Greenworks CSB410

I received an inoperable Greenworks CSB410 chainsaw for evaluation.  Unfortunately, Greenworks does not sell any electrical parts for repair, so you may need to get very creative to fix a dead saw. 

I expected one of the electromechanical parts (chain-brake switch, throttle trigger, or power relay) had failed due to mechanical damage. But they were all functional.  Even the motor and its thermal sensor tested as being good.  That leaves only the controller itself as the cause of the failure.  

Eventually the failure was discovered to be a shorted MOSFET on the low side of the 3-phase motor controller.  This is not something that can be repaired without at least partially de-potting the motor controller.  And even then, it would have to be a kludgy fix.

 Theory

The CSB410 is a brushless-motor chainsaw with an 18" bar.  It is rated 2.0 kW (2.68 horsepower) or 2.5 kW (3.35 hp) depending on….  I don't know what.  If you ask Greenworks, the official answer is that the CSB410 is a family of saws with different models and specs.  A more forthright answer would be:  Specifications subject to change without notice.

Either way, it's a fairly unimportant difference because the rating is peak power.  The system cannot sustain that power continuously.  Both the motor and controller have sensors to measure temperature.  Power will be reduced if either sensor gets too hot.  Two kilowatts with a fresh 80-volt pack implies a battery current of 25 amps before accounting for any losses in the controller itself.  Two point five kilowatts implies a battery current of over 31 amps.  Ultimately, the battery can limit how much power the saw will produce.

The chain brake is a mechanical device that inhibits chain movement as it would on an ICE saw.  Additionally, it provides an electronic safety switch.  The switch itself is a conventional lever-actuated microswitch.  Interestingly, the control interface utilizes both the normally-open (N.O.) and normally-closed (N.C.) contacts.

When the chain brake is in the run position and the throttle switch is operated, a large relay closes and applies motor power to the controller. (The controller also receives “housekeeping” power via other wiring as well).  You can hear the relay engage (there's an audible click) if the motor is not running.  After about a minute the relay disengages if the saw is not running.  

The no-load chain speed is 20 m/s.  More on this later.

Motor Controller

Physically, the controller circuit board is surrounded by a 100 × 70 × 15 aluminum “tray” filled with a hard potting compound.

The controller has several connections to other devices listed below:

• Motor Power: 3-position connector

• Motor NTC Thermistor: 2-position connector

• Chain-Brake SPDT Microswitch: 3-position connector

• Trigger/Throttle: 5-position connector

• Battery Box: 3 individual wires with spade terminals

• Power Relay: 2-position connector, plus 2 heavy wires soldered directly.

No rotor position sensing is used.

Several components are visible outside the potting.

Power MOSFETs are Magnachip MDP14N050 in a TO-220 package.  This part is a Chinese N-channel switch rated at 135 V, 120 A with an on-state resistance of 5.0 mΩ.  Because the controller has 12 MOSFETs, and a 3-phase H-bridge inverter requires 6 switches, we can assume two MOSFETs are used in parallel per switch.

There is an NTC thermistor to sense heatsink temperature for the power MOSFETs. 

DC-link capacitors comprise three 330 uF, 100 volt electrolytics.  It's possible that the high-wattage 22-ohm (red, red, black) resistor is used for capacitor precharge.  See photo below.

Motor

The motor is marked “Globe 6435 T0300371-00 231114-B6”.   It is not part of Allied Motion's Globe Line of BLDC motors.

Physically, it uses an outer rotor that has an OD of 70 mm.  It exhibits 14 detent (cogging) positions.  There's a 74 mm ID shroud that surrounds the motor to keep it cleaner.  I counted 12 slots (winding coils) through the ventilation openings.  This number must be evenly divisible by three for any 3-phase motor.  With 12 slots, the permissible number of poles (magnets) could be 2, 4, 8, 10, 14, 20, or 22.  The number would likely be at the lower end of the range due to space constraints.  I have not disassembled the motor to count them. 

Electrically, the phase-to-phase resistance is 84 milliohms (using a typical multimeter zeroed for test lead resistance will probably read 0.1 ohm).  Inductance when measured at 1 kHz was 83 uH in the quadrature axis (rotor at a detent position) and 60 uH in the direct axis (rotor in-between detent positions).

I tested the motor unloaded using a generic electronic speed control and an optical tachometer to observe the following: 

At 20 VDC, it ran at 2976 rpm and drew 1.4 A (28 watts).

At 10 VDC, it ran at 1480 rpm and drew 0.95 A (9.5 watts).  

From this we can assume it would run at ~11,900 rpm with 80 volts.  This is reasonable given that the speed of a high-performance ICE saw would be about 12,000 RPM.

The motor's temperature sensing thermistor measured 130k ohms at 63 degrees F.   The thermistor has a negative temperature coefficient, so as temperature increases, resistance decreases.

A photo below shows the $14 generic electronic speed control (ESC) I used to test the motor.  It is rated for 1000 watts with a DC supply of 6–24 volts.  You can read more about it on my other trials-related website via this link: https://www.ossa-efi.com/home/1-kick-starting/electric-start-investigation#h.chlzc01j34ps

Power Relay

A SPST power relay is in series with the battery and the motor controller.  The relay is part number HFV7-P / 024HST.  It has a 24-volt coil and the contacts are rated for 24 volts at 40 amps.  The 12-volt coil version of this relay is far more common, but I did find the 24-volt coil version on AliExpress.  This is one of the few electrical parts that could be purchased directly in the aftermarket.  Curiously, it is considered an automotive relay and the contacts are probably not officially rated for 80 VDC operation. 

Battery Box

I was surprised to see three heavy-gauge wires from the battery box to the controller.  Battery positive and negative, of course.  But what does the blue wire do?  My multimeter registered a small pulsating voltage between the blue and black wires.  Turns out this is a serial data communication line and does not need to be a heavy gauge wire.  See the section on the Greenworks 80-volt battery for more information.

Chain-Brake Microswitch

The Chain-Brake microswitch is SPDT with a 3-position connector.  The controller-side connector has male pins.  The switch-side connector has female pins. 

The connector is wired as follows:

• Red: Common (C)

• Black: Normally Open (N.O.)

• Blue: Normally Closed (N.C.)

Both the N.O. and N.C. circuits are used by the controller.  I'm not sure what advantage this offers, but I presume it is a safety feature.  Perhaps the controller can detect a failed chain-brake switch?

There must be continuity between the Red to Black wires for the saw to operate.

Trigger/Throttle

The Trigger/Throttle control utilizes a 5-position connector.  It contains both a SPST on/off switch and a ~34k ohm throttle potentiometer.

Chain Speed

Greenworks specifies the unloaded chain speed at 20 m/s.  Most Americans will find feet per second (or feet per minute) more familiar.  Regardless, I'll do the calculation using metric units.  The chain pitch is specified as 9.5 mm (3/8 inch).

The equation for linear chain speed in m/s is: 

2 × chain pitch (mm) × number of sprocket teeth × motor rpm ÷ 1000 (mm per meter) ÷ 60 (seconds per minute) 

((2 × 9.5 × 6 × 12,000) / 1000) / 60 =  22.8 meters per second

The factory-specified 20 m/s is lower than my calculated value by a factor of 0.88 (88%).  This is probably to account for a battery voltage lower than 80 volts.  Maximum motor RPM is directly proportional to battery voltage.  88% of 80 volts is about 70 volts, which would be equivalent to 10,500 RPM.  

To convert meters per second into feet per second, multiply by 3.28.  

Thus, 20 m/s is about 65 ft/s.

Teardown Video

Below is a 30-minute teardown video of a CSB412 saw I found on YouTube.  Honesty, it's pretty tedious.  I'd advise you to run at 2x speed and skip around. 

Testing N-Channel MOSFETs In-Circuit

A common cause of motor controller failure is a bad MOSFET.  Sometimes the damage is obvious — a MOSFET's package will be cracked, burned or have exploded.

More subtle damage can be detected using the diode test function of a DMM.  The diode test function measures the voltage drop across semiconductor junction.

In addition to the MOSFET's intended action as a transistor switch, all MOSFETs have a parasitic “body diode”.  The voltage drop across the body diode will be on the order of 0.5 volts.  Whereas the voltage drop across the drain-source path of the transistor will be on the order of 2 volts.  This voltage drop may be observed by reversing the positive and negative test leads in the procedure described below.  The exact value of the voltage drop is not critical — a shorted MOSFET will read 0 volts.

The adjacent diagram is a simplified version of the circuitry involved.  There will be other components (such as DC-link capacitors across the battery terminals and resistor voltage dividers to sense back-EMF from the motor) in any real implementation.  Fortunately, the “diode test” is still a reliable indicator despite the presence of these other circuit elements.