How does an alternator charge

Discussion in 'General Electronics Chat' started by Tobias, Apr 19, 2012.

  1. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    I hope this is OK under the automotive guidelines. I am actually using dc brushed motors in an agriculture application.

    The problem I am having is everything works fine on my test bench with a 12v battery and charger. Everything works fine on the tractor with a battery charger supporting the battery. When I start the tractor and rely on the alternator I am having MOSFETS fail. The MOSFETS fail instantly at start up. If the motors get moving the MOSFETS live.

    I have adjusted my code to start the motors gradually. I begin by using a 10 percent duty cycle for a quarter of a second then 40 percent for quarter of a second and then let my speed control routine take over. Even using the mentioned approach, randomly the MOSFETS fail going into a closed state. Physically the MOSFETS look like brand new.

    I inserted a TVS this morning and it didn't seem to help the problem. I clamped the voltage at 28 volts. The MOSFET is rated for 60v/40A. The motor requires 3.3A to operate.

    I am beginning to wonder if there is something else I am missing and not understanding and that is what really happens when the alternator decides to produce power. Any insight is greatly appreciated.
     
  2. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    I don't know about anyone else, but I'm having a little bit of trouble following your post. What do the motors do? Where are the mosfets?
    A schematic of what you have would help.
     
  3. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    Attached is a schematic. The MOSFETS are driving the DC brushed motors. Each control board (the attached schematic) will drive four motors individually. Power and Ground are supplied from the tractor through 6AWG cables directly from the battery through a solenoid that is switched by the key.

    For example, today I am working on a 16 row planter. There are four of the control boards and 16 motors. The tractor alternator is rated for 148A. This thing will randomly pop the MOSFETS right when the motor starts up. There is no consistency to one MOSFET per board failing.

    The drivers on the MOSFETS are operating at 2KHZ.

    So my question is what am I missing? This board works great on a test stand but not on the tractor. The only difference is the alternator. The board, harness and motors are all the same. I am thinking that when the motors power up the alternator creates a voltage spike, however the TVS I tested today really didn't support that theory at all.
     
    Last edited: Apr 19, 2012
  4. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Whatever these motors are doing, are you trying to start them under a load?
    Starting a motor under a full load will draw considerably more current.
    On the drawing you state the 24v is actually 12v battery, are the motors 12 volt or 24 volt?

    edit: Maybe I'm blind , but I don't see any voltage regulating or conditioning, is it there but not on the drawing?
     
    Last edited: Apr 19, 2012
  5. GetDeviceInfo

    Senior Member

    Jun 7, 2009
    1,571
    230
    automotive alternators typically produce a polyphase AC voltage, which is rectified onboard the alternator. Voltage is regulated by controlling current in the field windings. The battery will act as a smoothing filter at lower currents, but you may want to add an LC filter. A bad diode in the alternator could produce problems.
     
  6. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    When you are cranking the tractor motor, the voltage will drop low, down to 7-9 volts on a 12 volt system. When the motor starts the high current drain on the battery is gone, and now the alternator is charging. So you went from a low of7-9 volts to a spike which could be extremely high. Probably higher than the 60v rating of MOSFET
    Have you tried leaving the tractor motor running, then turn on whatever it is.
    Here again voltage regulation and conditioning is important.
    Connect a voltmeter to your battery and have someone crank the motor to verify what I said. You won't be able to see how high the spike is (too fast) but you will see how low it goes.
     
    Last edited: Apr 19, 2012
  7. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    I wasn't clear enough when I said the MOSFET pops when the motor starts up, sorry about that one. I mean the MOSFET pops when the electric motor starts up.

    The motors are under no load at startup. The motors are rated for 12v. The schematic was drawn up when we were going to use a 24v system.

    Today I made some gains. I placed a small battery next to each control box that controls four electric motors. That seemed to help alot. I am just wondering that 6AWG wire is not enough to sustain adequate current when all the electric motors are in the process of starting. Using an online gage calculator is looks like it will handle to load easily, as each motor requires 3amps to operate. But I am sure the current draw to get each one started is atleast twice that number.

    Its been tough to troubleshoot, out of 16 electric motors, the event is random to what row it will happen and also what start up event.

     
  8. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,782
    941
    This might sound like a strange suggestion, but go over all your physical connections in the battery system AND the grounding connections. make sure they are tight and the metal to metal contact is good(remove rust or corrosion). Also examine the crimp where the wire attaches to the connectors. Remove and replace any found to have corrosion present on the crimped copper area. Be thorough and don't miss any of the connections in the battery system.
     
  9. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    This is happening on several farm implements randomly. I have had some pop instantly, others after many stop/start events and even some have run over 200 hours with atleast 200 stop/start cycles.
     
  10. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Let me see if I have this correct. You have #6 wire from the battery(?) of the tractor going to this machine that has a couple (?) of these control boxes.
    You have 4 three amp motors connected to each control box.

    How far is it from box to box, what size wire are you using for these connections? How many of these control boxes are there? Have you metered the power to the motors or control boxes for voltage drops?
     
  11. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    #6 from battery at front of tractor. From the front of tractor to implement is appr 20' then the #6 'T-s' off to the left and right. Left and right of center 45" is a control box on an eight row configuration. There is junction block with two 3/8" studs at each control box. The longest the main power and ground supply cables will travel is 20' to the implement then another 14' to the most outer control box.

    The control box harness gets it power and ground from the junction blocks that connect to the #6. Its about 8" of #14 into the control box for the ground. #14 is used for the output ground going to the electrical motors, I am using N-type MOSFETS. The positive side of the control harness goes directly to the electric motors. From the control box the electric motors are either 60" or 30" of #14 wire.

    Total number of control boxes depends on the amount of rows. From 8 to 32 rows, so 2 to 8 control boxes. For the 32 row, I made a hydraulically driven 140amp alternator.

    The MOSFETS don't discriminate, they will pop on 8 row to 32 row applications.


    I only have about 0.3v drop at the last control box when all motors are running on the 16 row I am working on now.

    I have tried voltage snubbers(TVS), I have tried putting 25Kufd caps next to the control box, I have tried putting a lawn mover size battery at every control box and the same result. Random MOSFET failures. I even found another MOSFET part number and I don't know what it could be. Its always at the start up of the electric motor when the MOSFET fails and I can seem to get them to cycle on/off all day long if I am using a battery charger instead of the alternator.

    The only thing I really changed from last year to this year is the enclosure/header. Last year I used a CINCH and this year a Deutsche. The CINCH was only rated for 3amps, the Deutsche is rated for 7.

    I use a four layer board with a power and ground plane. I think 2oz maybe 3oz. For sure not just 1oz. The pad that the Source lies on is 0.225x.113. I have four vias connecting the pad to the ground plane. Its pretty much the same as last years board.

    I had one guy go six hours until a failure, another guy had three MOSFETS fail the first time he started the electric motors.



     
  12. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,782
    941
    Sounds like its time to try another alternator and either fix it or rule out that as the source of your problems.

    :(
     
  13. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Yeah, it sounds like you have everything under control. The wiring configuration sounds good, this is a tough one...
    Where are you getting the MOSFETS? any chance they're seconds or counterfiet?
    And as Kermit2 says check out the alternator..Do you have access to a o'scope, a shorted diode would be a huge problem.
     
  14. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    I actually rode on the back of the planter today with my digital scope hooked up to a Drain on the MOSFET. The output looked pretty good, when the MOSFET would turn off the most I noticed was about a 0.5 overshoot. So it seems like my kickback diodes are doing their job. The only problem is 'seeing' what happens the first time the MOSFET turns on and being able to study that first wave form. I believe that is when the problem is occuring.

    I lowered the PWM frequency today from 4Khz all the way down to 500hz. That seemed to really help prevent the MOSFETS from failing. It hasn't happened again yet. I think that maybe my start up routine needs some work and then I can increase the frequency. Or maybe I just need to leave it alone because it is working.

    This has been happening on many different makes of tractors supplying the power so I don't think its literally the alternator. One thing I did today on a planter is run double supply power and ground 6awg wire. I just wanted to be way overcenter on being able to supply current to my boards. I will know tomorrow how that one works out.

    The MOSFETS are a Fairchild piece from Mouser. I have used their MOSFETS on this project for a couple years and they have worked great. The part number was obsolete when I built this run of boards. I went through some other outfit ( I can't remember their name) and they provided me with an ON Semi part number that on the spec sheet looked identical. THat was when the problem started to come up. Last week I talked to Fairchild and they recommended another part number. I installed it and after adjusting the frequency to basically the same as their old part number I am getting it to live.



     
  15. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Sounds like the ON semi parts weren't up to the standards you needed.
    Looks like you've fixed it.
     
  16. Tobias

    Thread Starter Active Member

    May 19, 2008
    158
    0
    I think I have found the problem unfortunately. I say unfortunately because I am not to sure how to remedy it easily.

    Somehow I went from using regulated 5v to drive the gate on the MOSFET to battery power. The Max rating is 20V VGS!!!!! I bet I am getting a voltage spike once in a while that is doing this.

    My question is, can I use a lower rated TVS to fix this? Right next to the board I have a two post terminal block (3/8-24 studs) that feed Battery Power and Ground. What I am thinking is installing a 576-P6KE16 TVS in series with a 100ohm resistor between the two posts. THis should in effect protect the gate from more than 20V correct?
     
  17. gerty

    AAC Fanatic!

    Aug 30, 2007
    1,153
    304
    Just guessing here, but wouldn't it be easier to limit voltage to driver (5v) instead of each gate?
    Vehicle power is notorious for spikes and noise, the more protection you have the better off you are.

    edit: I didn't see a part number for MOSFET so I don't know if it's a logic level or not, in other words 5v on gate may not be enough to turn it on.
     
  18. crutschow

    Expert

    Mar 14, 2008
    13,001
    3,229
    Why the 100 ohm resistor? That will inhibit the ability of the TVS to limit the spike voltage. Normally they are just connected directly across the line to be protected.
     
Loading...