Wind turbine project... 3 phase rectifier-bldc motor controller

Discussion in 'The Projects Forum' started by Cbergerud, Dec 23, 2009.

  1. Cbergerud

    Thread Starter New Member

    Dec 23, 2009
    Myself and a few other mech eng students have built a vertical axis wind turbine. To make power we are using a custom built 3-phase 12 pole brushless permanent magnet alternator. This alternator puts out 3-phase AC of varying frequency and voltage, basically a max of 36V and about 50A. We take this AC and rectify it to charge batteries. However its a bit more complicated than that... Our design is a poor/non self-starting design, so we need to run this alternator as a brushless DC motor to get into operating speeds when the control system says so. In addition we have to brake the alternator when the wind is too high or the batteries/charge controller stop drawing power, to prevent overspeeding.

    We have a PIC-based control system that monitors wind speed and the turbine/alternators absolute position and rpm. It will be determining when to start, brake, or otherwise control the turbines output. It can output motor position in any form a controller might want, its just a matter of programming. This is working fine so far.

    Basically the turbine has 3 needed functions for a controller:

    1.) Rectification: During normal operation the 3-phase power from the alternator must be converted to DC efficiently.

    2.) Startup. The alternator is run as a BLDC motor. PWM could be employed to control the rate off acceleration until it hits about 150rpm. Once the desired speed is reached, the startup function shuts off.

    3.) Braking. When the wind gets too high, we need to brake the alternator by shorting its phases together (with high frequency PWM so the power output is not disturbed too much) This would have to be done on the AC side of rectification or we'd be shorting out the battery/charge controller as well as the turbine.

    Heres a few pics of the turbine to encourage responses :)




    OK heres a diagram of what I have in mind right now:

    So for startup mode (determined by PIC 1, gates controlled by PIC 2) The phases are excited in the right order using mosfets or igbts until the alternator reaches about 150rpm or so. PWM might be employed here to control the rate of acceleration. Once 150rpm is reached all the gates turn off.

    The 6 diodes act as a full-wave rectifier. When the output voltage of the alternator is lower than the load(battery) voltage, no load is placed on the alternator. The alternator is tuned so that it reaches 24V at about 175rpm. As the wind catches it from 150rpm and accelerates it towards its 300rpm max speed the voltage differential rises and puts power into the load.

    When the alternator begins to exceed 300rpm PIC 1 signals PIC 2 to brake the alternator turning on the top side and bottom side gates in an alternating fashion, with PWM used to control braking force.

    All of the PIC/logic side is fairly solid, what I need help with is the process of going from on-off signals from the pic to a PWM signal driving the gates of 6 IGBTs or MOSFETS. What kind of gate driver(s) do I need, do I need to isolate the PIC from teh gate drivers or the gate drivers from the IGBTs, and will the power supply to the PICs and gate drivers need some kind of conditioning?

    Ive been scouring tons of PDFs but I havent been able to find quite what I'm looking for.

    How close am I??
  2. SgtWookie


    Jul 17, 2007
    Interesting project.

    You should investigate synchronous rectification.
    In synchronous rectification, power MOSFETs are used as "ideal diodes", eliminating the traditional voltage drop and resulting power loss.

    re: high frequency PWM - keep in mind that charging/discharging MOSFET gates takes some power, and when MOSFETs are in transition, they are dissipating power as heat. You'll likely be better off if you stay under 30kHz with the PWM; lower is better.

    It is simply too late for me to give you more input.
  3. JDT

    Well-Known Member

    Feb 12, 2009
    You are very close.

    As SgtWookie says, synchronous rectification will be "the icing on the cake" and can be done with the basic circuit as drawn if you use MOSFETS. It's a little known fact that with current flowing through the body diode (the MOSFET reversed polarity) the device will still function. So you can reduce the diode losses by switching on the MOSFET while rectifying. However, you will need to monitor the voltage and the current flowing through the devices to know when to switch them on and off.

    As a rule I would completely isolate your PIC controller from the power section of the system. It should have its own power supply. One way you could drive your MOSFETS is to use a gate drive opto-coupler HCNW-3120-000E or similar:-

    There is a link to a datasheet on this page. Use one of these per MOSFET. For local gate drive power I would use a 15V 1W DC-DC converter, one per each device:-

    This will give a totally isolated solution. To monitor the current flowing through each MOSFET I would use Hall-effect devices like this:-

    Not only for controlling the synchronous rectification but also for over-current limiting while running as a motor.

    As always, the clever bit will be the control system!

    Very interesting project, especially the PM alternator. Been wanting to make something like this for some time!

    Let us know how you get on.