Wednesday, November 22, 2017

Setting up Nova (CX-20 PnF/ARF APM Open-Source) from scratch

OK, I'm down to just the one Quanum Nova now (just one good shell-body ... and plenty of spare parts). Last year, after this Nova #2 flew it's maiden and a couple of other flights, it blew an ESC (for no reason). After I repaired it, I was having problems getting the accelerometer calibrated. It's been sitting in this limbo state in the shop since then. The hardware is done and it's body shell is screwed-shut. It's fairly stock, but does have the newer APM-252_v5.0 FC, a Power Module, and PDB is raised a full 6mm (all best features/mods, unlike my old dead main Nova). I decided I'm just going to just "format it" and start over from scratch. 

From now-on, this model named "Nova 360" will be the best of the Quanum Nova parts, built-into a working model.

RC Radio Setup and Config:

Using OpenTX-Companion, copied my old Nova model to a new one called "Nova 360" and uploaded to Taranis. Radio is now mostly programmed for new model. Since this is a X8R, of course, the TX and RX must be bound-together using those normal procedures. You usually have to touch the RX, so do this before closing-up the Nova. 

Flight-Controller (FC) Software Setup and Config in Mission-Planner:

When you first use Mission Planner (MP) to Connect to the APM Flight Controller (FC), it retrieves the current parameters from it so you can work on them. When using MP, it's important to remember to click the Write button to write the parameters you just changed to the APM-FC before Disconnecting or closing the program. This includes any calibrations you just made, because calibrations are also just saved as parameters in the param config file. After you Write them to the FC, you can optionally Save them (as a file) for later Loading or Comparisons. Just because MP presents telemetry in Real-Time in some parts, don't confuse that with parameters. For params and calibrations, MP is just a temporary holding or work area.

One other tip for using Mission Planner v1.3.50. When programming the main parameters in the Full List, I found it easier to use a real mouse instead of the touch-pad on the laptop. If you were to accidentally brush-up-against a field (and erase and/or change it) that would be very bad. Take your time and be very precise. 
  • Connect USB cable from laptop to USB-Mini Port on bottom on Nova
  • Start Mission-Planner and click connect to test port and Connect ability.
    • Use Auto at first if you don't know com-port and baud rate
    • Click Disconnect but leave cable attached
  • Upload and Verify Custom Firmware file v3.1.5 (from VinnieRC) aka v3.1.5c
  • Click Connect
  • Reset (Parameters) to Defaults (for v3.1.5) and let it Reboot.
  • Save initial default v3.1.5 params to file (for future reference)
  • If not already Disconnected in MP, do so now.
  • Shut-Down MP, disconnect cable, and power cycle Nova.
Since I just re-loaded the firmware, I've learned it's normal for the Nova and ESCs to start beeping constantly on next full powered-up. Seems to indicate that ESCs aren't getting a signal because FC can detect that radio is not calibrated yet (getting that message in MP-HUD). Looks like the constant ESC beeping happens with ArduCopter v3.1.5 (or higher) and certain versions of ESCs. And, of course, what the APM-FC considers an un-calibrated RC radio.
    • Reconnect USB cable, start MP, and click Connect.
  • In MP, calibrate the RC Radio, 
    • The sticks, but since the radio is already programmed, step thru all the pre-programmed Taranis Flight-Modes on the switches. All channels should move on MP screen (except CH6 which is not assigned to anything on Taranis).
    • Write params, click Disconnect and power-cycle (reboot) FC.
    • If ESCs were beeping before, they should stop now.
  • Verify or Set Frame Type to X-Y6A (Nova is a X-4 quad-config)
  • Recall that you have the APM-FC mounted and orientation in normal/forward
  • Verify proper Compass hardware-config settings. Mine are fairly unique mounting for a Nova, requiring Yaw-90 degrees, and also Forced-External (which also disables any Auto-Detection). As you recall, the Nova has a single external-compass.
  • I would say these are the minimum parameters you want to set or verify before calibrating:

AHRS_ORIENTATION,0
COMPASS_EXTERNAL,2 
COMPASS_ORIENT,2 
COMPASS_USE,1
FRAME,1

  • Click Connect and calibrate the rest of the "Mandatory Hardware" devices.
  • Accelerometer calibration was NOT the usual procedure. I'm not sure if it's this special little APM-252 FC, the AC-v3.1.5, or how recent Mission Planner versions work with APM FCs these days ... but I was unable to calibrate the accelerometer the normal way. I had to use Terminal. 
    • Set to APM and click Terminal's Connect Button
    • Go to Setup / Accel
      • Remember that ? or help shows the current menu commands, and exit moves back one level
    • Follow directions. I find it easier to prop-up against a door because it must be keep still and perpendicular at each step. Use Enter/Return to go to next step.
    • It should say Successful at end. There were 6 shown, all between -0.03 and +1.4. I think it might Write them to the FC at the Successful end.
      • However, to be sure, I take a phone pic or screen-shot. Later, you can verify as INS_??????? in params list.
    • Return to Full Parameters List and Write them. You might need to Connect first. Finally, I Save as a File since we made it this far. 
  • Compass calibration, before I started, I went outside to get a good HDOP or at least enough SATs for a basic GPS-Fix (for Auto-Declination to function). Got a nice and fast (ie "clean") calibration. Actually, I got several similar good ones in a row, and kept the last successful one. My fairly unique Nova compass params (see above) I have been using for years. The new (to me) COMPASS_EXTERNAL,2 parameter seems to work fine also (Forced-External, no Auto-Detect) for its single external compass/magnetometer. Final Compass Calibration Offsets were -29, -68, -47.
    • I think they might auto-Write to APM-FC on success, but I go ahead and Write them manually anyway (and Save another File for this setup-progress step).
    • After testing that it new the basic 4 directions, I did notice that the Nova's exact North is about 7 degrees off as compared to my cheap analog hand compass. Since this old hand-compass does not take Declination into consideration, this is normal and correct. The Nova's compass appears to be calibrated properly. 

At this point, I'm still running the ArduCopter v3.1.5 Default Parameters and default tuning PIDs. Accordng to my notes, these work fine with a mostly un-payload-ed Nova (ie a Nova with a low-weight or THR_MID of 500-575). 





Calibrated ESCs (especially since this is a PnF model).

Set my Flight Modes in MP interface (and write/save):
1. Stabilize
2. Loiter
3. Alt-Hold
4. Drift
5. Land
6. Auto
... and in Extended Tuning, CH7=SuperSimple and CH8=RTL
(which I have each on dedicated Taranis buttons)

The above should change these Flight-Mode parameters:

FLTMODE1,0
FLTMODE2,5
FLTMODE3,2
FLTMODE4,11
FLTMODE5,9
FLTMODE6,3
CH7_OPT,13
CH8_OPT,4

Also, verify or change these parameters (some might be defaults):

ANGLE_MAX,4000
GPS_HDOP_GOOD,220
LAND_SPEED,40
LOG_BITMASK,958
MOT_SPIN_ARMED,0
RTL_ALT,2500
THR_MAX,1000
THR_MID,530                         (Only payload now is small SiK 915mhz Telemetry radio)
WP_YAW_BEHAVIOR,2         (Might prefer 1 for Auto-Missions that end in RTL)

BATT_AMP_OFFSET,0           (This set is for installed Power-Module. Notice PINs)
BATT_AMP_PERVOLT,19.1
BATT_CAPACITY,2700
BATT_CURR_PIN,12
BATT_MONITOR,4
BATT_VOLT_MULT,10.4
BATT_VOLT_PIN,13

FS_BATT_ENABLE,0              (This set is the Fail-Safe actions)
FS_BATT_MAH,0
FS_BATT_VOLTAGE,6
FS_GCS_ENABLE,0
FS_GPS_ENABLE,1
FS_THR_ENABLE,1                (Very important)
FS_THR_VALUE,975

Tested Fail-Safe with no props installed (Passed).

Completed Nova assembly. Since everything else was done, really just had to install props (watch CW and CCW) and add a bit of clear tape to compass-pod cover.

Armed and carefully did low-throttle "hold by landing gear" test. Also, verify flight modes (all Passed).

Looks like its Ready for it's test flight after this rebuild.

No comments:

Post a Comment

The stupid spammers have now forced me to approve each Comment before it appears (but I am usually pretty quick about it).