This article is a report on a Turnigy 2730-1500 brushless DC motor.
Before proceeding, reader's attention is drawn to an article on a similar motor which failed dangerously during testing, see Turnigy 2211-2300 brushless DC motor. The experience highlights the need for an effective high impact face shield when testing these motors.
The motor is used for testing ESCs, and the requirement was a motor that drew nearly 10A on 3S (12.6V).
Specifications from Hobbyking website:
Dimension: 32mm x 27mm, 43mm(with shaft)
Weight: 28g (kv1300) (not including connectors)
Diameter of shaft: 3.1mm
Length of front shaft: 9.6mm
Lamination thickness: .2mm
Magnet type: 45SH
Max current: 3.5~7.5A/20S
For small parklyer and gliders.
The motor on first inspection is very nicely made, and appears stunning value if it works to specification.
The motor was modified as described at Turnigy 2211-2300 brushless DC motor to reduce the risk of separation of the motor from the mount, and the grubscrews tightened with threadlock.
The motor was secured to a piece of timber, and a 8045 propeller cut down and balanced to load the motor to 9A at 12.6V fitted.
Fig 1 shows the mounted motor. Note the rotor magnets are very close to each other. Although a fractional pole motor, it exhibits quite strong cogging.
Fig 2 shows the cut down 8045 propeller which is 5" in diameter and has a pitch of 4.5". The propeller is fitted for clockwise rotation which produces a downwards thrust.
The cut down propeller measures 4.9" diameter and 4.53" pitch. On measurement, it is quite similar to a 5.1x4.5 APEC E propeller.
Fig 3 shows the characteristics of a 5.1x4.5 APEC E propeller which is used as an estimator of the actual propeller load.
Fig 4 shows the characteristics of the drive system using a 2730-1500 and 5.1x4.5 APEC E propeller.
Measured figures are 8.9A current and 13700 RPM with an Afro30 ESC set to 11°+14Ás advance (optimised timing).
Fig 5 shows the characteristics of the drive system using a 2730-1500 and the cut down propeller with an Afro30 ESC set to 15° advance. The throttle curve is quite reasonably linear.
As mentioned, optimum timing is 11°+14Ás advance, it delivers rapid acceleration with no faltering, stable operation at all speeds and acceleration / deceleration without any loss of sync.
Fig 6 shows the combination timing as both degree and microsecond equivalents.
The motor is an interesting one in that it requires quite a lot of advance, at least 10° equivalent to be stable and maintain sync.
Fig 7 shows the effects of loss of sync with 5° advance, current peaks reached 30A, the motor windings were very hot after just a few seconds of this sync recovery attempt.
Measurements were made of the zero crossing timing shift a range of speeds. The characteristic is specific to the device under test.
The ESC motor timing was set to a fixed 50Ás advance so that it would run over the full range of speeds, but without variable commutation advance which might alter the measurements.
Fig 8 shows in yellow, a rotor position reference signal from an IR reflective sensor. The blue channel is one of the motor wires with the motor at 3100rpm. The point of zero crossing is where the 'sense' voltage of the un-driven winding passes through half supply voltage.
Fig 9 shows the motor at 6200rpm.
Fig 10 shows the motor at 13400 rpm.
Fig 11 shows the shift in zero crossing position vs rpm, including a linear predictor of that shift.
Realise that any advance applied within the ESC is additional to this apparent advance in the magnetic neutral plane.
Fig 12 shows the motor at 12800 rpm with 25Ás advance. Expected commutation would be 30°-25Ás=24Ás after ZC but in fact it is 42Ás which suggest path latency of 18Ás.