[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

(ET) IGBT and FET and modern controls



GE fans,
    IGBT stands for "Insulated Gate Bipolar Transistor".  FET stands for "Field Effect Transistor".  SCR stands for "Silicon Controlled Rectifier".  And when you see below any mention of a single transistor, understand that properly paralleled transistors actually act as a single large transistor.
    There is a reason that most electric vehicle controls use FETs instead of IGBTs or other non FET silicon devices.  Most silicon transistors have a 0.7 volt drop across each junction of which there are two in a transistor.  That translates to a 1.4 volt drop across your power handling armature controls when IGBTs or standard bipolar transistors are used.  (Yes, I know that some modern silicon devices have somehow sidestepped that 1.4 volt drop across two junctions.  So now you have 0.6 to 1.0 volt drops in some devices, but never 0.)  When you only have 36 volts to play with, you waste (1.4 V / 36 V) x 100% = 3.89% of your energy.  This is independent of and in addition to any resistive losses.  In electric vehicles, that is a significant loss of energy.
    On a 480 V AC drive that actually has a charged capacitor bank of 650+ VAC, (1.4 V / 650 V) x 100 % = 0.22% is the energy loss due to the voltage drop.  Not only is 0.22% a much lower loss, you also have the entire electric grid to draw from instead of a few golf cart batteries.
    On the other hand, when "on" FETs act more like resistors and have no fixed voltage drop.  Voltage drop will follow ohms law, V = I x R.  Make an FET big enough, i. e., low enough in resistance, and the voltage drop across it becomes insignificant.  This is why electric vehicles tend to use FETs despite some difficulties in their application.
    One other method of DC motor speed control, now less often used, was to use a GTO SCR (Gate Turn Off Silicon Controlled Rectifier).  Normal SCRs when turned on stay on until their input voltage drops to 0 volts.  They are still used on DC motor drives powered by AC where the input voltage is a sine wave that constantly crosses 0 volts.  The GTO SCR was developed for electric vehicles, primarily fork trucks and the like, with DC (battery) power sources.  The "GTO" allows them to be turned both on and off, hence Pulse Width Modulated.  The GTO SCR was a compromise in that any SCR has a single junction, so the voltage drop was only 0.7 volts, as opposed to a bipolar transistor with two junctions yielding a 1.4 volt drop.  Losses were lower than in transistors, but not as good as more modern FETs.
    I've considered updating my E15, but presently my "backward" relay controls work.  However, my C185 Wheelhorse and ER8-36 are good candidates.
    Jeff Dailey correctly states that you need two controls when powering the shunt wound motor driven tractors, the E15 and E20s in particular.  (Jeff, I'd like to see your control design, posted if possible.  A friend with an E8 is thinking of designing one, and seeing your design could be a great help.)  What the commercial SEP-EX (for separately excited armature and field) do is make the armature high current "control" single transistor control that does not reverse.  The low current field has a full bridge reversible control that need four transistors.
    I was given a Curtis "SEP-EX" control with a damaged output section intended for an E Z GO golf cart.  (One of their models uses a shunt wound motor.)  It has many paralleled FETs for the motor armature and four bridged FETs for the field.  If it is not total toast we'll replace and upgrade the armature FETs.  At first glance the unit is not rugged enough for tractor duty, but FETs are easy to replace and upsize.  This is likely a spring time project.
    A permag motor needs either a reversing contactor and a single transistor control or a four transistor bridged control, a much nicer but much more costly arrangement.
    One weakness of many solid state control designs is that they don't have regeneration.  In our tractors regeneration is only needed if you use motor braking going down hills, which, of I course, I do.  The relay controls, at least in the E15s and E20s when the motors were at the speeds requiring full armature voltage, did give regenerative braking.
    And speaking of Technical Service and Bill Gunn, Bill used to offer a solid state upgrade that replaced the armature resistors with a GTO SCR (Gate Turn Off Silicon Controlled Rectifier).  I don't know if the GE circuit boards went away or were used in conjunction with the Technical Service circuitry.  If Bill still offers these, it may be a quick way to get a pre engineered drive upgrade.
    Well, that's enough for now.  Happy reading.
 
Steve Naugler