It's so sad to see all your hard work fall from the sky like a rock.  This past Friday after work, I tried to fly my usually circle flight plan at a field next to my work.  Everything seem normal, I switched to ALT HOLD and then switched AUTO and off it flew to waypoint 1 at 20 meters right in front of me.  It had a 10 second delay, which it never finished.  I'm not sure yet what happened, but it just fell like a rock.  It landed upside down and damaged 3 arms and 3 motors, plus the stack on top broke all it's supports.  About $90 of parts, which I ordered Saturday.  And, I'm happy to report my order from 3DR has shipped tonight, I should see them Friday. I'm really happy they turned my order around so fast.  Brownie points to 3DR and staff.

Now can someone give me some pointers on what I should look for and test to see why my copter failed?  I really don't want to fly again until I figure out what went wrong.  Last week I had tuned my gimbal thinking it was causing power problems and some weird twitching and 1 minor crash. I recently just added 3DR power module and pulled JP1, hoping this might help with power issues.  I had 2 successful flights and now this crash.  I also have a 5 amp BEC plugged into the output side of APM, plus 3DR PM.

I'm sorry to say the GoPro failed to capture the last few seconds of the flight.  Bummer, the last time I had a brownout the GoPro showed a great tumble to the ground.

I've attached my tlog, kmz and log files for review.

Firmware is 2.8.1.

Recently balanced props and mounted APM on foam corners.  I was just about to upgrade to 2.9.1. :-(

Thanks for the support.

Richard

Views: 3260

Attachments:

Reply to This

Replies to This Discussion

wow that was a high fall hope you can get it back together 

Do you have your 5V supply do your servos for your gimbal connected to CH10/CH11 (ie. A10/A11)?

See my other reply to you here http://diydrones.com/xn/detail/705844:Comment:1125129

Even with the PM the Inputs 5V supply and the A0-A11 is limited to MAX 500mA due to an in circuit solid state fuse. 2 servos will draw in excess of 1A, so can cause a brownout.

Maybe that explains the crash?
My gimbal controllers are hooked up via ch10 and ch11. I have JP1 removed and 3DR power module plugged in. I also have an external 5 amp BEC plugged into pin 1 on the output side. Should I have my BEC plugged in on the side where my gimbal controller pins are plugged in? Would this then just power the extra stuff like gimbal and LED lights, where the APM gets its power from the 3DR PM?

The question is, where do those servos get their power from?  You say the gimbal controller is plugged into 10 and 11, is the gimbal controller passing power through to the servos from Ch 10-11?  If so, that's likely your problem.

You are power the gimbal servos via the APM PM. You should only connect the signal lines to ch10 and ch11 when using the APM PM.

If you plug you BEC into the inputs side this would power the APM and gimbal servos. But not power the output. You would also need to disconnect the PM or cut lines 1 and 2 and the PM power connector.

The reason for the problem is that there is a 500mA fuse between the PM 5V out and the 5V to the APM 5V inputs rail. This cuts out at 1000mA two servos are 400mA each and the APM is 250mA.

There's another solution that will work and that is to short the fuse. I personally think it not required. It looks to me that the idea was taken from the fuse for the USB port. But that served a different purpose. Protection of too much draw from the USB port. The fuse from the PM limits current from the PM to 500mA with JP1 removed.

And another With JP1 in place you get 5.3V from the PM on the outputs 5V rail. You could do that and utilize the extra capacity for two servos. I would power any more than two as you get too close to capacity of 2.25A. Though don't connect you BEC in this case.

I think that fuse is part of a Crowbar circuit designed to protect the system against over-voltage being applied.

I just looked at the design and you are right. It looks like the Zener starts to draw the extra current, and when the over voltage triggers the circuit to pull more than 1000mA (500mA rated fuse) the solid state fuse will trip.

The unfortunate consequence of that is that with JP1 removed is the PM is pretty much locked to 500mA and with the solid state fuse being a resistive load, Vcc changes depending on current. (bad for ADC measurements)

Maybe for the crow bar to work and still have the PM rated at it's 2.25A you would need a Zener 6V@2W with 2A fuse.?

Can the PM actually operate at 2.25A?  Thermally?

A more sophisticated solution, but costly and complicated would be a UVOV controller like this:

http://www.linear.com/product/LTC4365

see http://www.ti.com/lit/ds/slts235c/slts235c.pdf

It states 85DegC at 2.25A, so I don't see why not. It needs to be tested to obtain the real A that can be delivered. But i definitely think you can get more from it. Two servos and the APM (1000mA + 250mA) should be no sweat. Anything more would be a problem. The only issue is where you connect the servos. That crowbar circuit is the issue, since A0-11 and Inputs 1-8  5V lines are connected to Vcc.  A9/10/11 which are used for servos as well should really be connected to the Outputs 5V rail, not the PM.

Maybe some solder pad for a choice on the bottom of the board would be useful to select 5V Outputs for 5V inputs for those extra servo channels?

So what to do?

I fly APM2.5 and plan on including a camera gimbal with two servos.

After reading about a crashed caused by the drawn of too much power I think about using a second UBEC (5V 3A) (http://hobbyking.com/hobbyking/store/__4319__TURNIGY_3A_UBEC_w_Nois...)

just for +5V of the servos. Another UBEC powers the APM2.5.

Will this work?

I mean: Ch10 / Ch11 just GND and Signal and +5V from UBEC?

@CyberCrash,

Yes, that combination will work. Just make sure that you have jumper J1 removed so that the input and output power rails are split.

Regards,

TCIII

I see possible four options to power the APM with camera gimbal servos of CH10/CH11 (and same applies for trigger servo of CH9/A9)

Option 1) Dual Regulated Supply APM PM & ESC with UBEC

With JP1 removed with APM PM. 5V/GND from PDB to Output, you can connect the servo's 5V & GND to the Outputs 5V/GND pins. Take the signal lines only from Ch10/Ch11

Option 2) Dual Regulated Supply with separate UBEC (or BEC)

With JP1 removed, No APM PM, 5V 3A UBEC on Inputs (no need to separate power for CH10/CH11 in this case (if you connect the APM for I and V sense, cut line 1 & 2 of the connector to the APM, this is the 5V line)

Option 3) Single Regulated Supply from APM PM

With JP1 in place, APM PM, PDB 5V not connected to outputs. Connect 5V/GND from servos to Outputs 5V/GND pins, and signal lines from CH10/CH11 (Note do not use more than two servos connected to Outputs with this method as it uses the APM PM and it's limited to 2.25A total (This suggestion uses 1.25A nominal))

Option 4) Single Regulated Supply from ESC with internal UBEC

With JP1 in place with APM PM. 5V/GND from PDB to Output, you can connect the servo's 5V & GND to the Outputs 5V/GND pins. Take the signal lines only from Ch10/Ch11. (if you connect the APM for I and V sense, cut line 1 & 2 of the connector to the APM, this is the 5V line)

In summary the only time you can power a servo directly from CH10/CH11 5V pin is when you power the APM from the inputs using a UBEC.

Reply to Discussion

RSS

© 2018   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service