EPEVER MPPT Tracer3210AN Teardown

  Uncategorized
  • Very audible whine at 8355 kHz in standby state. goes away after some minutes of standby. is it just my device?
  • Idle Power Consumption: 400 – 500 mW (varying) at 26.5V
  • very fast start-up: 5seconds until full power
  • Coil , core geometry: T184, material: sendust? winding: 23? turns, 2.5-2.6mm wire diameter
  • 40khz frequency
  • LS gate rise time 445ns, fall time 282ns, 12.8 V
  • backflow switch: HGQ018N03A (similar HYG018N03LS1D) (30V 1.4mOhm)
  • All HS/LS gate drive resistors 10R, no diodes
  • Gate Driver IRS2186 High-side/Low-Side 600V, 4A, 170ns, no dead-time
  • DCDC Switches: STP150N10F7, 100V, 3.6mOhm (typ), Qg=117nC
  • Snubber: 6.8R and 2.4nF (92ohm ESR, 1206)
  • Snubber HS: 2.2nF

TODO

  • thermal image
  • gate drive circuit + wave form
  • SW node wave form
  • components (caps, switches, snubber, MCU, coil)

thermal image at full power 780W (Vin 62V, Vout 26V)
zoomed at the output switches and traces

DCDC Switching

HS Gate Cycle (270ns on
HS Gate On
HS Gate Off
HS Gate Off
Switch Node Wave Form
Switch Node Turn On (30 ns)
LS turns on, HS turns off (500ns dead time)
LS turns off, HS turns on (500ns dead time)
HS turn on and SW node

Control Loop

Disconnect battery under max power (62V input). green is the HS PWM signal.

Tracking

output power at PSU CV 62V to 26V battery. Tracking with 24W power jumps. Appears not very well implemented, need to test a real solar panel

Conversion Efficiency

Conversion efficiency at constant voltages (61Vin, 27Vout) gdoc

Coil Current

Vin=33, Vout=27.5, Iout=28A, Ipp=3.3A, fsw=40kHz
=> L_dc = 34.7uH

Iout = 7A, Ipp=2A (rising over time after 28A- test, temp?)
=> L_dc = 132uH => core Al=250nH (22.5 turns) => similar to MS184125

wire diameter ~2.6mm (AWG#10)

LEAVE A COMMENT