Tuesday, December 26, 2017

PixRacer Advanced Setup (PixHawk - ArduCopter Latest Stable)

My first PixHawk (PixRacer) build starts here.

FRS FrSky S.Port 
This should enable a lot of Telemetry to now run over the FrSky's 2.4ghz main link/channels, instead of using a dedicated SiK telemetry radio (for example).

On FRS connector, connect both signal lines together, and then connect to the S.Port line at X8R. Also, connect GND (black wires) but NO RED (Power) WIRE (X8R is already powered by PixRacer through RCIN connector).



Upgrading Tarasis to OpenTX v2.2.1

I had to upgrade my older OpenTX version before the newer "Discover Sensors" was even available for use on Taranis.

It seemed to work fine going from old Open-TX v2.0.12 all the way up to v2.2.1 (all at once) on my FrSky Taranis X9D. Rememeber that you must also upgrade your OpenTX Companion for Windows to version 2.2.1 (so they match exactly):

I followed DKEmxr's forum-post instructions and also some tips from Oscar Liang at this page. I backed-up first, but luckily, I didn't seem to lose any existing models or programming.

As I recall, I used these Settings in OpenTX Companion v2.2.1:
My Radio
FrSky Taranis X9D (specify exact model)
en (English for me)
lua and luac (only those checked, I dis-liked the way sq5font looked, so i re-did it)
Mode-2
AETR (what ArduCopter uses)
-
IIRC, I also erased the SD-Card and loaded a new set of files from sdcard-taranis-x9-2.2V0013.zip
And finally toward the end, I found and reinstall Amber (for Open-TX v2.2.x) .
Ok, now all it right with the world :)

With Open-TX v2.2.1 installed, now the "Discover Sensors" option is actually there. But so far, it reads RSSI (for RX) and RxBat (the default ones).

In Mission-Planner, the PixRacer's FRS (FrSky) port is Serial-4. So, set Baud = 57 and Protocol=4. Save that and Reboot PixRacer. Now, when you Discover-Sensors on Taranis, many (all ?) should now populate the List of Telemetry Sensors.

In the Model's Telemetry screen, you can assign Sensor fields to Main Page Flying Screen.



PixRacer ESP8266 WiFi Module

Since it's not "keyed" be careful to always connect like this little WiFi board this way. Be sure it's turned like this, and always use the 8 right-most holes.



On mine, the SSID was "AI-THINKER_code" by default. This is a good sign that the firmware is out-dated. While I could connect the laptop's WiFi to it, I was unable to establish a connection through it with Mission-Planner. DKEmxr says this tiny ESP8266 must be flashed with better firmware before it will work properly.

It apparently must be flashed to v1.1.1 (or maybe at least v1.1.0) . I used NodeMCU-Flasher. It will set SSID to "PIXRACER". It will also set the ESP8266's WiFi AccessPoint to be password-protected now (pixhawk). After running it, you should (might) be able to do further Firmware upgrades "Over the Air". While working with NodeMCU, it might help to disconnect from Internet and connect laptop only to ESP8266 SSID (AccessPoint/ HotSpot) via WiFi.

If your FTDI-wiring is correct, but you have trouble getting NodeMCU-Flasher to work, try swapping the wires for TX and RX (nothing should blow or fry just swapping those two lines). This ended-up being the reason my ESP8266 couldn't initially get MAC addresses in NodeMCU (and finally initiate the Firmware Flashing process). I followed the (simple) cable pin-out exactly so I think that maybe the silk-screened TX and RX labels on my new 3.3volt FTDI-USB Interface might be incorrectly swapped-around.

Communicating with PixRacer using Mission-Planner over WiFi (ESP8266 Module)

In the top right corner of Mission Planner, choose UDP. I don't think baud rate matters. Click the CONNECT button. If it asks for the Local port, enter 14550.

While I haven't tried it yet, I've read that it also works with QGroundControl.

Friday, December 15, 2017

Eachine E010s FPV Indoor Quadcopter

Winter is here, but that doesn't mean you have to stop flying :)

After comparing several inexpensive indoor FPV racing-drones, I found this Eachine E010s quad. It has a nice camera and F3-class Flight Controller. The 65mm ducted-fan design also provides prop-guards (to save the pets, furniture, and walls). Sure, it has brushed (core-less) motors and matching ESCs, but that helps to keep the price and weight down. I've since read the pilots call it a cheap Tiny Whoop or Inductrix clone. It's even Bind-and-Fly so you can use your favorite hobby-class transmitter. I have a Taranis, so I got the FrSky RX version.

 Eachine E010s FPV Drone


While the directions are fairly-poorly translated (no surprise there ... but we are used to it, right :), there ARE some good specs and tidbits of info. These are readable if you open them in a new window (and zoom-in). Hopefully this blog-post will fill-in the missing pieces and get you started in the right direction. I know the Blue-LED is for the FrSky (PPM) RX radio. Additionally, Red-LED appears to be Power, and I think the Green-LED might be the F3-Evo  Flight-Controller (complete with brushed-ESCs).


Radio Setup

I thought the setup was easy or fairly normal (if you have some previous experience with FrSky and CleanFlight).

First, you Bind the RX to TX and setup a new model on the Taranis. The Mode is D8 and Channel Range is CH1-8. I saw some pilots thought the Binding procedure was difficult, but I think it was easy and like any other FrSky Bind. Like always (with any FrSky equipment), the E010s should not be too close (within 3 ft) to Taranis or you might have communication problems. These FrSky radios are tuned for long-range.


  1. When you start, the E010s' Blue-LED for RX will be flashing (means it's un-bound).
  2. At Taranis, select Model. Go to Model Setup and click Bind. Taranis will start Beeping
  3. Carefully, hold-down tiny button on E010s FC PCB and apply power (connect it's battery completely). Blue-LED will turn off
  4. Stop the Bind-beeping at Taranis
  5. Cycle power on E010s (disconnect and re-connect it's 3.7v 240mah 45c 1-cell LiPo battery).
  6. The Blue-LED should now be solid-on (indicating it's successfully Bound).
  7. On the Taranis display, you should now be getting basic FrSky Telemetry and it will show the E010s' battery voltage (similar to your full-sized FrSky RX equipped multi-rotors).

Like other quads with F1-F4 based FC, the actual programming at the TX radio is fairly basic. You just need TAER1234. CH1-4 is TAER for main-control and 1234 are for any (AUX) switches you want to assign to CH 5-8.



I'm used to setting up for SBus (not PPM), but this seems to work (including switches). TAER is the important part. If you wanted to lock it into Angle-Mode, I think it might fly with as little as 4-channels.



CleanFlight 1.x Setup

Like many older F1 (Naze32) and even smaller current F3-based quads, you should use a CleanFlight 1.x. Even if you already have a newer v2.x version of CleanFlight installed, the old v1.x can also be installed or "side-loaded". v1.x can be installed as a Google Chrome App in Extensions Developer Mode. I still use CF v1.2.2, but I've also read about pilots using v1.2.4 (which is the last stable v1.x release).

Of course, (since you are using CleanFlight) you also have to install the SiL-CP210x Drivers (in my case, for Microsoft Windows-10). You will know that they are correctly installed when the (USB connected) quad appears in "Devices and Printers" with a Comm-Port at the end of its name.

Remember that while using CleanFlight ... after you make a change on any page, scroll to the bottom of that page and click Save. Also, I'm not really going to document each setting here now (especially if it's a default that is already set correctly). After you get CF installed and running, select your Comm-Port and then click Connect (top right).

  • Setup
    • Backup is the first thing I suggest you do. This should be the E010s' current "shipping defaults" ... to a file for safe-keeping.
  • Ports
    • Nothing to change here. Remember to never change or mess-with the first (UART-1) line, or you will lose USB communication with the FC :(
  • Receiver
    • Channel Map change to JR/Spektrum (and then click Save) to get TAER1234. At this point, sticks on Taranis radio should be moving the proper channels.
    • Direction say to set Serial Receiver Provider to "SPEKTRUM1024". It's the default first item so no changes required. However, since this quad is setup on PPM, I'm not sure it's even relevant.
    • I've never been able to find an actual "radio calibration" in CF over the years, and it seems to work fine without it. Of course, the Taranis itself is already calibrated (but that's not something you would do in CleanFlight).
  • Config
    • Mixer stays on Quad-X
    • Receiver Mode changed to RX-PPM . As far as I can tell, no SBus on this one (but it's not like we need the extra channels anyway).
    • The directions say to set Serial Receiver Provider to "SPEKTRUM1024". It's the default first item so no changes required. However, since this quad is setup on PPM, I'm not sure it's even relevant.
  • Modes
    • Aux-1 (aka 1st assigned switch on TX). I used Angle (flight mode) for all 3 positions
    • Aux-3 HeadFree. Set to the 3rd assigned switch. Bar takes-up 2/3 of line so it's active when the switch is in Position 2 or even Position-3
    • More experienced pilots might want other modes (like Acro) assigned to different switches and position. I probably add that or similar later.
    • I didn't have any problems Arming the quad the normal way (Throttle stick down and to right for a few seconds), but some pilots apparently do. They assign a switch to Arm.
  • CLI
    • I see a note in the directions-sheet to SET motor_pwm_rate=1000. Actually, I did NOT set this yet, but I'm thinking it's already set correctly because the E010s flies fine.
  • Setup
    • Setup screen one last time and calibrate accelerometer. It's real easy in CF since quad just needs to sit still on level surface for a few seconds.
    • Backup final settings to another file (since it's setup properly now).
    • AFAIK, there is no barometer, magnetometer (compass), or any other sensors to calibrate on this quad.

Plastic Body (Protector Shell)

Yeah, it's a strange looking robot head, but this is Quadcopter-Robotics :) Seriously, I thought it was a good idea to install it to help protect the camera, antenna, and FC-PCB. Installation is self explanatory. I did have to shave-off about 0.5mm of the two top guides to allow the camera to slide into place completely (you want the front of lens pretty close to flush in front). I think it also does a good job of supporting the camera/antenna.



I'm surely not taking off the body any time soon. I don't expect needing the Bind button again soon. However, it might be a good idea to carefully Dremel a little slot for the USB plug. I also saw a mod that puts a "arch protector strip" or loop over the antenna (for additional protection).

Conclusion

I tried Angle and HeadFree flight-modes so far, and it flies good for me indoors. Even for beginners ... while some simulator-time is a good idea, the Eachine E010s should also work nicely as a FPV trainer (when indoors or even outdoors with no wind).

Blue-LED (on steady) = FrSky RX and TX R/C radios are bound and connected
Green-LED (on steady) = F3-Evo Flight Controller is Armed (also visible thru body)

If I do any more mods to it I will try to post them here with an edit.

Tuesday, December 5, 2017

Hobbico Rise RXD-250 Racing Quad - FPV Trainer

I saw this for $50 (new-in-box) over at the RCGroups Classifieds and could not resist :). Here is what it looks like so far. I'll add more later.



It has an (older) CC3D (6DoF) and they suggest you use LibrePilot. Sure, why not ... What is that now ... 3 or 4 different GCS ... what is one more ?! :)

I'm currently using the included JST-SH_ 8pin-to-Dupont-RC RX cable with a FrSky-X6R and it seems to work good. I do find it strange that the manual and included cable steer you toward using a larger (non-SBus) RX, but they really don't provide much room to install it (between the FC and battery wiring) ... the X6R barely fits.



I think there might be a way to set-it-up with Sbus, but for now ... I used the included PWM cable, and connected each plug to a separate channel.

I bound the X6R to the Taranis with the "No Jumper Required" Mode-5 D16 CH1-6.

Model-04 at Taranis:
Mode: D16
Channel Range: CH 1-8
Receiver No: 04

Like my EB-185 and/or Tyrant-180 (they also have STM32-F1 based FC's) ... I setup basic Model Inputs as TAER, and Mixes as TAER1234 (1234 are your favorite 2-3 position switches on TX).

I installed and used LibrePilot v15.09 because that's what was shown in Hobbico Manual and I didn't want to mess with possible Firmware mis-match or upgrades right-off the bat. First thing I did in LibrePilot was to Export current UAV-Settings (including the Yaw-90 and 2-motor-swap). I think I also read in forums to not use the Setup Wizard (so I didn't, which is fine).

After I calibrated the TX, it let me Arm the motors to do a little motor-test on the bench.

Camera & VideoTX Install

I added a Foxeer HS1177 v2 600tvl CCD Camera and Aomway TX001 600mW 40ch 5.8ghz vTX.



Here are some pics of HS1177 Foxeer camera's mount. After test fit, screws must be cut to the proper length.




Installed the props for the first time. Armed quad and did a basic "control surfaces" run-test while carefully holding by landing-gear.




Maiden flight was hovering around the back-yard (aka, obstacle course) in the cold-dark night. It flies fine and fairly stable in Angle-Mode. Flew-thru a battery and did a little grass-mowing :-)






Eachine Blade 185 - RX upgrade and GPS install

I needed the FrSky-X6R for another project, so I installed a smaller FrSky X4R-SB in the EB-185. The X6R was really just for testing anyway (but I did fly it that way several times). I think the X4R-SB (also using SBus) is a better match. I mounted it with a square of black HobbyKing Zeal-clone. The blueish silicone tubing is just long enough to contain the 2 antenaas. It's wire-tied to the edges but away from spinning props.

Like before, connect cable from EB-185 PCB to the RX. That's the supplied cable with the little white 3-pin connector (SBus w/ Power) to FrSky X4R-SB using the black DuPont-RC plug on the other end.

On the Taranis:
Mode: D16
Channel Range: CH1-16
Receiver No: 03

I bound it in D16 Mode with Telemetry (CH1-3 SBus). That is the "No Jumper required Bind" and is comparable to the similar Mode-5 (D16) on the X6R/X8R.

Hold-down the F/S button on X4R-SB and power-it-up (plug-in EB-185 battery). IIRC, it's LED will be red. On Taranis' (Model-03) program screen, click Bind. Power-cycle EB-185. LED on X4R-SB should now be steady green (showing it's Bound). The EB-185 should also stop beeping now.



I also installed a GPS-unit I ordered for it a while back. They said it was for Naze32 (Rev6 10DoF) and had the proper connector already to PnP on this EB-185. I thought I would try it first on the low-deck where it's protected and there is plenty of room. I thought it might get a good-enough signal from the sides and even the top a bit. If not, I'll create and install a little platform for it behind the battery.

I plan to get back into CleanFlight and add some new Flight Modes and see what's up with the GPS features. I'll add to this post later if I come across anything interesting.



The Eachine Blade 185 is still flying good. My skid-plate mod seems to be keeping the VTX and its all-in-one PCB from getting damaged.

Thursday, November 30, 2017

PixRacer Bench-Testing and Setup

I flashed my PixRacer to ArduCopter v3.5.4 with Mission-Planner v1.3.50.

So, with nothing else connected, it seems this is the order of the two internal compasses
1. MPU-9250 IMU
2. HMC5983 dedicated compass chip (internal, on PixRacer PCB)

I say that because:
1. Calibrates with high offsets or sometimes completely fails to calibrate at all.
2. Always calibrates fine (but one axis is usually around +/- 200)

So, in the above case, you would Disable #1 and make #2 the Primary. Personally, I think it's good to try to calibrate just the Internal ones first (even if planning to add an External one). Not only as a preliminary test (find early which are troublesome), but also to help identify them later by extreme axis offsets.

When you connect an external GPS/Compass, it becomes #1, is auto-detected as External, and they all shift one number across.
1. External HMC5883L
2. MPU-9250 IMU
3. HMC5983 compass chip (internal)

It's actually pretty easy to see (watch for similar extreme offsets). So, in this case, I Disable #2 and make #1 the Primary. Coming from APM, I like the idea of having a redundant something if possible (especially something important like a compass). This is my current testing config.

Notice that with the PixRacer and latest ArduCopter, you can calibrate them from MP, but it's actually facilitated by the PixRacer FC itself. I just slowly turned quad on all 3 major axis, and then moved around randomly a little more.

I fixed this old Beitian GPS/Compass combo unit and it works great now (both the Compass and the GPS). I cut-off the old Pico-Blade connectors and soldered on the included PixRacer 6-pin cable.

I used this pic (thanks Gervais from RCGroups) to wire the GPS/Compass cable for the RixRacer and it worked the first time. Basically, you connect:
1. Power (VCC and GND) to Power ... "straight-across".
2. The two compass signals to the same compass signals on PixRacer (again, straight across).
3. It's the GPS's signals (TX and RX) that get criss-crossed to each other (TX-RX and RX-TX).


While new (and possibly a little hard to find) ... I'm really liking the new PixRacer connectors. They are smaller and with a higher pin-density that the old PixHawk-style Molex Pico-Blade connectors. Also, much easier to insert and remove (but still lock-into-place nicely).

Radio Setup
Like my APM and full-sized PixHawks, I bind (X6R and/or X8R) RX to TX in Mode-5 D16 (with Telemetry). I setup basic Model Inputs as AETR, and Mixes as AETR5678.

Note: I plan to add to this post as I come across something new or important.

Net Neutrality is GOOD - Killing it would be BAD.

In case you aren't aware, Net Neutrality is in danger of being repealed and removed by the FCC.

I think this is a good explanation in layman's terms. It also clears some confusion that some people might have. (The square icon in lower-right-corner switches the videos to Full-Screen)



Kinda strange because Jon never really uses this language in reports normally.



Not doing good work these-days guys. It's almost like Trump wants to undo everything Obama did while in office (no matter what the cost).

https://www.theverge.com/2017/1/23/14338522/fcc-chairman-ajit-pai-donald-trump-appointment


And don't think we are just talking about your landline broadband at home. This issue also includes your cell-phone data plan and mobile computing. A recent Stanford article reminds us that local news will also be affected.

An older video, but very good points. It seems this problem of Net Neutrality being adopted and staying in effect is a constant battle (Note: This video contains some adult language).



And now, John Oliver's latest video (and again ... John's language is Rated-R :)




All joking aside, if you want to keep using an internet with no-blocking, no-throttling, and no-paid-prioritization protections ... you should definitely visit www.GoFCCyourself.com (then click on Express) and let the FCC know that you support Net Neutrality. Then, just to be sure your wishes will not go ignored, you should also contact your government representatives.

Looks like sites and services like YouTube, Amazon Video, and NetFlix will be the first to be throttled-down or held ransom for extra cost plans.

Here is a newer video. The Cnet guys debate Net Neutrality ... I guess to try to present it from the ISP's point-of-view. Personally, I'm not buying it.


Another good recent article. You gotta read this stuff carefully friends :)

Upon finding updated but disturbingly unofficial source documents, David Gewirtz recants his earlier statement that the FCC changes are benign.


Time to save the Internet ... AGAIN!



MIT Media-Lab : What we lose when we lose net neutrality


We lost the battle but hopefully not the war. The December 2017 FCC vote was close at 3-2 :(

Update: January 17, 2018

Mozilla and 21 states Sue the FCC over internet freedom

=========================================

Counter Point 

I don't agree, but obviously some people think that Net Neutrality is bad (or it conflicts with their agenda). They think that government regulation of the Internet like a public utility (like they have been doing for the past 2 years) is a not a good idea. Even if that regulation is in place to keep it Open and freely accessible. As usual, some in the "conservative right" label any opposing views as "Fake News".

It's hard to debate or argue the issue without knowing how they see the issue or what they believe. So, to be fair, here are some counter-point links. These are the minority and not easy to find.

https://www.rushlimbaugh.com/daily/2017/10/13/millennial-asks-for-net-neutrality-explanation/

https://www.reddit.com/r/The_Donald/comments/7eqtrn/must_listen_rush_explains_why_liberals_want_net/

https://news.iheart.com/featured/rush-limbaugh/content/2017-11-27-net-neutrality-explained-via-professor-hazletts-great-book-the-political-spectrum/

https://www.rushlimbaugh.com/daily/2017/11/27/net-neutrality-explained-via-professor-hazletts-great-book-the-political-spectrum/

It's just my opinion, but I find the Rush Limbaugh articles a little hard to read. Some of his referenced tech history is mis-understood. Some of his basic explanation of how current technology works is also inaccurate. It's hard to "talk technology" or take tech advice from someone who doesn't completely 100% understand it. But it's SOME not All, so it's worth a quick read.

One law professor's overview of the confusing net neutrality debate:
https://www.washingtonpost.com/news/volokh-conspiracy/wp/2017/11/28/one-law-professors-overview-of-the-confusing-net-neutrality-debate/?utm_term=.93d4fec72250

Wednesday, November 29, 2017

Successful Nova-2017 Flight

We were watching the birds at the feeder today, and I realized there was no wind. I thought to myself "I should go maiden the re-built Nova-2017".

I went to the smaller Rec-Center field. Sure enough winds were 1mph. Got a good HDOP/SATs. Pretty nervous, but tried to remind myself I used to know how to do this :)

Flew mainly Stabilize and practiced landings. Tested the following Modes: Stabilize, Loiter, Alt-Hold, Simple-Mode, Land, and did a test RTL (set to 25 meters). One of it's auto-Landings was soft, but the motors ran too long so it tipped itself over once (not sure why the Nova does that sometimes). Anyway, I flew through my 3000mah battery without a hard-landing or really even a scratch. 

Some neighbors stopped to watch the show. I still got a little nervous when it was high, far-away, or near the main street or nearby houses.

Using Tower-Beta v4.0.1. It's nice flying a quad with a Power-Module again. Angle_Max = 4000 is fine (very fast), but THR_MID = 530 is not quite enough with 3000mah  battery installed (and Power-Module and SiK 915mhz radio). Land_Speed = 40 seems fine.

Here is the Nova after I got back (ignore grass-stains on props :)


Monday, November 27, 2017

Turnigy SK-450 PnF quad - The latest addition to the hanger

Here's the latest addition to the hanger. I caught it on sale at HobbyKing for around $100 (and with Black-Friday free shipping from USA warehouse) and could not refuse. Lets call it a birthday present :)


I plan to figure-out KK-2.x FC just enough to test the ESCs and motors. I've got some spare FrSky X8R RX's, so I'll drop one of those in there. However, I'm pretty sure it's getting a PixRacer and GPS pretty soon. I've be wanting to install the spare PixRacer in something for a while now. Actually, wanting to get a PixHawk-anything in the air :)

It's good to be back "in the hobby" more these days, but I need to stay away from hobby websites :)

Either I've been busy or the weather was bad, so I'm still looking for a good time to go fly the rebuilt Nova.

EDIT 12-2017
PixRacer Install


I got the PixRacer and Power-Module installed. Just using the single external Beitian compass/gps unit. FrSky-X8R fits nice up-front. Taranis programming was mostly the same as Nova/CX20. In fact, I just copied over the Nova model to the SK-450 model.

After clearing/working some strange Pre-Arm errors in Mission-Planner:
- Check Firmware - Already had ArduCopter v3.5.4. Flashed to ArduPlane and back-again
- Check FRAME-CLASS - Was X4, but switched it back-and-forth. Saved and Reboot
- Check 3D Accel - Calibrated it again. I guess required after major firmware switching
- Check Compass - Set to use single external one and Calibrated
- Check RC Roll not Configured - Re-Calibrated Radio (all sticks and switches)
- Check FS_THR_VALUE - Radio throttle stick goes to 982. Had to lower from 975 to 950.


I finally got the PixRacer Armed and spun the Motors for the first time. The ESCs also had to be calibrated.


It came together real fast. Before maiden flight:
- PixRacer needs to be mounted more prematurely ( I found my stash of Zeal gel-pad).
- More settings in MissionPlanner.
- Need to setup my first PixHawk (PixRacer) with my preferred APM params equivalents
- Wiring and antenna-mounting will need some clean-up

Notes:
WiFi AP is AI-THINKER_code


Thursday, November 23, 2017

Auto-Declination and verifying your Multi-Rotor's Compass Calibration.

After I properly and successfully calibrated the compass (magnetometer) in my Quanum Nova (APM-252 with single external compass) the other day ... I noticed that what it considered to be exact North was about 7 degrees off from what my analog hand-compass read. I was outside, with a GPS-Fix and Auto-Declination was on (which is best). I remember seeing something like this before but I was pretty sure the magnetometer (and whole quadcopter for that matter) was calibrated properly, so I thought I would investigate further.

I posted a message in the forums, and forum-member DKEmxr had this to say:
"Your hand compass will read Magnetic North. After calibration and auto declination correction the magnetometer will read close to True North. So it depends on where you are in the world as to how much this differs."


So, according to this article, Central Texas is about 7 degrees East Declination to Magnetic North (what hand-compass shows as exact North). Or, said another way, the hand-compass' Magnetic North minus 7 degrees = True North (what APM-based multi-rotor shows as exact North).


I also found this article.  For example, a 15 degree error, on a mile-long flight puts you a quarter-mile (440 yards) off-course or away from your desired destination. I think I will investigate buying a compass that takes declination into consideration. Until then, I will just subtract 7 degrees to measurements taken with my old hard-compass, or expect it to be 7 degrees off when comparing it to the Nova quadcopter or its software.

So, when the multi-rotor is facing its exact North at 0 Degrees, the non-declination-aware old-school hand-compass will read more like 7 Degrees (slightly East).

I'm pretty sure this applies to all other navigational systems on earth. I'm not 100% sure about DJI and other UAS, but it does also apply to PixHawk-based systems. 


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 (and with 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, unlike my old main Nova). I decided I'm just going to just "format it" and start over from scratch. 

RC Radio Setup and Config:

Using OpenTX-Companion, copied my old Nova model to a new one called "Nova 2017" 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.

Tuesday, November 21, 2017

Cheap 3D gimbals for sale (Tarot & Feiyu-Tech)

I was just browsing online the other day and came across these inexpensive 3D Tarot gimbals. I don't actually own one yet, but I remember that the Tarot name is good and it shows by the reviews. I've had good luck with this place over the years. No problems shipping from China and even acceptable shipping speed.

I noticed they also have the popular Feiyu-Tech 3D gimbals on sale. Check-out their accessories, it's like they thought of everything. For instance, no hardware hacking is required to switch between Xiaomi Yi cameras and others.

Speaking of cameras, if you don't have a GoPro yet, I think the GitUp Git2 is a good lower-cost alternative.

I guess it's that time of the year again. Some of these Black Friday RC deals look pretty good.

I've been working on my Novas a bit since Saturday. Between the two of them, I hope to get a single one flying soon.


Sunday, November 19, 2017

I've had better days at flying field. Rough Nova RTL landing with damage.

TL;DR - Pilot error and high-winds results in rough off-target RTL landing, with Nova's egg-shell body cracking in previously repaired spot.

With my life schedule and the weather, it's been a couple of months since I've had a chance to go fly.

I started watching the weather, and this weekend looked good, except for a cold-front blowing-in on Saturday. Since we had plans for Sunday, it was Saturday or try again next week.

On Saturday, it was Clear and 75f, but winds at 2:00PM were 22mph (with gusts to 30mph). I'm a little rusty, so I don't think I will do good enough in 22/30. And since it's likely faster higher-up, I'm not so sure about stock Nova/CX20 either. About 4:00PM I saw that winds had calmed to 12/20 so at the last minute, I decided to go for it after all. Over the years, I have successfully flown the Nova "in some wind".

At the school's big flat football field, it seemed really windy after all. I installed my small 2200mah battery. Within a few minutes I had 11 SATS and HDOP=1.5. All systems go. It was a little scary at first, but I flew through the first small battery successfully. I kept it fairly low, flew Alt-Hold, tested Loiter (noticed it fighting from being pulled down-wind),  and practiced my manual Landings in Stabilize.

Second battery was my larger 3000mah. Still had SATS and around HDOP=1.5. Flew around the field at little further out now, and trying some other flight modes on the Taranis. Then, I heard it ... "Proceeding to WayPoint-2". I was like "What The Hell is Happening!". I looked down and I had apparently accidentally flipped the Auto-Mission switch on the Taranis somehow.

At this point, the Nova was pretty high and heading for the back-corner edge of the school property ... the baseball diamond about 150 meters away. This was an old mission from a couple of years ago. It was programmed at high-altitude due to the metal high-power towers in that back corner of the school property. Across the street corner, there is a tall church and houses all around. I PANICED since I've never un-intentionally started an old Auto-Mission. I turned the Auto-Mission switch off and switched to Loiter, but the quad was now high and far away ... it seemed to be drifting away.

I tried to control it back closer, but it seemed I was making it worse. At this point, I'm pretty frazzled and I just want this flight to be over, so I switch on RTL and it starts coming back really fast (with the wind) but still pretty high since IIRC, my RTL altitude is set to 40meters (to prevent hitting anything along the way). When it gets above the Landing spot, it starts to descend finally. But of course the wind decides to start gusting. The quad is fighting it, descending, and drifting down wind.  It finally lands, but brushed-up-against a small tree (on the other edge of school property now) so the last few feet was a little too fast :). To tell you the truth, I was just glad the ordeal was over. Tower informed me of the crash (thanks Tower) and the Nova was still on and sending telemetry.

The only damage was the Nova's egg-shell body cracked and broke where I had glued it back together 3 years ago. The props, motors, ESCs , landing-gear, (everything else really) looks fine. If it wouldn't have been for that tree (in the way of that off-target RTL) I think it would have survived un-damaged.


So, I guess I have plenty of spare parts for my other Nova now :) Time to get that one calibrated-up and flying again ( EDIT DONE ). Also, a spare FrSky-X8R-RX for a different quad. Can you tell I'm trying to look a the "bright side" and stay positive? :)

I've tried to evaluate this flight and plan to be more prepared to do things differently next time. I also notice my ANGLE_MAX,3000 .  I had it raised a bit, but still kinda low to be docile. I plan to raise it higher since that would have probably allowed the RTL to land more on-target on the football field (even in the steady 20mph wind).



Friday, November 17, 2017

Mission Planner v1.3.50 for Windows (for APM and PixHawk)

Well, I recently posted about the other GCS I use (or plan to use) so I thought a post about the main one I use would be in order. Glad to see Mission Planner (for Windows) is still going strong (thanks to Michael Oborne , all the project contributors in the community).

I came across an interesting glitch late last year. It has apparently since been fixed in the latest version of Mission Planner (v1.3.50), but I thought I would post about it anyway.

With the old version of Mission Planner v1.3.41 (current back in October 2016) I tried to load it on a fairly new Dell laptop (with Intel mobile CPU and Intel on-board HD4000 IGP) running Windows-10/64bit. It seemed to lockup while trying to initially load and start running. I didn't think much of it and just put that laptop aside. Instead, I used my slightly older AMD-based Lenovo laptop (also running Windows-10/64bit). It ran fine there, so no worries and back to flying :)

I recently found out that it wasn't actually locking-up on the Intel-based Dell, it just goes unresponsive for about 4 minutes and then finally loads, initializes, and runs properly (if you wait-it-out). I was still running v1.3.41 and holding-off on updates. However, Google searches reveal other pilots having similar problem with versions as late as 1.3.45 and 1.3.49.

So, in the v1.3.50 Change Log, we see:
HUD: fix Intel graphics freeze

However, it might have been a different issue that was fixed. Anyway, its fixed and now works on both laptops. The Intel-base one is running Windows-10 Pro 64bit Version: 1607.

To get Mission Planner v1.3.50, you can update from inside the Mission Planner program itself, or just install the latest version from the installer-file.


Thursday, November 16, 2017

QGroundControl v3.2 - Initial Impressions

Looks like the QGroundControl is coming along and maturing nicely. I finally had a chance to load this open-source GCS on my Android tablet.



I think it looks great. I think I also prefer the UI and compared to what Tower has become, including the representation of the fundamental Aircraft Instrumentation.

Amazingly, it connected right-up to the APM in my Nova over the SiK-915mhz radio ... and the UI came to life with real-time telemetry from the aircraft. The Android-girl was even calling-out Mode changes as I tested the Taranis.

However, I immediately got a pop-up notification warning-message that v3.3 or higher is required for 100% proper operation. I assume it means ArduCopter because this little APM-252 FC is just running v3.1.5. Ground testing seemed fine, but as I explored the app, I continued to get various other warnings about yet other parameters being missing. It's not a fault of the App, just that this version of ArduCopter isn't new enough (and can't particular FC really be upgraded to v3.3.x).

I still plan on using Mission Planner for Windows for quadcopter and drone setup, calibrations, and parameter tweaking ... while at the workbench. While this seems to be barely working on my old APM aircraft (on the ground) I think I will stick to Tower 4.x for them. I'll reserve this QGroundControl for my PixHawks.

Saturday, November 11, 2017

Tower (for Android) 4.0 and Tower Beta 4.0.1 - GCS Testing

Tower v4.0.0
Well, the first couple of times I ran the Tower GCS, after a few minutes it just quit while it was running and dropped to the tablet's HomeScreen. While later researching Tower for this post, I came across some post by other pilots describing similar crashes. Not sure what's up with this, but it's rather off-putting ... I use Tower during most flights and I can't have it crashing. I don't recall any old versions over the years crashing for no reason.

Since they aren't required any more, I tried a long-shot fix ... I un-installed 3DR-Services, and rebooted. Then, and un-plugged/re-plugged the USB cable for the 915mhz SiK-Radio and answered the prompts to use Tower as default App.

Luckily, it seems to run fine now. I ran in many times and no more crashing (but I'll watch it closely). In fact, it works fine now except no RSSI is shown (under the triangle-looking icon). Just confirming a know bug.

Tower Beta v4.0.1
Installed from Google-Play store and rebooted.
Un-plugged/re-plugged the USB cable for the 915mhz SiK-Radio and answered the prompts to use Tower Beta as default App.
Set my desired Settings and Preferences.

Connects fine and NO crashing to HomeScreen... seems to run fine.
However, there is still no RSSI shown. Interesting, because Release Notes say that it was fixed by the new Developer/Maintainer. I was hoping it would work, but I did read a forum post by another user that said it was still broken for them also.

Hmm. Maybe there is a newer compiled version in .APK file, that just hasn't been vetted or uploaded to Google-Play yet?

There is not really a full manual for Tower that I've ever seen. If you are new to Tower, or just need a little refresher course, these 4 videos by RNDrones seem informative and helpful to me. They might be for a 2016 version, but other than not needing 3DR-Services now, they should work fine with v4.x .


Thursday, November 9, 2017

What's going on with the Tower app (for Android) lately ?

As you might remember, I use the Tower app (previously known as DroidPlanner) while flying my APM-based quads.

So, I'm trying to spend more time with this hobby lately. One of the things I was checking was what software I should be using, and that everything is up-to-date.

I don't really use Android for anything other than this R/C Tablet, so bear with me (in real-life I live in Apple ecosystem). For some reason, the "auto-update-apps" setting in Google-Play Store was off on my R/C Android tablet, so I turned it back on and various Apps updated. I'm not sure if my last flights were on a late release of Tower 3.x or v4.x.

When I started Tower the first time, I got a "Change-Log" screen. Among other things, it said that the dependency on 3DR-Services had been removed. At the time, I didn't think much of it, other than the fact that my R/C tablet's software-build could now possibly be a little leaner.

At first, I thought the Tower v4.0 release was new. However, after doing some research, it looks like v4.0 was released back in Q4-2016.

Maybe the removal of 3DR-Services had a deeper meaning of things to come?

It appears that Bill Bonney  https://github.com/billbonney has come to the rescue.

https://discuss.ardupilot.org/t/tower-4-0-1-beta-released/17037

Bill, thanks for Tower Beta v4.0.1 . I think this might be the main GitHub dev page

https://github.com/billbonney/tower

Looks like its in the Google-Play store, so we doesn't have to hassle with a .apk file.

https://play.google.com/store/apps/details?id=org.droidplanner.android.beta

The main Tower dev page is here.
https://github.com/DroidPlanner/Tower

It appears to being updated slightly recently, but a new version with those updates is not always uploaded to Goggle-Play-Store. Evidence of that is its Tower is stuck at v4.0.0 since 10-2016.

https://play.google.com/store/apps/details?id=org.droidplanner.android&hl=en

Next, I need to fire-up the Nova and Tower v4.0 and see if the RSSI for my SiK (915Mhz) Radio is truly missing. I do like seeing that . Not just when flying far away, but also as a Pre-Flight check.

Tuesday, November 7, 2017

Cheerson CX-20 still for sale

Looks like the Cheerson CX-20 (APM Open-Source version) is still for sale.

If you are still flying this nice little brushless quad-copter, now might be a good time to pick-up another RTF package. You will get a spare battery and lots of spare-parts (landing-gear, body shell, motors, ESCs with LEDs, APM-FC, etc.).

Shipping from USA-warehouse will likely be pretty quick.

8% OFF Coupon Code: bggcx-20 (special limited-time code for CX-20 Quadcopter)

If this is your first CX-20/Nova, you might want to read my blog's archives for tips. Like ... throwing-away the included cheap battery-charger (and buying yourself a "real" LiPo charger for $30 usd) :) The LiPos for your new hobby will thank you.

UPDATE:

Right after I posted this, I found another tip posted by DKEmxr in the forums. Looks like HobbyKing still has some Nova's left as well. It's the PNF package (no battery or radio) but it's only $129usd. Remember that the Cheerson CX-20 is the quad that HobbyKing cloned to make the Nova.

Monday, September 11, 2017

I'm back. My weekend Flying Session and Multi-Rotor Sale notification

I bet you all thought I disappeared from the scene :) No, I'm still alive. It's just that "life" has been getting in the way of my hobby :) It's been really crazy around here lately.

I think I got to go fly my quads like maybe 4 times all spring/summer. FPV racers are mostly just sitting on shelf and custom-build 550 hexa is still waiting for assembly. I have been maintaining my LiPos though (or, at least checking them).

It was kinda windy this past weekend, but the weather was otherwise good, so I decided to take the Nova over to the middle-school for a little flying.

I flipped though the Taranis switches and tried to remember how to do a proper pre-flight and actually fly this thing. It took a while (updating GPS almanac ?) but I finally got a good GPS lock (sats and hdop). I thought it would be OK if I use Manual-Mode sparingly ... instead relying on Loiter, Altitude-Hold, etc.

I had a couple of scary moments, but I flew-thru 2 batteries without any major damage. I also practiced my landings. I was glad to see I could just go fly the Nova after all these months. Not just with my rusty skills, but also that the Nova was still calibrated. Not sure if the version of Tower on my TX-attached Android-tablet is the latest, but it seemed to be working (but didn't try any auto-missions or anything).

Hey, one of my primary affiliate partners just notified me of a limited time sale on multi-rotors and drones. This link and code are only good for a limited time, but if you need something, you might want to check it out.


Extra 20% Off For Multi-Rotor & Accessories DealsUp To 71% Off For FPV Racer and Multi Rotor Parts.



Saturday, January 7, 2017

Electronics and computerized machines fail early due to Lead-Free Solder and RoHS techniques

Have you ever wondered why electronics and computers from the 90's and early 2000's seem to last forever? But now, you are lucky to get 3 years from just about anything computerized. Now-days, just about any machine (including vehicles, boats, and airplanes) have some kind of logic-board or PCB installed. From coffee-makers to TVs, computers, and spacecraft ... and everything in-between.

So now we know why. It's not just a conspiracy theory. Are the machines you purchased any less valuable or important that the ones that NASA's and the US-Military bought?

Lead-Free Solder - A Train Wreck in the Making

The advisory concludes with this guidance: “Though there are many alternative solder alloys available to replace traditional tin-lead, none of them has passed the reliability testing required of aerospace-quality hardware.”

Killed (Not Saved) By A Whisker - Why Our Electronic Gizmos Inevitably Die

The military and NASA were especially concerned, so they generally operate under waivers allowing lead solder in the gear on which our space program or national security supposedly depend.

Military Suppliers Confront the RoHS Challenge

The two main reliability concerns with lead-free processes are solder joint reliability and tin whiskering ... Most discrete parts and parts with leads can withstand the temperatures of both leaded and lead-free processes, but BGAs can be a problem.

Hazardous Materials Statement - Amphenol Aerospace (PDF)

Due to stringent customer and Mil-Spec requirements, AIAO will not pursue 100% transition to Pb-free products ... MS and other mil-specification part-numbers requiring leaded solder will not be made available in lead-free or RoHS compliant versions until the applicable MilSpec is revised to incorporate this change.

I think this inferior unleaded solder is killing more electronics than we realize. Reliability and longevity seems to have fallen from around 95% to around 60% ... on just about everything. I used to think it was just poorly executed Lead-Free solder techniques, the additional solder-heat weakening the soldered components, and more recently ... a large industry shift to BGA components. While I'm sure all this is still true, now we know the main reason. It only takes one bad/cold solder connection or weak PCB multi-layer-via to render a machine inoperable. Your thoughts?