Data glove --> throttle control and flight instrument

This is my latetest gadget, a "data glove" controls the eHayabusa and serves also as flight instrument. The OLED display is very bright and clearly visible from a distance. It shows the system status (System off, flight data) and does not disturb the hand movement in any way. Very handy.

Here is a short video of a test run to see how it functions. The system is still modular and can display any information provided: flight time, sensor information, heading etc. It is connected to the Ardunio by a I2C bus which works fine with 1m cable.

New battery position

The belly pack was nice, but has an inconviniency while starting. It changes the harness fit and limits the movement. So I finally decided to put the batteries in the housing. 2 sets of 8000mAh 12s Lipo cells fit perfectly and are secured with a velcro strap.
The batteries weight in total 4 kg so that the overall weight of the eHayabusa is from 4-6kg depending on batteries.

Controller change

I changed the controller from an old Arduino Uno to an Arduino Nano. There is not really a weight problem, but the smaller Nano fits better in the housing.
Here you can see the size difference:

The Arduino Nano controls now also the "data glove" flight instrument.

Heat managment (2)

To activily control the power output due to heat dissipation I installed 2 NTC sensors. One for the motor and one for the ESC. As the heat dissipation is slow, there has to be an offset for the cut-off temperature. I tried several configurations and about 45ºC gives good results for motor and for ESC. The real temperatur of the motor is about 80ºC when the sensors gets to 45ºC.
The motor is than powered down in a cooling mode, not completly shut-off. For about 2 minutes the motor can only run with low RPM and the small propellers realy cool down the motor.
Here is the photo of the motor rod with the ESC and NTC sensors.

Heat management (1)

The heat management of the motor is essential for an effective use of the energy. About 50% of the electric energy is wasted into useless heat. The heating of the motor and components reduces their efficiency and causes more loses (resistance).

There are mainly 3 components which dissipate the heat:

• motor
• ESC
• wires

My strategy is as follows:

• reduce the thinnest wires (ESC to motor) to a minimum
• increase wire diameter from battery to ESC to a maximum 
• have the ESC cooled externally
• have the motor cooled externally

The result is that I changed the battery cable to and placed the ESC on the motor rod. This reduces the wire length from ESC to motor and gives the ESC a cooling from the air flow of the propeller.

I also installed some additional propeller blades to cover the center of the 40" propeller.
The cooling of the motor was really bad before, as the direct driven motor does not get a good air flow from the 30" or 40" propeller.
With the additional 2 small blades (about 10", from a helicopter), I cover just the center of the propeller. Looks cool (and it is) and should give 500g more of thrust :-)
You can see the ESC (red circle) and the additional blades (green circle). The propeller is still well balanced after the change.

RPM measurement

Today I did some RPM measurement to calibrate the throttle control. As I have a Arduino to produce the input signal for the ESC, I can put it easily to any desired value and watch the RPM count from the Hall effect sensor.
I made the measurement with the 40" propeller and here is the result. The input signal varies from 0-179 and the RPM output has the expected curve. The max. RPM for this motor is 5800. At 5400 RPM I measured a thrust of about 13kg which is in accordance with the previous measured valued.
The output signal is now limited to 150 to not exceed this power setting.

Hall sensor for motor RPM

To know the RPM of the motor is a good thing. As the throttle control is done by a Arduino, I can use the RPM value to prevent the motor from overheating and adapt the throttle to different propeller configurations by software.
I made some research how to install a RPM sensor in a model motor. There are basically three posibilities:

• getting a signal right from the 3 wires of the synchronous motor --> that needs some soldering and signal processing
• putting a optical sensor for the propeller RPM --> possible, but too much components for electronics
• Hall effect sensor with magnets glue to the motor --> needs a least 2 additional mangets

I than realized that the motor has yet a lot of magnets inside. There are 40 of them ... so why do not use the Hall efect sensor with them?
This magnets are glued to the inner side of the moving motor housing. I have to get as close as I can to get a good signal. I tried some configurations with a standard Hall effect sensor and it worked.

The Arduino software checks now every 200ms the counts of the magnets and calculates the RPM.
Nice and simple. The value can be used to adapt the ESC signal to control motor speed.
The photo shows the position of the sensor, glued to the motor mounting. You can also see the motor mounting rod, made of carbon fibre. There are some wires left for 2 additional temperature sensors: one for the motor and one for the ESC.

 

Maiden flight

Today was the big day, the maiden flight of the ascending aid. I choose a small hillsite near Toledo (Magan, Cerro del Aguila) for this first test as the height is only 90m and I had to go back by walking with 15kg of equipment and 4kg of motor/battery.
The wind direction was perfect for this site, north-west with about 5-8 km/h. The test was conducted with the old battery set and the 30" propeller with 10kg thrust.
The ground handling is very easy, the battery pack does not disturb too much. The start procedure is like a normal paraglider. The motor is switched on when airborne to prevent any hazzle with the lines. The motor is also switched of before landing for the same reason.
I flew 2 nearly parallel tracks to compare the sinking speed with and without motor. You can see the details in the track log. The 10kg thrust gives me a ascending rate of about 1m/s. It could not compensate the 1,2 m/s sink rate at this moment, but it would have made the flight much longer if I wished.
Next time I will try the 40" propeller with 15kg thrust. In this configuration, a positve ascending rate should be possible.
Here is the link to the video :

And here the track log:

Arrival of new batteries

Today I received also a new battery pack. The new Turnigy Lipos are 15C rated and weight a little more than the first ones (only 10C). Now I have 6 x 8000mAh 6s batteries. Next test will be conducted to compare the performace of each one. With all batteries I should have a autonomy of about 20min full power. The battery weight is 6kg and the housing motor 2kgs as before. Here a photo of the two battery types:

New BEC

Today I got a new BEC to power the Arduino. The former setup was a smaller BEC rated 6s, which burned down after 5 minutes. Than I installed a 9V battery. with the new beter BEC I do not have to check for the voltage of the 9V battery. as far as there is a 6s 24V LiPo connected, the Arduino has power. Here is the photo of the old and new BEC (old, small one is already junk):

Ground test with video

In this viedo you can the the assembly of the eHayabusa to the standard harness Swing Connect Reverse. The test was conducted with the 30" propeller. Than I tried the 40", which gives >15kg of thrust. But in this configuration the heating of the wires is to much, after 60 seconds I stopped the test. I will install a new software version for the 40" propeller to limit time of full throttle to prevent any overheating.

https://www.youtube.com/watch?v=C8QBHXXR0wA

Continous run test for 6:00 minutes

One of the most important test was conducted today, the continous run test. I started full throttle with nearly full batteries to check the power consumption of the motor and the heat dissipation.

The result is very convincing: at full throttle with the 30" propeller I operated the motor for exactly 6:00 minutes. The cells were depleted afterwads to 22.7V that means 3.78V (LiPo 6S).

That is a fairly good value and there would be still some reseserve power. As the cells had seen some short tests before, I estimate the total operational time of about 7:30 minutes.

That is also more than initially expected.

The ESC has a shut-off controll for low voltage to protect the battery, so there is no problem of damaging it with to low voltage depletion.

Another important point is the temperature disipation. I checked with an infrared thermometer and the motor housing is getting to 78ºC at the end of the test. The power cables are also getting warm, around 40ºC. The battery itself only changed to 35º, also the electronics is not getting any warmer.

Static thrust on ground

I measured the static thrust on ground with a normal digital scale. The measured result was 10.1kg with the 30" propeller and unbelievable 14.9kg with the 40" propeller.

From the photo you can see that the axis has a inclination of about 25º. That means the real thrust in propeller direction is about 10% higher: 11kg for the 30" and 16,5kg for the 40".

That is far more than I expected for the 40".

Measurement setup with a sand bag to avoid the movement of the housing on the scale:

Propeller

The standard propeller for this electro  motor is a 30x8 inch plastic propeller. That is a diameter of 80cm. As the effectiveness of a motor depends on the propeller diameter, a larger one would be more apropriate.
For this reason I converted a 30 inch propeller with carbonfiber tubes to a 40 inch propeller, 101cm.
The thrust of the 30" should be about 120 N, increased with the 40" to 150N, nearly 15kg.
The propeller is well balanced and does not produce vibrations at higher RPMs.
Here is the comparative photo:

Next Steps

The first trial with full throttle was successfull. For about 30 seconds I switched on the motor mounted in the Swing Connect Reverse harness. The thrust is about 10kg as expected, an exact measurement will be done later.
But I have to modified the electronic box. Right now the control (an Arduino Uno board) is placed in an experimental plastic case. The wiring has to be more robust to prevent the cables to get loose.
In the next days I expect the new components and switches.
I will also replace the 9V battery as power source for the Arduino board with a BEC connected to one of the main batteries.

Throttle control

The throttle control is a modified comercial throttle. Unless a gasoline powered paraglider, the throttle is not operated with the fingers. The throttle is controled with the thumb which is far more effective and allows to use the throttle grip as brake handle.
The throttle/brake can be mounted parallel to the normal brakes. So in normal flying conditions you use the normal glider brakes and when you change to motor-mode you simply grap the throttle/brake handle and push the throttle with the thumb.

Her some pictures of the throttle before and after attaching it to the brake handles:

Battery pack

The battery pack is stored in a Lipo-Save bag for safety reasons. The bag is attached with velcros on a sports vest. The battery pack weights only about 2kg in the basic configuration.
If a problem with the battery occurs in flight (overheating/fire), it can be easily desconnected with a handle and thrown away (no mini-parachute installed yet).

Physics behind the electrical ascending aid

First of all, what power do I need to keep a paraglider in the air?
From a very simple model, we can estimate the power we need to keep us in the air (without ascending or descending).
If my paraglider has a sink rate of about 1m/s and weights 110kg (weight force: about 1100N), than I need a power P of : P = F x v to keep me in the air.
That means only about 1100W. That is independant of the power source. With an propeller and a efficiency of about 50%, we need 2200W on the propeller to keep us up. Every additional Watt would give me a lift.

Second, what power do I need to get me in the air from ground?
That is far more complicated. The moment with most power consumption is the phase when the paraglider is just behind you at the start. This is the "parachute" moment. The full canopy is open, but not above you, it's behind you and causes a tremendous drag.
The power of the paramotor needs to compensates this drag force and needs to deliver a push force of 500-800N (50-80kg) in a normal condition.
Just for a few seconds, but without this power the canopy will not rise.
As a foot starter you deliver also a lot of power to overcome this drag with your feet. In a trike configuration, the motor has to due all the work and needs even more power.

What is the best canopy?
Normally a reflex canopy is used for paramotors, because it is more stable at higher velocities. But the reflex has a drawback, it has a lower glide ratio as a standard paraglider. The glide ratio defines also the vertical sink speed at a given horizontal speed.
Just an example: the glide ratio of a reflex canopy is half the glide ratio of a normal canopy. That means roughly that the glider has double the sink speed. That means we need to double the power to maintain us in the air.
So the best choice is a standard canopy. We can't fly to fast, but the goal is to reach the next thermic, nothing else.

Picture of prototype

This is the design/prototype of the eHayabusa ascending aid for Swing Connect Reverse harness:

Project description

This a my blog about the project "eHayabusa", a lightweight electric paramotor system ment as ascending aid.

Project philosophy:
The eHayabusa is no "full" electric paramotor for a simple reason: the technology and cost to achieve a full electrical paramotor are not feasible until now.
Nevertheless, an electric ascending aid is possible and economical.
The main difference between an electric paramotor and an ascending aid is the power and the maximum time of power deliverance.
With an ascending aid it is more difficult to make a foot start from even ground, but once in the air, it has nearly the power of an traditional paramotor.
How is it possible that such a low power motor can fly a paraglider? Finally it has only 12kg of thrust. We will come to this point later on.

Project data:

• Weight: basis version only 4kg
• Push force of system: about 120 N 
• Battery capacity: 5 min of full throttle
• Maximum height gain: 150m 
• Assembly time:  < 5min, ready to start
• TOW: <110kg

The design is specially fit to my harness, Swing Connect Reverse, but it's adaptable to nearly all standard harnesses.