I have 2 problems with the Ardupilot Mega 1.
Since 2 days I am struggling with it.
1. When hovering, it sometimes makes a turn at the yaw axis for
several degrees. It is a quick turn, like jerk for about 20-30degrees.
When I use the terminal test functions and enter compass, it shows me
a lot of good values, and then some lines with "not healthy" and
sometimes wrong values.
When I remove the GPS module, it works fine. Now "not healthy"
messages, and in flight, it does well.
2. When I setup PID values for the first time, I used the channel 6
function to tune the values in flight. This worked very well. Now,
since 2 days, it does not work any more. When I set a value on channel
6 and set the range for example to 3 as minimum and 7 as maximum, than
use the pot on my transmitter to change the value, the value becomes 0
or 7. Nothing between.
Do you have any idea what could cause this problems?
What could I do?
yaw issues are 9 out of 10 times compass related.
a. disable the compass & gps. if yaw is solid then lots of possibilities. If it still yaws, then it's ESCs, motors, or PIDs. And check in that order (ESC throttle cal, motor axis angles/bad props, yaw rate P). If it flies solid, then it's usually a power or interference issue (especially the 'not healthy' message).
b. if you re-enable the compass and the problem reappears--the not healthy says 2 things:
power, magnetic interference, or i2c interference (yes, it's present with a 900Mhz telemetry), Check in that order.
+ power: ensure the compass is getting 4.985V or higher at the Vref/Gnd pads of the compass module. Anything lower will cause the "not-healthy" issue (had 2 boards with that problem). Also check for cracked/releasing pads--the compass connection is NOT robust. (Also the telem port ground connect is not robust as well!). If you're getting undervolts, then you can check the air speed port or telem port for a good 5V and wire tap them to the compass directly to power it--have fixed one board this way & it's a hack solution...
I doubt the i2C connections are bad, but you may want to check continuity between the SC lines to the back side of the APM (not oil pan) pins.
+ mag interference: take your props off. connect the APM to mission planner, record the compass reading in the HUD disarmed. Arm your motors and power up to 40% throttle. If the compass drifts more than 10degs--mag interference is your problem. Move power wires and ESCs away from the oil pan... repeat until you hardly get any drift.
+ i2c interference: move your RF equipment away from the APM.
Hope that helps...
Cliff, thanks a lot for your reply. I will check this all tomorrow.
Today I updated the firmware to 2.6. The not healthy still exists, but it doesn't show up in flight any more. So I think 2.6 filters this false informations out and doesn't use it.
So tomorrow, I will check all your suggested possibilities.
Many thanks and kind regards
Great reply, Cliff-E. Really well said.
Cliff, just measured the voltage on the board. No matter where I measure, I get 4,65V on the upper board (blue) and 4,70V on the lower board (red).
When I measure direct at the RC pins, I get 5V.
So there must be some problem with the voltage on my board. I am not an electronics expert, so I do not really know where to look for the problem. So I will try to measure all V and GND pins I can find on both borads and hopefully I find the place where the voltage disapears.
Is there any description, where a non electronic expert can find a hint where to search?
I found 2 points on the red board marked as VIN + and - with & to 12 V. So I soldered to pins on it and connected a 2S LiPo for testing. Now I have 5V all over the board. Looks great. IIf that works, I will connect this directly to the PDB.
Will test later, lunch is ready now;-)
I powered the board using the BEC of the first ESC.
Now I discovered that it is much better (5V everywhere), when I power the board from the VIN 6V-12V on the red board.
I tested it using a 2S LiPo. Now I am thinking about connecting it to the PDB, where I use 3S LiPos. Is there a problem, when it gets 12,5V from a fresh LiPo?
Is it better to use a seperate 2S LiPo to power up the board?
It is normal to have some loss due to resistance and connectors. You may have more loss than is ideal from some cold soldered joints, at the servo rail, and/or between the boards.
I would not recommend powering directly off a battery anywhere. I would personally prefer to use a switching BEC, like Castle Creations. You also don't need the extra weight of a second battery. This can be wired into the same flight battery you use.
Castle Creations, and many high quality BECs, are programmable. For that product, you'll want a castle link device to set the output voltage.
You can adjust the output voltage slightly, but do not get too close the 5.5v. Personally, I would not generally increase the voltage to more than about 5.25v. Or you can use the VIN, and bump the voltage up higher, to 6+. But even if I did that, I would keep the voltage down just above 6v and I would use a BEC. I'm don't remember what the voltage regulation circuit looks like from VIN, and you want to avoid too much down-conversion-through-heat, as that is just wasteful. Another reason I like a switching BEC over what are too often linear BECs built into cheap ESCs.
But there are lots of ways to do this, and so you'll want to consider different opinions from my own.
Hello Cliff and Mike and the others that are reading this.
It seems, that I am a real donk. After lots of testing and writing to several forums and reading a lot of articles, tody I discovered the reason for the compass not healthy problem.
I red the assembly article in the wiki, as I did on the first time. I setup the APM1. Then I saw this :"Finally, attach your GPS module as shown below (MediaTek module shown). It goes in the connector on the APM board, not the similar one on the IMU shield (that one, which says "No GPS!", is an I2C connector for the optional magnetometer or other I2C sensor)."
The first time I setup the APM1 I did it right, but now, when I unplugged everything and reconnected everything I forgott to connect the gps to the lower board. I connected it to the upper blue board. Now everything is fine and works like a charm.
Sorry for my dullness and sorry for bothering you all.