Conversion to quadrotor (parts)

After having struggled to keep the vibration caused by two of the motors on the hexrotor down I decided to convert it to a quad, thus eliminating the two bad motors. Luckily the website that sells my hexrotor frame (www.cnchelicopter.com) also sells other frame configurations that use the same arms. This means that I am able to convert my hexrotor frame to a quadrotor frame simply by changing the center plates.
Because I was already paying for shipping I decided to pick up several other spare parts along with the body plates.
To complete the conversion I needed to reprogram my CRIUS SE. This was a problem as although I had an FTDI programmer, the FTDI power supply was shorted with the main rail (see First flight). This meant that I would have to reprogram the board via the ISP (in system programming) port. I found a USBasp AVR programmer for sale at www.hobbyking.com and, as with my other purchase, decided to order some additional parts along with it as the shipping was fixed.

 

My purchases from cnchelicopter:

XAircraft X450P Upper Plate

XAircraft X450P Lower Plate

XAircraft X450P Arms (one damaged a crash)

XAircraft X450P Feet

Xaircraft Hexa Upper Plate (cracked in a crash)

XAircraft X450P Motor Mounting Plate (one damaged in a crash)

XAircraft X450P Guide Ball Mounting Rod (current rod is not long enough to span the wider 90 degree angle of the quad arms)

Xaircraft Hexa Frame Electronics Mounting Plate (my current one has a GoPro mount attached to it that I can’t remove. Although designed for the hex frame, it fits on the quad and is larger than the quad version)

 

My purchases from hobbyking:

USBasp AVR Programmer

5 XT60 Female Connectors

4mm Banana plug to 6 x Male XT60 in parallel (because hobbyking was currently out of stock of their 5-piece bags of male XT60 connectors I decided to buy this adapter and cut the XT60s off)

Twin pack charge lead (2 x 3S)6S w/ XT60

Flat 26AWG servo wire 1mtr (R/B/W)

Turnigy Pure-Silicone Wire 14AWG (1mtr) RED

Turnigy Pure-Silicone Wire 14AWG (1mtr) BLACK

Turnigy 5mm Heat Shrink Tube – RED (1mtr)

Turnigy 5mm Heat Shrink Tube – BLACK (1mtr)

Advertisements

First flight

After having charged the batteries for both my hexrotor and transmitter I went out to fly and discovered several things:

1. The default PID settings worked well although there was considerable wobble (addressed below). Here’s a screenshot of the settings in the MultiWii Config menu:

2. The wobble was caused by the severe vibration of one of the six motors. The sensors on the board (accelerometer and gyroscope) were getting false movement signals as a result of the vibration, thus throwing off the stability. I managed to abate the vibration by balancing the bad motor with a zip tie (the ratcheting portion of the zip tie is heavier than the rest so it can be used as a weight is on a wheel).

3. The CRIUS SE controller board has a short on one of its power rails. Each ESC (electronic speed controller see First parts) has three wires that go to the controller board: Positive, ground and signal (marked +, G and S respectively) (see picture). The ESC has a circuit called a BEC (Battery Eliminator Circuit) that would, in a conventional circumstance power the receiver (at 6 volts). However, multirotor aircraft have a middle-man so to speak. Instead of the speed controllers getting a signal directly from a receiver, a controller board gets signals from a receiver, processes the commands from the receiver and the data from the sensors, then decides what commands to send to the speed controllers. The power distribution runs in the opposite order. The BEC in the ESC supplies a steady 6 volts (versus the full power of the battery) to the controller board (in my case the CRIUS SE), which in turn distributes the power to the receiver. This is not an issue when using in a hex configuration (six motors) because there is power being supplied from the the speed controllers for motors 5 and 6 to the good rail, but when in a quad configuration (four motors) there are no longer ESC wires attached to motor 5 and 6 pins. However, and this is strange, despite the board being unable to accept power from the shorted rails, it is able to supply power to the receiver via the rails. This may be due to the receiver pins (marked THR, ROLL, PITCH, YAW and MODE) having a power circuit separate from that of the motor pins. If in the future I wish to change the motor configuration to quad, I will have to figure out some way to supply power to the good power rials (most likely by attaching an extra ESC with no motor to work as a power supply). Here is the pinout of the board with the rails in question labeled:

Additional parts required to get flying

After purchasing most of my multirotor parts from my friend (see First parts), all I need to get flying is a battery charger, transmitter and receiver. I looked around for awhile and finally decided on the FlySky TH-9X transmitter for its relatively low cost and large number of channels (9 channels).

Here it is:

As for the charger, I had several 50 watt chargers recommended to me, but after doing some math decided that I needed at least an 80 watt charger. Here’s why:

The batteries I will be charging have an average voltage output of 11.1 (3-cells each at 3.7 volts) volts over their discharge period, but are charged at 14.48 volts (this is due to the electrochemistry behind the reverse reaction in the lithium-polymer battery). It is a general rule to charge your batteries at an amperage equal to the capacity of the battery. For example, my batteries are 3,300mAh (milliamp hours) or 3.3Ah (amp hours) so I would charge them at 3.3 amps. This means that my charger must have an overall power output of 3.3 amps x 14.48 volts = 47.784 watts. This is below 50 watts so a 50 watt charger would be able to handle this, but if I would like to charge my batteries at a higher amperage once in awhile for a faster charge time or purchase batteries with a higher voltage or capacity, a 50 watt charger will not be able to supply enough power (the charger automatically limits the current delivered to remain within its power rating (watts) while keeping the voltage at the correct level for the voltage and chemistry type of the battery being charged). Therefore I selected a charger with an 80 watt rating, specifically the Thunder AC680 for its relatively low price, built in power supply (transformer) and good selection of adapters. The reason I did not select a higher wattage charger is that chargers become considerably more expensive once they go over the 100 watt power rating and they require a separate external DC power supply, increasing the cost further.

Here’s the charger and adaptors (AC power cable not shown):