# EKF errors, catastrophic tumble, need help understanding log **URL:** \ **Category:** Cube Autopilot\ **Created:** [8 October 2026 21:21 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974 "2026-10-08T21:21:02Z")\ **Posts on this page:** 5\ **Page:** 1
**Author:** ![pauljfpv](https://discuss.cubepilot.org/letter_avatar_proxy/v4/letter/p/cc9497/32.png) [@pauljfpv](https://discuss.cubepilot.org/u/pauljfpv)\ **Post date:** [8 October 2026 21:21 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974/1 "2026-10-08T21:21:02Z")
Hi, Was testing out a new DIY heavy-lift quadcopter build with the following configuration with Mission Planner, but after flying fine for a few minutes hovering in Loiter mode, suddenly the quadcopter started to tumble uncontrollably down and pancaked on the ground. I don’t understand logs yet, but did notice a couple EKF errors in the attached log, so please help identify what caused this to happen and how to prevent/correct it, thanks: [2026-10-08 12-10-51.bin](https://discuss.cubepilot.org/uploads/short-url/7jNUucdWlIG3SkEUuYEcikmrrZn.bin) (6.8 MB) CubePilot Orange+ Here4 GPS RFD 900ux-US 24" wheelbase 60 amp ESCs 4320 350KV motors 2x 6S 6500mah batteries I had everything calibrated and I’ve completed initial tune successfully, had made a handful of successful flights, and was testing out a digital camera/VTX. I also verified the batteries still had plenty of charge (23.4 volts). But I noticed the following messages in Mission Planner right around the time it crashed: 10/8/2026 12:15:12 PM : EKF primary changed:2 10/8/2026 12:15:12 PM : EKF3 lane switch 2 10/8/2026 12:15:12 PM : EKF primary changed:1 10/8/2026 12:15:12 PM : EKF3 lane switch 1 Any help will be appreciated. Thanks, Paul ---
**Author:** ![xfacta](https://discuss.cubepilot.org/letter_avatar_proxy/v4/letter/x/5f8ce5/32.png) [@xfacta](https://discuss.cubepilot.org/u/xfacta)\ **Post date:** [9 October 2026 06:01 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974/2 "2026-10-09T06:01:01Z")
I would say the cause is a physical problem with Motor4/Prop4. Look for anything like a broken prop or mounting issue, even a twisted/loose motor mount or broken arm. Since both motor4 and 2 are effected, it could even be a serious weight shift to the rear, such as the batteries coming detached - but you probably would have seen this, and they may have disconnected too (which was not evident) I’m fairly certain this is not an ESC desync though, as that usually takes some time to recover and is more evident - the suspect motor/ESC would be commanded to maximum for a longer period of time. You can see in the graph that pitch and roll go bad slightly after Motor4 is commanded to maximum (a sure sign of no thrust) and also motor2 is commanded near maximum possibly indicating a problem with both rear motors or props. ![image](https://discuss.cubepilot.org/uploads/default/original/2X/a/a888e2681eacd9b20b4cfd54c77011b935ae8b91.png) The EKF lane changes only come after the copter is tumbling down and the flight controller is trying anything it can. You dont have any battery monitoring configured, and this would probably have helped diagnosis. Not only that, battery voltage is **essential for safety** , and current is desired for monitoring and diagnostics. For reference: ![image](https://discuss.cubepilot.org/uploads/default/original/2X/1/1a68d6aefb39ac96fea8b88910126467409030a1.png) ---
**Author:** ![pauljfpv](https://discuss.cubepilot.org/letter_avatar_proxy/v4/letter/p/cc9497/32.png) [@pauljfpv](https://discuss.cubepilot.org/u/pauljfpv)\ **Post date:** [9 October 2026 17:10 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974/3 "2026-10-09T17:10:12Z")
Great info, thanks for the analysis Shawn! I verified that the clamps where the 2mm thick carbon fiber arms attach to the main body plates (also 2mm thick) were tight before flight, and I thought that because I had the arms “locked in” with an extra bolt through the body plates and tube, that it would prevent the arms from twisting in the clamps. But when it crashed I noticed that a couple of the arms had rotated in the clamps about 10 to 15 degrees from level. I thought this was from the crash, but it looks like those clamps could allow the twist in flight… The copter was hovering in Loiter at the time and I was starting to fly it back to home and that is when it upended and crashed… It had a takeoff weight of under 9 pounds (with twin 6S batteries and standard ESCs instead of a 4-in-1), so I need to find better clamps or a better way to attach the arms to the main body plates for a heavy copter like this, but it’s been difficult to find a better mechanism. Any recommendations? I tried a couple large frame kits, but they also use the same clamps and have the slippage issue… Also, I tried setting up battery monitoring a few times with the Power Brick Mini that comes with the CubePilot Orange+, and it shows voltage (battery percentage is not accurate), but very soon it says Bad Battery and goes into battery failsafe, even though the batteries are still nearly full charged… Looks like I need to get one of those Mauch sensors… Any recommendations for a model that will work with the CubePilot? Thanks! Paul ---
**Author:** ![pauljfpv](https://discuss.cubepilot.org/letter_avatar_proxy/v4/letter/p/cc9497/32.png) [@pauljfpv](https://discuss.cubepilot.org/u/pauljfpv)\ **Post date:** [9 October 2026 18:14 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974/4 "2026-10-09T18:14:27Z")
Thinking about the frame some more, I bet I could get a better arm to body attachment by going with square carbon fiber tubes instead, and not use clamps (just reinforced with carbon fiber plates and bolts at multiple places to form the attachment)… ---
**Author:** ![xfacta](https://discuss.cubepilot.org/letter_avatar_proxy/v4/letter/x/5f8ce5/32.png) [@xfacta](https://discuss.cubepilot.org/u/xfacta)\ **Post date:** [10 October 2026 19:52 UTC](https://discuss.cubepilot.org/t/ekf-errors-catastrophic-tumble-need-help-understanding-log/17974/5 "2026-10-10T19:52:50Z")
The mini power brick will work OK if the overall current draw is within limits. There will be a parameter we need to adjust if you get the Bad Battery message.