600, 3200 grams).  Then add the weight of the motor, prop, prop attachment, motor attachment, and 2 grams for every inch the prop is over 10".  So a 15" rotor weighing 100 grams would be tested at the following thrust levels:  330, 530, 930, ... grams

o fly the rotor so the force is applied towards the scale.

o to avoid ground effects on the props and scale, the motor is best mounted to the end of a rectangular one meter light rod with the smartphone mounted vertical just below the rotor (in close like that the phone will not receive significant prop wash).  velcro the bottom end of the rod to the scale.

o test the smartphone first to see if it gives you want you want.  place the phone on top of a foam pad on top of your computer speaker.  online, find a frequency generator.  turn on 200 Hz.  See what the phone says.  the accelerometers in the phone may or may not have enough throughput to identify frequency.

Some test results about our prop discussion:

I tested the same craft, MT3515-400KV motors, same all up weight, same APM vibration dampening with two sets of propellers:

-APC 15x4 electric : i used to like the APC slow fly I had used in 10x4.7, 11x4.7 and 12x3.8 formats. Thay gave me previously good results with the MN3110-700Kv motors. Now with these 15x4 I measure vibrations at minimum +3/-3 which is awful. I balanced them as good as I could, it did not change a bit the level of vibrations.

weight of these APC props : 38g

-XOAR wood props 14x5, weight 27g (10g lighter!). Reduced the vibrations in half : +1,5/-1,5

I love these props, not only are they efficient but also they are nice looking!

Wood props are gorgeous ... as are wood canoes and kayaks.  Love the craftmanship.

Will be watching for the carbon results.  I'm typically pleased to see 0,3g vibration at about 280 net grams of thrust.

Hello Forrest

with great interst I follow this forum and learnt a lot! Sorry for my english.

My question where do I find your excel file for the motor calculation.

I have some HP2212 1000kv from rctimer left from my quad and i like to fix them on a v ship.

did you test this motors as well or calculateed? are they suffizient to use for a v octa?

I used up till now propellers 9x4.7 and 30 A esc.

Thx for your reply in advance

Regrds martin

I updated the Custom Octa CAD worksheet for the V tab. So the V tab is good. But I haven't checked the other tabs. If you have any questions, let me know. The worksheet has macros that are safe, so OK to enable.

You will notice that the prop rotation is different than what is stated on the DIY website. That is because this particular prop spin layout solves to a better solution (otherwise all of the roll factors are the same and the roll factors are different, which is less than optimal).

Attachments:

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service