Monday, September 29, 2014

Flight #5 - Testing Nova and Mods

At the small neighborhood field and light winds. Overcast skies.

After making sure Nova would take-off and land ok, I did the AutoTrim procedure. No problem getting a GPS 3D-Lock ... Telemetry revealed sats=7 and hdop=1.8 . Tested currently assigned Flight Modes ... Loiter, Stabilize, Alt-Hold, Land and RTL.

This is also my first-ever YouTube video. The local edited 1080p file looks smoother than the uploaded YouTube video (even playing on 1080p). The YouTube video looks more like 20-25fps than 30fps. Overall, impressed with Mobius Camera. The sun-exposure glitch (pointed-out in video) has been mostly eliminated.

Everything on quad seems to be working fine after mods (see list in previous post). This includes Telemetry radios, which I have been wanting since I ordered the Nova. Nice having laptop and MissionPlanner at field, and not have to connect USB cable any more.

For the camera, I am leaning toward a FPV setup (therefore, the static camera mount), mainly because that's what many pilots say works best. I suppose Loiter assisted footage was acceptable for AP.

Here are some notes about GPS and Arming ... Green light on-steady means GPS 3D-Fix ... likely sats=5 or better (hdop is irrelevant). Allowing Arming in Loiter means hdop=2.0 or better. Number of sats is irrelevant (it's more about QUALITY of GPS signal as a whole). I always Arm in Loiter ... for best chance of GPS dependent functions to work. If at the flying field without your laptop, but having trouble, try this GPS App on your phone (just search by zip code).

Saturday, September 27, 2014

Testing & Setup of re-configured Nova

Since ...
- FC boards have been completely unplugged, removed, soldered on, and re-installed.
- FC, RX, and GPS Module re-mounted (with new foam tape).
- ESC PWM control cables disconnected from FC
- GPS cable connector broke and was replaced
- Way RX and FC get power was changed
- Taranis firmware upgraded
- Telemetry 915mhz v2 radio installed
- Mobius Camera installed
- Any number of other things changed

... this post will try to document the various tests and procedures that I deem required to get the Nova ready for flight.

Final Internal Config. Everything is close without actually touching.

Initial quick/easy tests and observations:

At this point, the FC, RX, and GPS-Module have all been re-mounted with new/fresh foam tape. Nothing is actually touching (not even micro-usb plug and RX below it).

In Terminal/Test/Motors ... Motors seem to spin in proper order and direction. Testing order is different that FC connection number.

GPS-Module seems to be working (tiny blue led on it's lower-pcb is flashing). Inside house, I get nothing (sats=0, hdop=100) but this is normal for my house. On the porch, I can get a 3D-GPS-Lock and sat=9, hdop=3.0 ... I think ok for an over-cast night.

Compass, Accelerometer, and Barometer (air pressure altitude) seem to be basically working ... according to MP-HUD (but will still re-calibrate).

Radio (as a set) seems to be working. Nova will Arm and spin motors. The sticks and battery meter have been calibrated at the radio itself (or at the "radio level"). Radio calibration can now be done in MP.

External LEDs are working.

Telemetry radios seem to work fine. Having on-board Telemetry radio On and/or in use does not seem to affect main RX reception (RSSI values don't really change or get worse).

Camera is basically stand-alone, but seem to work during testing.


Done like before (see previous posts) in MissionPlanner. First center your radio's Trim-buttons, and then calibrate Radio. Then, calibrate the Compass ... point each of 6 surface-faces to the ground and slowly spin around with it for each one ( aka "compass dance"). The globe model will end-up having 6 small circles embossed on it. The wireless telemetry radios make this east now. Current compass offsets are -54, 25, 2. And finally, the Accelerometer ... hold very still perpendicular-ly as instructed or prop-up.

ESCs calibrated (All at Once) in case end-points changed (center Trims first). FailSafe doesn't need to be changed but should be tested (see previous posts).

Carefully, did a "Hold by Landing Gear Test" ... with other hand, Armed and tested basic directional controls with TX.

All completed and/or passed.


In MissionPlanner/Config & Tuning/Extended Tuning ... give Opt 7 a function and also Opt8 (I set to RTL). In Taranis (Mixer), assign both to their own dedicated switches. Check all Modes on FailSafe screen.

Next Flight Objectives:

First, I will see if it will still take-off and land. Then re-do AutoTrim. Check currently available Modes (some new to me) ... Loiter, Stabilize, Alt-Hold, Land and RTL. Test Telemetry. Try to get my first flight-video with Mobius.

Re-Assigned Radio Channels

For the following reasons, I have decided to re-assign my available radio channels:
- I don't plan on installing a gimbal any time soon.
- Mobius camera has it's own battery. Might as well use it.
- I could use some extra FC Aux channels for dedicated Flight Mode switches.

RC Cable 1CH-1 VerticalCH-1 Vertical
RC Cable 2CH 2-4 Horizontal (signals only)CH 2-4 Horizontal (signals only)
RC Cable 3CH 5-7 Horizontal (signals only)CH 5-7 (FC Aux 1-3) Horiz
A-JumperCH 8CH 8 (FC Aux 4)

  • BEC-5v goes to RX and then to FC. FC (Flight Controller) really gets 5v from CH-1 connection.
  • Vertical connections are the standard way, with standard RC-Cable radio plug (Gnd, Power, and Signal).
  • I used "black to the right" when plugging in horizontal "servo-like" RC Cables.
  • A-Jumper is a standard (single-pin) 3inch Arduino Jumper cable
  • While Channel-Signal-Rail is on FC bottom, it's on X8R's top (opposite).
  • By using previously unused SBUS socket for power, frees a standard channel. It shares the same power-rails.
  • At this point, I can't even see a reason to need BEC-5v routed to lower connection panel. If need arises in the future, I can just use a RC Y-cable.

Good pic that someone made.
The white boxes are where you connect Dupont (servo like) RC-Cables. Mine is mostly like this. Only difference is that I went ahead and connected CH-6 also to RX (in case needed later). Red VCC and Black GND drawn boxes are just informational.

Friday, September 26, 2014

FrSky Taranis - OpenTX 2.x Firmware upgrade from 1.x

Taranis Radio has been working fine with my Nova quad. I thought this would be a good time to go ahead and upgrade to the latest OpenTX firmware. Might as well be running the latest code. I also wanted to start using OpenTX Companion, which is software you can load on your computer that allows you to program the radio from a computer (instead of just the radio itself).

My understanding is that the radio ships with a special FrSky version of OpenTX. In my case, it was several versions older.

This example uses my Taranis (Original, Model-B). It's a few months old and running FrSky stock firmware.
SVN: opentx-r
DATE: 2013-10-18
VERS: opentx-x9d-v1.1.02

Taranis should be disconnected from computer and switched Off.
This is the initial procedure when going from stock FrSky v1.x firmware to OpenTX 2.x (there is no easier way at this point).
To read or write Firmware to radio, Taranis should be Off when connecting to PC.
To read or write Models (Eeprom) to radio, Taranis should be On when connecting to PC.

After reading, to get started ... download, install, and use the proper version of Zadig (to install Windows driver for Taranis)

Follow the rest of the steps from to get OpenTX Companion installed, matching firmware downloaded and written to the radio. The current version at this time is v2.0.12.

After you get the firmware written to radio, the first time you turn on the radio, the current Eeprom will upgrade itself to a version that works with the new firmware. In my case, all the Models programming was still intact (but first flights will be made with caution until proven).

I then re-did the sticks and battery calibration (just to be on the safe side). Center your Trims first.

Here are some simple rules to follow while using OpenTX Companion (basic work-flow):
  1. Always start by reading the current models and settings (called EEPROM) from the radio.
  2. Save the models and settings to the computer as a backup.
  3. Make changes to the EEPROM directly or by copying from other EEPROMS.
  4. Save the EEPROM when all changes are made.
  5. Write the original (now modified) EEPROM back to the Transmitter.
In addition to having your models set-up, the EEPROM contains all of your individual radio settings and more importantly - your radio calibration settings.

There are more advanced tricks, but for now, I will stick to the core-rules and work-flow above. Seems like this prevents programming made at the radio itself (ie., at the flying field) from getting accidentally over-written. The radio will always be the "current live copy", with backups on the computer. In the even that I have to restore the radio, I just need to redo the battery and sticks calibrations, and I'll be back current.

Thanks to the helpful folks in this forum that helped me get through this easily.

Here are some newer Taranis general usage notes and tips for using OpenTX-Companion.

Taranis-Plus Notes:
I think the Taranis Plus also ships with a FrSky firmware, however, it's derived from OpenTX 2.x, so upgrading it's firmware to a true OpenTX 2.0 firmware will be easier than above (I was going from 1.x to 2.x). I think you just connect Taranis-Plus to Windows computer via USB with "3-Finger Salute" and go from there. Also, in OpenTX-Companion you must let it know you have a Taranis-Plus (and not a plain Taranis). Hope this tip helps, but I don't actually have a Plus to test it.

Sunday, September 21, 2014

Mobius Camera Install

Used most of the supplied mount to mount Mobius (Lens-A) camera. Drilled a 1/4inch hole in the middle of lower plate and used a (cut to exact length) 1/4in. nylon bolt to attach it. A little foam tape at front and back (just to make sure it doesn't turn and/or loosen). Between upper plate and body, I used two small nylon washers to help even-out upper plate due to rubber dampening ball's grommets in corners.

The wire-ties and fishing line are for emergency tethers (and are required ... balls won't hold in crash). One loop through all plates and parts. Even in this default position, landing gear does not appear to be in Mobius wide-angle lens.

Willing to give-up tilt to have camera up higher. Might adapt or hack a Mobius mount later. Obviously, no gimbal yet, but not really any "jello" either so stock mount is fine for now.

Current Settings (updated 2015-03): I just set my Mobius exposure to "Sunny" instead of "Auto". Just one flight that way but seems much better (especially when facing the sun). Instead of the camera temporarily making the rest of the scene dark, and then slowly transitioning back, it's more of a constant brightness. The sun just looks bright and there are some cool halos now. Anyway, thought this tip should go here.

Update 2015-06 Changed Primary recording mode to 1080p/30fps Narrow. Will decrease "fish-eye effect" when camera is pointed down. While the end-effect looks like a zoom, it's actually a pixel-matched 1080p recording from middle of lens. I also tried 720p/60fps/Narrow ... while it is smoother, it's not as sharp when played back on 1080p monitor. I have it setup as Secondary.

Sunday, September 14, 2014

Telemetry Upgrade Mod

The parts arrived last week, so I did the Telemetry upgrade on my Quanum Nova. Yes, connector is missing, but you can see that PCB is ready for it. This post has been recently updated, so read the whole thing before attempting.

HobbyKing Telemetry Radio set v2 (915 Mhz) ... aka SiK or 3DR Telemetry Radios
Molex PicoBlade 5-pin connector (53047 Series)

Phase 1:
Removed lower connector board from FlightController.


Soldered in PicoBlade connector (watch orientation). You want red wire on cable to connect VCC power to 5v on radio. Solder with plenty of light and a magnifying-glass helps (the solder-pads at 1.25mm spacing are very close).

After soldering, I checked for unintentional shorts and proper continuity with an ohm meter. All checks out good. 

Installed the 4-wire 6-inch cable (supplied with radios).

Re-assembled enough for testing and to prevent any chance of short-circuits/damage.

Phase 2:
Connect one radio to Nova via PicoBlade cable, and the other radio to computer via  supplied Micro-USB cable.
Connect LiPo battery on Nova.
Start MissionPlanner. Set to "Auto" and then click connect.
Connects and works fine. See Artificial Horizon and Compass Heading respond on computer.

Phase 3: Mounted here with foam tape for now. The slot cut for wire access is in "channel" next to battery area (as to not affect body strength). According the RSSI bench-tests, having this 915mhz radio on doesn't really affect FrSky 2.4ghz radio (even though the 915 antenna and 2.4ghz antennas are fairly close in this location).

I have since moved Telemetry radio inside Nova.

It's come to my attention (from RCGroups posts) that this mod might only work if the FC is still running the stock-factory Arducopter (in my case v3.1.2). I put too much work into this, so I'm staying on ArduCopter v3.1.2 for now.

This version of the HobbyKing Quanum Nova Flight Controller are marked like ... PCB-1: APM_V252_V2.0 and PCB-2:  CON_V2.0 . In Jan-2015, some newer revision board are showing up and it sounds as if the boards are different and connector pre-soldered-in (but accidentally backwards, according the established 3DR pinouts). If using pre-made cables, check your VCC power with meter first.

The direct USB connection and Telemetry radios CAN NOT be used concurrently. Pick one or the other each time. If using Telemetry radios, do not connect normal direct-connect USB cable to bottom of Nova.

My radios came pre-programmed with matching settings (required for basic link and normal operation). There is an utility in MissionPlanner that can be used to verify and/or change them if needed, but I didn't need to use it in my case. IIRC, it is like in the Settings section and radio module on Nova just needs to be getting power also.

While the Telemetry port is 5-pin places, there are only 4 wires and/or connections. Two are for power (VCC+5 & GND) and should never be swapped (verify with volt-meter). The other 2 are signal RX & TX. Between the board and radio, TX goes to RX and vice-versa. If TX and RX are swapped, you won't damage anything, it just won't work. The supplied cable was already wired properly in my case.

To use the direct USB connect to computer instead ... leave the computer's radio disconnected. Pick the other com-port and set port speed back to 115200. I don't have to disconnect the Nova's radio. It doesn't seem to care if it's powered-up. It's ok to use the Auto com-port setting in MP if you need to.

Telemetry radios sets come in 915mhz or 433mhz. You should probably get the correct frequency for your country. Power should be left at max or 100mW (MilliWatts).

The clones (HobbyKing and generic Chinese) of 3D-Robotics Flight-Controllers and connecting sub-devices tend to use Molex PicoBlade 53047 Series connectors, while devices and cables purchased directly from 3D-Robotics use Hirose DF-13 Series connectors (to match the Hirose plugs on genuine 3DRobotics Flight-Controllers). Both are 1.25mm pin-spacing but are slightly different otherwise. The plastic connector-housing is shaped a bit different. Additionally, the Hirose DF13 pins themselves are square-ish, while the PicoBlade pins are rectangular/flat like "blades".  This makes the male and female pins incompatible in my book. I personally would not mix any parts of the two connector types. Instead, I would build or buy the proper cable.

The idea for this mod. is based on this blog article. Many thanks to Jean-Louis Naudin. However, I highly suggest you take the time to solder-in the PicoBlade connector.

With the release of (custom-compiled) Special CX-20 APM252 version of ArduCopter v3.1.5, there is now really no reason for anyone to attempt to do DogDude's (risky) originally popular "Resistor Hack" mod. Install the Molex PicoBlade connector and run an appropriate version of custom ArduCopter.

ArduCopter Versions that Support Telemetry on this APM_252

I guess what it comes down to is that this Cheerson APM_252 Flight Controller requires some special ArduCopter code for the Telemetry/Uart port to work properly in the Nova and CX-20. The common/public versions do not have the special tweak, so you have to run a version that has been custom compiled.

Here are the various versions that I know about:

ArduCopter v3.1.2 (Special CX-20 APM252 version). It's only available pre-loaded from factory (no way to download it if it ever gets erased). This still appears to be the "shipping version". Pretty stable code and should fly your quad fine at first.

ArduCopter v3.1.5 (Custom CX-20 APM252 version). Last stable release of 3.1.x based code. Still has Terminal and this simpler ArduCopter code seems to run good on this tiny processor. This is what I run now on this quad's Flight-Controller. Many thanks to VinnieRC.

ArduCopter v3.2 (Custom CX-20 APM252 version). Reportedly, you can also successfully run this Custom Compiled Firmware if you wish (Thanks to Frank/ Funtastix). Code is more complex and Terminal has been removed. Watch your Logging level (due to processor over-head) and PM-LongLoops.

Cheerson APM_252_5.0 PCBs (Update Feb. 2015)

According to RCGroups Forum reports, this mod works with both v2.0 and v5.0 PCBs in CX-20 APM252 FlightControllers. What I have been calling "Telemetry Port", Cheerson/HK is now calling "Uart port" ... it's the same port. Version 5.0 boards not only have Molex PicoBlade socket pre-soldered-in for Uart port, they also have ADC Port for easy PowerModule connection. So, these APM_v252_v5.0 boards have the following connectors soldered in:
UART (Telemetry) : 5 pin Molex PicoBlade
I2C (Compass/Magnetometer) : 4 pin Molex PicoBlade
ADC (Power Module) : 3 pin Molex Pico-Blade

While Cheerson has added Uart PicoPlade sockets on v5.0 PCBs (so you don't have to solder-in your own), they might be soldered in backwards. If using any pre-made cables, be sure main voltage is on correct pins (VCC polarity), or you will destroy Telemetry radio module instantly. You do this by verifying voltage polarity on Uart pins with a volt-meter before connecting cable. Also, you can try swapping TX-RX wires if no communication after trying everything else (only TX-and-RX wires can be swapped without fear of damaging anything).

For example, telemetry radio just randomly disconnects and re-connects. These are listed in most-common problem order first.

Be sure Settings match in both radios. Try the default settings. Baud Rate should usually be 57 (57,600) at normal distances. You do this from a laptop. I wouldn't deviate from these settings much, but if you know what you are doing, here are some details on Advanced Settings , and also this page.

The default for ECC is ON. While it is 50% slower (due to the over-head of error-checking code), several docs say it is "highly recommended" and they also mention longer ranges.

The first time you get it working should be from Windows computer running MissionPlanner (other radio on quad).

These 915mhz v2 radios have very good range and Frequency Hopping Spread Spectrum (FHSS). I have seen them reach 0.5 mile easily, and they likely go further than that. If there is a lot of multi-rotor activity around you, try changing the Channel (aka Net-ID).

If you are using DroidPlanner (on Android device and with USB-OTG cable), instead try a Windows laptop, MissionPlanner, and a different cable. See where problem follows. Also, vice-versa (if you normally use MP, try DP).

Try to observe power lights on radios to make sure you don't have a power problem and/or a simple cable problem. USB cables (especially the smaller form-factor ones are historically not durable).

While I use MissionPlanner, here is a tutorial using APM Planner and what appear to be a bundled 3DR RadioConfig. Nice graph might be helpful in troubleshooting.