0
\$\begingroup\$

I have a damaged RC drone/quad. It has an all in one FC/ESC. The "all in one" is brand new.

One of its motors is twitchy/does not spin. That motor is tested to be good. I tested it by swapping a new motor and saw the same result (twitchey ness).

The drone software is able to view the microprocessor that controls the bad motor. It looks fine.

The motor has three wires on it. Each wire is connected to a pad where there are 2 MOSFETs per pad. One will switch the negative, one will switch the positive. There is probably a chip to switch these in sequence and measure back EMF. There appears to be no damage visually. No motor pad is shorted to ground or positive battery terminal.

When the motor is powered, it twitches. Again, it is not a motor problem, it is an ESC problem in the fc/"all in one". I have read elsewhere that if a single wire is bad on the three phases going to the motor, the result will be twitchiness. That is not the case, all the wires have good solder joints.

I also tested that if the motor is disconnected, there is no short between any of the three ESC pads. I tested that none of the three ESC pads has a short to either the positive or negative terminals leading to the battery.

I want to narrow down to which of the three ESC pads is bad in need of further debugging. If it is a bad MOSFET that will not switch on on say pad one, then I can replace it and hope all is well. It is a small "all in one."

I have a hot air station and oscilloscope. I wanted to minimize desoldering any components, especially those feeding ESC pads that are in good shape.

I also have a "smoke stopper". If I power the FC and spin the motor, it will twitch and then the stopper will shut off power (keeping the smoke in.) I do not remember any particular ESCs being warm to the touch after this has happened.

Using my oscilloscope, I am how wondering how I can connect it to the three phases/pads along with the connected motor and see what trace is "not like the others" and then pull MOSFETs and check components along that channel.

I envision that once connected, I can power the ESC, see the motor twitch, and then look at what trace is missing on the scope. I can then check MOSFETs along that channel as well as.

How can I attach my oscilloscope to the twitchy motor to locate what to do next? Where would ground connect? Should I add resistors between the scope probe and motor lead? I do not want to damage the scope with back EMF or high voltages. The drone uses a 12 volt battery.

My hope is to see just enough to narrow to a channel. I suppose if all looks the same then it is the controller along that channel.

\$\endgroup\$
9
  • 2
    \$\begingroup\$ You measure the inductance and resistance of each phase. But you also need to rule out the motor driver. I suppose you could also run the motor as a generator and put the waveforms on a scope. \$\endgroup\$
    – DKNguyen
    Jul 2, 2022 at 14:10
  • \$\begingroup\$ @DKNGuyen The motor is good. I have checked the inductance of the windings as well as spin it with a Dremel and watch it on an oscilloscope. I also tried a different/new motor. The problem lies with the esc. \$\endgroup\$
    – Jeffrey Edward Messikian
    Jul 2, 2022 at 14:37
  • 1
    \$\begingroup\$ Most common fault on ESCs is burned MOSFETs. With the motor disconnected, measure resistance from Drain to Source and Drain to Gate of each FET (can be done in circuit). If any are bad (low resistance) then replace all the FETs in that phase. \$\endgroup\$
    – Bruce Abbott
    Jul 2, 2022 at 15:43
  • 1
    \$\begingroup\$ @JeffreyEdwardMessikian Is that what your question is asking about? Because title and body all talk about the motor. Or was that previous comment meant to be an update? \$\endgroup\$
    – DKNguyen
    Jul 2, 2022 at 16:46
  • \$\begingroup\$ ROFL You say twitchy motor and bad motor in question and problem lies with the esc in the comments. That comment should be placed central to the question and your title changed. You will not damage your scope. We don't mind helping but we really do not like spinning our props. \$\endgroup\$
    – StainlessSteelRat
    Jul 2, 2022 at 16:46

1 Answer 1

Reset to default
3
\$\begingroup\$

To rule out the motor, you measure the inductance and resistance of each phase. I suppose you could also run the motor as a generator and put the waveforms on a scope.

For the motor driver, the most common failure is a blown MOSFET and MOSFETs (all semiconductors really) tend to fail short. So measure the resistance between the source and drain of all one MOSFET in each high-side and low-side of all three phases. If any do, you know at least one of them is blown but you won't know which one since it is in parallel.

So at that point you either:

  • Replace that entire group of parallel MOSFETs with the same part number
  • Remove each MOSFET from the group one by one and test it until you find the bad one then re-install/replace the ones you removed with the same part number
  • If you cannot get the same part number you will need to replace ALL the power MOSFETs, everywhere, that share the part number with the blown MOSFET with another type (i.e. all high-side, all low-side, or possibly all high-side and low-side MOSFETs). You don't want mixing and matching of part numbers which are in parallel or in the same position on other phases.
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.