Flight controller loses power during High current draw. Any ideas?

Thread Starter

bigjoncoop

Joined Feb 1, 2019
73
I'm in the racing quadcopters and I have a bunch of electronics that don't work for one reason or another and when it comes to Tiny SMD components it's a tall alien to me.

I have a brand new flight controller that dies if you go about half throttle which is pulling about 40 50 amps. But everything else connected to it still powered. So basically the 5 volt regulator and 12 volt regulator are still working it's just the processor itself the 3.3 volt stuff loses power and will not work until you unplug the battery and plug it back in . Then it works fine until you again give it more than half throttle.

Any clue where I should look or Focus my attention.
42945323-0261-489d-ae74-a43c5755d2e1.jpg42945323-0261-489d-ae74-a43c5755d2e1.jpgMatek-MATEKSYS-F405-AIO-BetaFlight-F405-CTR-STM32F405-Flight-Controller-Built-in-PDB-5V-2A-BEC.jpgMatek-MATEKSYS-F405-AIO-BetaFlight-F405-CTR-STM32F405-Flight-Controller-Built-in-PDB-5V-2A-BEC.jpg
 

MisterBill2

Joined Jan 23, 2018
5,228
If that high current is passing through any part of that controller board shown in the photos then I suggest first considering the battery supp[y impedance.
The two things that will upset a processor are power dropouts and supply voltage spikes. If there is any way to power the controller logic from a separate power source that might be a solution. A second possible fix may be to add an external "super capacitor", 1 or 2 Farads, across the 3.3 volts line on the control board, to keep the 3.3 volts constant. There might also be a heat issue someplace. That is a huge amount of current for a circuit board assembly.
 

jpanhalt

Joined Jan 18, 2008
8,765
With electric-assist sailplanes, we usually used a different regulator for the radios/servos and motor. The built in BEC for the motor just didn't have enough oomph for 6 hardworking servos.
 

Thread Starter

bigjoncoop

Joined Feb 1, 2019
73
The current passing through that board is normal these quads pull an excess of 100 amps burst constantly all day everyday.

And we do use 1000uf low esr caps.

These boards always handle that amount of current and beating. Just this one in particular is having this problem. So I am asking which component on the board would be the likely culprit not how to solve running so much current do these boards or how to power them

Thx
 

jpanhalt

Joined Jan 18, 2008
8,765
The current passing through that board is normal these quads pull an excess of 100 amps burst constantly all day everyday.

And we do use 1000uf low esr caps.

These boards always handle that amount of current and beating. Just this one in particular is having this problem. So I am asking which component on the board would be the likely culprit not how to solve running so much current do these boards or how to power them

Thx
You have not given a whole lot of information about your set-up. The obvious answer, look at the component that cannot provide more that 40 to 50 A and the current needed by the servos. In the models I referenced, we always looked for at least 10 A for the servos.
 

Papabravo

Joined Feb 24, 2006
12,953
You may think those boards are capable of handling that current load, but it is not obvious from the pictures why that might be the case. What you have is a multi-layer board with ordinary size vias and pads. Maybe the fabricator used extra copper on one or more layers, but again that is not obvious. You have a problem with at least one board, so you have to start eliminating culprits. I woud start by instrumenting a board to actually measure and record both the current draw and the voltage regulation. Switching regulators are notorious for having poor transient load performance. I cannot imagine anybody using linear regulators with that kind of current draw.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
5,228
If other identical boards cab function correctly in the same arrangement then the first thing to suspect is a connection that is not quite good enough. That may be an inadequate solder joint of a surface mount component or an inadequate vie in a plated through hole. So a stub of wire soldered into the several plated through holes might be a solution. That sort of flaw would be very challenging to find, but a resoldering of the unused plated through holes on a board that size should take a soldering master less than 5 minutes. An intense inspection of all the component attachment joints, and possible resoldering, may take half an hour or more. If the installations are identical then that is my suggestion.
BUT if there is some wiring difference then that should also be examined closely.
 

Chris65536

Joined Nov 11, 2019
164
Maybe the controller is falsely sensing an overcurrent, and shutting down? I'm guessing that the 0m50 component is the current sensing resistor. It could be that's it's not soldered perfectly, so the shunt resistance is higher than it should be. If you dare, you could try re-soldering that part.
 

Thread Starter

bigjoncoop

Joined Feb 1, 2019
73
Hey guys it's great to see so many people chiming in. Let me explain a little bit more on how this board works and where the high current is going.

This flight controller is my go-to board when I don't want to worry about it and just want something that works. It's also true for many other people.

So that high current draw is not going through any voltage regulators etc... that just passes through the board from the battery input pads straight to the ESC/motor pads. The only thing that the board is doing with that is current sensing.

There are two voltage Regulators on board. One of them is 5 volts to power camera, receiver, video transmitter etc.. that cannot handle battery voltage.
The other voltage regulator is 3.3 volts which Powers the STM32F405 micro Processor and the other IC's...

The 3.3v regulator or that circuit is what seems to be failing. When the board exhibits this problem that its having, the 5 volt regulator is still working and providing power, and of course the motor outputs still have power because they're just passed through from the battery connection.

so I am assuming that the 3.3 volt regulator circuit or whenever is attached to it is the problem area. So what I'm asking is if anybody could maybe look at the picture and point out an area or component that normally goes bad when there's a lot of noise or something like that. Just to help narrow down my search because I cannot really test it on the bench.

I guess I can connect to 50 amp test load to one of the motor outputs and raise it if need me until the processor turns off and then check to see if it's still being supplied 3.3 volts and if not then I know it's coming from the actual regular.

normally I would just throw this thing in the garbage or keep it for parts and buy something else. The problem is that we fly these quadcopter so hard we break things on a daily basis and are constantly buying stuff. This board is brand new was never flown before and this problem is happening and I've had it for so long sitting in my supplies I cannot return it for warranty so just be nice to get it to work
 

MisterBill2

Joined Jan 23, 2018
5,228
OK, the explanation does help. Is the board a multilayer PCB, or just a two-sided one? What will really upset a regulator is impedance in the leads, especially the supply or ground leads. AND the small capacitors between input to common and output to common are quite important. So if those 0.1 mfd caps are damaged, or have found series resistance, then that is what to address and repair. Worst case is the ground trace, because any impedance there leads to coupling of things that should not be coupled.
 

MrSoftware

Joined Oct 29, 2013
1,676
You first need to figure out why it's shutting off; either (a) the 3.3v is sagging past the brown out point for the processor, or (b) you're getting so much noise through one of the processor pins that it's locking up. If I were you, what I would do first is try to determine if (a) is the case. Attach a volt meter to the output of the 3.3v regulator, or any point on the 3.3v bus (solder some wires on if you need the reach), hold the quad down with something and give it the throttle that normally makes it turn off. Did the voltage sag below 3.3v? If it happens very quickly then set your volt meter to max/min and it will remember the minimum voltage, or if you have an oscilloscope then you can set it to trigger on a falling edge and keep moving the trigger point lower to see how low the voltage gets. Come back with your answer and we can take it from there.

edit--> option (c), you might have a cold solder joint that looses connection when vibrations get high. I would still start with (a) above and move forward from there.
 

Thread Starter

bigjoncoop

Joined Feb 1, 2019
73
You first need to figure out why it's shutting off; either (a) the 3.3v is sagging past the brown out point for the processor, or (b) you're getting so much noise through one of the processor pins that it's locking up. If I were you, what I would do first is try to determine if (a) is the case. Attach a volt meter to the output of the 3.3v regulator, or any point on the 3.3v bus (solder some wires on if you need the reach), hold the quad down with something and give it the throttle that normally makes it turn off. Did the voltage sag below 3.3v? If it happens very quickly then set your volt meter to max/min and it will remember the minimum voltage, or if you have an oscilloscope then you can set it to trigger on a falling edge and keep the trigger point lower to see how low the voltage gets. Come back with your answer and we can take it from there.

edit--> option (c), you might have a cold solder joint that looses connection when vibrations get high. I would still start with (a) above and move forward from there.
thx for the info mrsoftware... i think i will try to reflow the board first and see if that helps.

holding down the quad copter is no really a option (unless its outside and i am very far away!!!) these motors/props spin at 40,000 RPM at the throttle range in need to get to where the problem was occuring!!!:oops: 2400 rpm x battery voltage.. so 2400 x 16v = death / scary ass sh*t
 

MisterBill2

Joined Jan 23, 2018
5,228
thx for the info mrsoftware... i think i will try to reflow the board first and see if that helps.

holding down the quad copter is no really a option (unless its outside and i am very far away!!!) these motors/props spin at 40,000 RPM at the throttle range in need to get to where the problem was occuring!!!:oops: 2400 rpm x battery voltage.. so 2400 x 16v = death / scary ass sh*t
I wondered how those things flew with such small props. Very high RPMs explains it. Is it possible to disconnect the motors and hook up some other load, such as 12 volt light bulbs of some kind? Much safer to work around. And quieter as well.
 

MrSoftware

Joined Oct 29, 2013
1,676
lol, can you pin it to the ground? The quad I have is older and larger but has landing gear, so when I need to give it a brief load test I just push down on top real hard to keep it pinned to the ground.
 

MisterBill2

Joined Jan 23, 2018
5,228
lol, can you pin it to the ground? The quad I have is older and larger but has landing gear, so when I need to give it a brief load test I just push down on top real hard to keep it pinned to the ground.
That is probably not a good choice with the higher powered Quad Copters. That is why I suggested connecting an external load resistor. It is MUCH safer.
 
Top