I’m getting more into FPV, but it’s hard to find enough time for the planes. I have to go to a place where I can send them up, and make sure that there isn’t anything I can fly into etc, so I was thinking that a quad might be a better platform for the short afternoon flights.
I’ve never had a quadcopter before, so I started looking at HobbyKing, where most products suit my wallet. The Quanum Nova quickly became a good option, at $319 with everything included to fly except 4 AA batteries, it’s an excellent deal, and with free shipping with Swiss Post as well!
Four days later the box was at my doorstep, in perfect condition! I started unboxing it, and wow, it’s a lot of boxes. So here’s a quick unboxing gallery, with what’s included:
The first thing I noticed, was that the charger was rather underrated. At 500mAh, it’s going to take hours and hours to charge this thing, I’m probably not going to use this a lot.
It seems like the quad can be modified to fit a larger battery, 2700mAh is really not a lot for four motors. The power distribution board could be moved up a bit to allow a thicker battery to fit, but the battery hatch is a bit small, so it might not work. I’ve read about someone successfully adding a mount underneath it instead, and fitting another 2700mAh in parallel, or a single 4000mAh. Maybe I’ll try this later if I need longer flight times.
The two holes are for the GoPro mount, these go into the battery compartment as you can see above, be careful not to use longer screws here, they could hit the battery! There’s also two +5V ports (for optional gimbal) and two +12V (for optional FPV gear) ports at the bottom, in addition to the USB-port used to connect the flight controller to your computer. I’ll probably make use of the FPV option in the future. You can also see the antenna to the right, which is supposed to be mounted to the landing gear after fitting it.
I mounted the included GoPro mount, there were no instructions, but it was fairly simple, mount the four vibration dampers in one of the frames, mount the GoPro mount on the other, use two screws to mount the frame with the dampers, and pull the rubber vibration dampers through. This means that the only thing holding up the GoPro, are the rubber dampers, so I might add something to secure it, just in case.
After mounting it, I noticed that the compass cover was starting to come off. It’s very loose, and would probably fall off after the first flight. Will probably tape this to make sure it’s tight.
Just to check that the electronics were OK, I plugged it in my Mac and started APM Planner 2.0, and also plugged in the battery (although not fully charged yet) to stop the beeping.
After connecting, I got a message recommending me to update to ArduCopter 3.2 Stable, so I did (had to click disconnect first). This was a simple process, and finished quickly. Then when I connected again, I got a new message: “PreArm: Compass not healthy”. Not good…
This meant opening the unit, and try to find the source of the problem. The magnetometer is located at the top, the “radar” looking thingy, but everything looked fine there. I still had to disconnect it to open the Nova. It’s easy to open it, remove all hex screws except the four for the motor, and a philips inside the battery compartment, and the top comes off. Still nothing out of the ordinary, a couple of solder points at the ESC’s were rather bad, but this has nothing to do with my problem – so I opened the flight controller, and there it was:
The connector was ripped off the board. Also notice the red wires, they all have marks from being squeezed between the board and the housing of the flight controller. There was an extra support pin on the casing, just where the connector is, making it even tighter. Weird that they didn’t use a downwards facing connector here, as the cable goes out towards the other side. Also I think the connector should be glued to the board, not just being kept in place by the pins, this is a weakness. I’m not sure what the cause for the connector being ripped off is, the tight squeeze or if the cables have been pulled, all I know is that it doesn’t work.
While I was in there, I figured I would check the rest of the electronics as well. Everything seems OK except rust on the GPS and a couple of melted wires at the solder points. These will probably not cause any problems, but it’s wise to inspect these before first takeoff in my opinion.
I contacted HobbyKing straight away, and was told to issue a RMA. After a couple of days, I still hadn’t heard anything, and contacted them again, I asked if there was any possibility of receiving a new flight controller (HK sells them separately), but they told me no. I had to send the whole quadcopter in so the product specialist could take a look first. I could send it to the local warehouse rather than to China, but still. It isn’t in stock anywhere at the moment, so that would mean waiting a long time. At the end we agreed that I could fix it by soldering it without voiding the warranty, which is fine. The problem is that after trying to solder it, the solder pads melted away almost immediately, even with a 30W iron! So there’s definitely a low quality on some of the parts. I contacted HK again, saying that I’ve tried to fix it, but it’s starting to go beyond repair, and told them to please reconsider, as this part was defective at arrival. After a quick moment they came back to me and told me that a new flight controller would be sent to my address as soon as possible – Excellent customer service! 🙂
While waiting for the new FC I continued with the propellers, I recently bought a DU-BRO 499 prop balancer, so this was a nice chance to test it. The propellers were definitely off balance. I just a 320 grit sandpaper to get it balanced, and it took a long time to sand down enough, white powder everywhere!
Both the blade and hubs needed to be sanded on all propellers to get them balanced, some of the hubs were pretty poorly balanced, like this one where a big portion of the hub had to be sanded down.
New flight controller
After 10 days I received my new flight controller, I opened it to check the connector, and this was actually different from the board I had. My original one said APM_V252_V5.0, while this was APM_V252_V2.0. I guess these differences are mainly the connector board, and I actually prefer the V2.0 as it has the connector in the middle of the board, facing downwards instead of sideways towards to casing.
While I was waiting for the new flight controller, I read other peoples experiences with the Nova, and a lot of them recommended you to calibrate the radio and ESC’s to make sure the motors are fully in sync. I tested myself, and even though they weren’t way off, they definitely weren’t in sync.
I started out by calibrating the actual transmitter. It has its own internal calibration for the sticks and should be done before the ESC calibration. It’s very simple:
The ESC calibration is a bit tricky, as the stock Nova radio has to have the receiver powered on before the transmitter to bind, not the other way around which is common with “all” radios. My ESC’s are soldered, and can’t be disconnected, meaning I can’t easily achieve this. Some have had luck by connecting a USB cable on the Nova to power only the receiver and flight controller, and later connect the battery to turn on the ESC’s. This didn’t work for me, as plugging in the USB cable also activated the ESC’s and therefore wouldn’t let me put them in calibration mode.
I tried disconnecting some of the leads from the receiver, and connecting my old flight controller to it, to use it as a +5V power source, but the new APM flight controller didn’t seem to want to enter calibration mode. I tried all kinds of different connections, but whatever I did, I couldn’t get it calibrated! After reading about ESC calibration at ardupilot.com I noticed this at the end of the page:
The All-at-once ESC calibration mode simply causes the APM to pass through the pilot’s throttle directly through to the ESCs. If you power up the APM while in this mode you’ll send the same PWM signal to all the ESCs. That’s all it does.
Meaning that it doesn’t actually use the APM flight controller for anything while calibrating the ESC’s, the signal from the receiver is just sent directly to all ESC outputs, meaning the flight controller can be bypassed for a simple 1-1 calibration with the receiver instead if it doesn’t work.
I ordered a servo splitter cable from HobbyKing (these can be found all over) so that I could connect the one throttle output from the receiver (port number 3) to all the ESC’s. Then I disconnected all other cables from the receiver, except the aileron which I put in my old flight controller used as a power source (this can be any +5V power source, battery etc) and powered on the receiver then the transmitter.
The receiver lit a steady red light, meaning it has a successful bind, so I turned the throttle to full, then connected the battery to the quad – two beeps from the ESC, lower throttle to zero, and the ESC initialize sequence started. The ESC’s are now fully in sync, and I could connect all wires back to the way they were.
What I learned from this experience is that the PNF would probably be my choice if I were to buy again. Any normal radio would easily be able to calibrate by turning the transmitter on with full throttle then the quad – and you would be done! Requiring the receiver to turn on before the transmitter makes it impossible without opening the quad! Anyways, connecting the ESC’s directly to the radio receiver and doing it the old fashion way seemed to be the quickest and simplest solution if the ESC’s are soldered.
After connecting the Nova with a USB cable I opened APM Planner and went through the initial configuration. I started out with the radio, which was fairly simple (remember to also use the flight mode switches during calibration!) AUX 1 and 2 won’t work here, as they’re not connected to the APM, they go straight to the gimbal outputs.
6, 7,8 is disconnected. 5 is the flight mode switch.
Then I ran the accelerometer calibration. This is fairly simple, put the quad on each of its 6 axes and click a button. But make sure you hold it perfectly still and in an correct angle. I used my laptop as a reference for a 90 degree angle, and some spice containers to make it lay flat on its back (the disc with the magnetometer is in the way). I used a 10ft USB mini cable (PS3 accessory) from Nyko to hook it up, this made it easier to move around.
As I had no idea which numbers I were looking for, I did it three times, the last one was when I used the spice containers, the other times I just held it steady on its back. It turned out like this:
AFAIK the second and last were pretty similar, so I guess I’m done with the accelerometer, the next is the compass.
While still having the quad connected, I started the compass calibration routine. I rotated it around its own axis, instead of doing the whole “calibration dance”. The process is similar to the accelerometer calibration, flip it on one of the sides, then turn it 90 degrees, stop, another 90 degrees etc. four times, then flip it to the next side. I did this for all 6 axes, which is probably not necessary for the compass, but it seemed to turn out fine. The offsets were very low, so it should be OK. I wasn’t sure what I was really looking for, so I tried several times to see if the numbers were similar.
Last, I also did a compass magnetic interference calibration. There shouldn’t really be much interference on this quad, as the magnetometer is all the way at the top, away from the motors, but I just did it while I was at it. You’re advised to flip the propellers, and switch the counterclockwise with the clockwise rotating ones, so it’s pushing downwards. As these propellers can’t be flipped, I just switched them so they were rotating the wrong way. The procedure is simply done by arming and slowly throttling up to about 75%, and you’re done. I held the legs on the quad down while doing this, making sure it didn’t move, while also making sure my arm wouldn’t touch the propellers. They did bend a bit downwards when throttling up, but still had half an inch left before hitting the quadcopter’s arms. The result shows that there wasn’t really much interference anyways.
This should be it for the calibration part. Now what’s left is testing it, and to set the trim.