High Amps V Reg.

Discussion in 'General Electronics Chat' started by mwalden824, Jan 14, 2014.

  1. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    Hello. I have been working on a UAV quadrotor design for about a little over a year now. Picking it up, then taking breaks. It was my senior design project and unfortunately I picked an extremely hard project to do and didn't finish because I wanted to design absolutely everything from as scratch as possible. I am almost done with the electronic speed controllers for the brushless DC motors finally, but right now I am working on the power supply for the main board on the quadrotor that will supply the 16 volts needed for each of the 4 motors. The motors can handle 30 amps max according to the conflicting data from the RC hobby website I got them from (http://www.hobbyking.com/hobbyking/store/__25555__AX_4008Q_620KV_Brushless_Quadcopter_Motor.html).

    I am going to go by the lower values to be safe. But since I have four of these I would need a voltage regulator that could handle 120 amps at 16 volts.

    I have been researching this and most or pretty much all voltage regulator let you use external pass transistors to increase the current flow. I have a few questions I hope someone could guide me in the right direction.

    Can you pretty much use any regulator and just select a very high current external pass transistor to do the job? with a large heat-sink on the transistor or transistors and regulator of course.

    If not, where do I go about researching a design like this? I found a website where a guy used an LM723 regulator with MJE3055 external pass transistors to achieve like a 105 A supply. And this would probably work as I may not need the whole 120 amps of current. The motors are a little over 500 watts each and the quad only weighs about 4 or 5 pounds.

    Can a SMPS be designed for this current, or would I have to resort to linear and which would be more expensive or complex?

    I took a class in Power electronics but we just went over theory and didn't get much into practical designing. I still have the book that we only got about a third of the way through. I will be looking through it but I thought I could get some advice from you guys.

    And yes I have a battery that can source 200 amps constant. Not for very long of course but it can handle the regulator.

    Thanks for your time and any advice or guidance you may provide,
    Michael Walden
     
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    Try using lipo's

    Wait...200Amp battery. ? Can the UAV handle the weight ?
     
    mwalden824 likes this.
  3. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,801
    1,105
    Might be easier to make four 30A regulators. Do the motors even need regulators if you're using PWM?
    500W x 4 sounds a lot for a 5 lb quad (but I'm no expert).
     
    mwalden824 likes this.
  4. #12

    Expert

    Nov 30, 2010
    16,321
    6,818
    Why not use batteries that add up to less than 18 volts? It would save massive amounts of transistors and heat sinks if you did that.
     
    mwalden824 likes this.
  5. MrSoftware

    Active Member

    Oct 29, 2013
    504
    124
    Some of the beefy LiPo's can put out a significant amount of current. For example, a 5000mAh LiPo rated at 40c can deliver 200A. Google for 5000mAh 40c and you'll get a pile of results.
     
    mwalden824 likes this.
  6. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    How wonderful this project, but I only saw the videos and no chance to build one.

    The AX-4008Q-620KV Brushless Quadcopter Motor.
    The Rated Watts : 1000W @ 18.5V, 800W @ 14.8V

    So if you using 16V for the power is OK, but if you want to be lower the values and using the uA723 regulator and 2N3055, maybe you can using the MOSFET to replace the 2N3055, using the MOSFET will reducing the numbers of parts and choosing the MOSFET which has the low Rds<10Ω and enough Voltages and current as :

    1. STH260N6F6-2,Nch,60V/180A/300W,2.4mΩ.
    2. FDP8440,Nch,40V/80A/306W,2.4mΩ.
    3. FDP5800,Nch,80V/60A/242W,6mΩ.
    4. IRF3205s,Nch,55V/110A/220W,8mΩ.
    5. IRFPS3810,Nch,100V/170A/580W,9mΩ.

    The more Rds will producing more heat, so choosing the parts you can buy to suit your need, and using the current less than 1/3 as shows above.
     
    mwalden824 likes this.
  7. tcmtech

    Well-Known Member

    Nov 4, 2013
    2,039
    1,667
    I don't see any point in adding an external regulator systems when your speed controller and drive, if designed right, should easily be able to limit the effective peak voltage the motor is getting.

    I don't see any logical reason for regulating the voltage anyway. The motor itself is rated for 1000 watts @18.5 volts and 800 watts @ 14.5 volts which to me means that if you can get your battery pack voltage anywhere close to that its going to work.

    To me this all sounds like too much electronics theory and not enough education in practical hands on application. :(
     
    mwalden824 likes this.
  8. MrSoftware

    Active Member

    Oct 29, 2013
    504
    124
    Having been involved with flying toys for a while, I can tell you 2 things. First the amount of time that you're going to spend at full throttle is very limited. During normal flight, unless you're doing stunts or something, all motors are not under full power much if at all. If your weight load is imbalanced then one or two motors may run closer to their peak, but you should make efforts to keep your weight as balanced as possible so that each motor has enough reserve power to maintain stability in the wind and when maneuvering. Also your speed controllers are going to be the final word in how much power goes to each motor. Just make sure your speed controllers are capped at how much power they can put into your motors and then the motors will be fine. If the speed controller can output 100A and the motor is only rated for 50A, then just limit how much throttle that speed controller will respond to so that it never provides more than 50% power. The brushless motors are 3-phase AC motors.

    I'm not sure how you're designing yours, but mine is a MikroKopter and each of their speed controllers has a processor running firmware. The speed controllers communicate to the main flight controller via an I2C network. If you do something similar then you can have your speed controllers behave any way that you want. Maybe this link will give you some educational reading, and good luck!

    http://www.mikrokopter.de/ucwiki/en/BrushlessCtrl
     
    mwalden824 likes this.
  9. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    Yeah I am using a lipo. It's has 6 lipo cells in series which amount to about 25.2 V fully charged. The battery pack is 5000mAh which can be discharged at a rate of 40 C safely. So that's 5*40=200 amps constant current. I am sure it will drop as the battery discharges. Here is the battery I have:
    http://www.hobbyking.com/hobbyking/store/__10308__Turnigy_5000mAh_6S_40C_Lipo_Pack.html

    Yeah I built the frame out of carbon fiber and the battery weighs around 2 pounds or 830 g I think. The whole setup is really sturdy.

    It's amazing these little motors that can fit in the palm of my hand can handle this much current

    I bought high power motors and batteries because I plan to add a lot of weight with extra sensors, boards, gps, kinect, video camera's and transmitters or I could use kinects camera. So eventually it will weigh a lot more.

    You are right it would be easier to make 4 30 amp regulators and I could even put them on the Electronic Speed Controller boards. Only thing I am worried about is that obviously they will be slightly different in their tested specs because of the tolerances of the components and maybe other factors, but I maybe could account for that in the firmware on the ESCs. So yeah I may do that. Or find a way to make 4 separate regulators but somehow design the control loop to take feedback from all four keeping them running rather consistantly with each other.

    I am not sure exactly what the quad weighs right now, I was just estimating, it may weigh a little more, but like I said earlier I need the extra power for future addons that will add lots of weight.

    The ESCs do use PWM to control the average current flowing through the phases of the motors. But my design for the ESCs have 3 output wires, one for each phase A, B, and C, and two power wires that should come from either the battery itself if I decide to put the power regulators on the boards separately, or the big 150 amp regulator. There will be one more input wire that will control the speed of the motor. It will be an analog voltage 0-5 V where 5 V will be its maximum RPM and 0 off. This is accomplished from the main controller board MCU to the MCU's on the ESCs that use the PWM to drive the motor. So the analog 0-5 V input signal to the ESCs is converted to the duty cycle of the PWM that controls the 3 half bridges that control the current flow through the motor.

    A good article on how BLDC motors work and tips on how to design ESCs is:
    http://ww1.microchip.com/downloads/en/AppNotes/00857B.pdf
    **PDF Warning**
    It's an excellent read if you're are interested in BLDC motor control.


    What do you mean exactly? Like use a smaller voltage battery pack and design a step up converter? Or just choose a batter pack that closely matches the voltage needed for the motors and discharge C rating?

    I just chose a bigger battery for 2 reasons. One, the battery will make the quad last longer with the extra energy, and two, I plan to add more stuff to the quad that will make it heavier so I will eventually need the power.

    I am going to be running a lot of tests on different configurations. But I was wondering are high currents, like this, capable with switch mode power supplies but just more expensive to build than the linear versions? I know they are more efficient but I would love to rather have a SMPS than a linear one for that reason.

    Thanks for all your help and advice,
    Michael Walden
     
  10. tcmtech

    Well-Known Member

    Nov 4, 2013
    2,039
    1,667
    5000 mah x 25.2 volts = 126 watt hours of energy.

    126 Wh / 16.8 volts = 7500 mah. Same battery power but at a lower voltage and no energy being lost to regulator voltage drop or added machine weight.

    Didn't they cover Volts x Amps = Watts and its numerous variants in any of your classes? :rolleyes:
     
    mwalden824 likes this.
  11. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    Well I said my power control class only covered theory. I am trying to learn more about if but apparently you didn't read my post carefully. But thanks for the advice anyway.

    I will test different things out, including just a battery straight to the ESCs without any regulation, and maybe that's all I need. I hope so. But it seems to me as the battery discharges the voltage rail to the half bridges will lower and the current to the motors will decrease also. But I guess I could account for this in the ESC's MCU firmware by measuring the voltage rail and increase the duty cycle according. This is if I choose a battery closer to the actual voltage needed by the motors.

    I chose battery with more lipo cells, which means more voltage, because I eventually need the extra power. I just need a step down converter really. I guess regulation isn't really required. Look I said previously I don't know too much about power electronics and I am still learning, and all I am asking for is help. If you say it is too much theory and not enough experience, then what is you advice exactly on what I should do? Design a Buck converter that can handle the high current? Because I need the 6 cell LiPo battery, which means approximately 25 V fully charged.


    Hey thanks for the references to those transistors. They look like what I would need.

    Yeah, I have an MCU on the main board that handles communication through XBees, takes the incoming signal and converts it to an analog voltage using the DAC that's fed to each ESC and the MCU on that board converts the 0-5V analog signal to the duty cycle needed to control the half bridges and motors. I'm using a Playstation 2 controller to control it, with a simple LDO regulator and batteries and the other XBee. I plan to design an IMU. I have a 4 by 4 mm chip with 3 gyros, 3 accelerometers, 3 magnetometers and a barometer. I will use a Kalman filter and control loop to stabilize the quad.

    Thanks for everyone's advice,
    Michael Walden
     
  12. MrSoftware

    Active Member

    Oct 29, 2013
    504
    124
    That's a seriously big battery in terms of output power. On paper that battery will give you a peak discharge rate of 50C (for 10 seconds). That's 6300 watts, the equivalent output of almost 8.5HP (not accounting for losses). Are you going to put a lawn chair on that thing and fly yourself around?

    If you're putting a lot of money into this (and for safety since it's going to be large), you might want to consider swapping your single big battery for a pair of smaller batteries in parallel. It is not unheard of for a LiPo to fail in flight, and having 2 in parallel might allow you enough time to land safely. Actually that's quite a large battery, you might have overshot the mark on that one by a tad. ;)

    I'm tending to agree with tcmtech. A voltage regulator is going to add complexity, weight and it's going to lose a lot of power to heat. I think you would be better to get a battery with fewer and larger cells, and just limit your peak output power in your speed controller firmware.
     
    mwalden824 likes this.
  13. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    Why do you have to be a smart ass. I am just asking for help, no need to be rude. I have done said I am an amateur in Power electronics because all we covered was theory. Didn't they teach you to read in what, 1st grade?

    All right, all joking aside, I do appreciate your help. But I don't understand what you are saying here. You aren't being very clear. How exactly are you telling me to lower the voltage? (This was pretty much my original question by the way, how to get a regulator, step down converter, or whatever to get the voltage down to the 16 V)

    Are you saying get a smaller battery? or are you saying just use a step down converter to drop the voltage down to the 16 volts?

    I am confused at what you are trying to get me to do. Because I have been saying this whole time that is what I wanted to do, was step down the voltage to 16 volts.

    Thanks,
    Michael Walden
     
  14. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    Okay I get it now. You explained this much better. I guess I did over calculate. I made the mistake of thinking more lipo cell battery pack with higher ratings would equal more energy so longer lasting flight times.

    I just need a new battery I guess, Well that simplifies things a lot.

    Thanks for being more understandable. Some people make you feel stupid just because you forgot something or didn't realize something.

    I am not a power guy at all. I went the digital electronics, embedded systems route. I used to hate power stuff, but I am learning to like it as I learn more. I did make a fundamental mistake as tcmtech pointed out.

    But again, thanks everyone,
    Michael Walden
     
  15. MrSoftware

    Active Member

    Oct 29, 2013
    504
    124
    He's saying you can get the same total power capacity using a battery with lower voltage but larger cells. This way you have the same total power capacity, but without the need for a voltage regulator.

    Consider the units for total power capacity are Watt Hours. You can get 100 watt hours either by supplying 100A at 1V for 1H, or you can supply 1A at 100V for 1 Hour. Both give you the same total power output.

    So you can trade your 25.2V 5000mAh battery (25.2V * 5Ah = 125WH) for a 16.8V 7500mAh battery (16.8V * 7.5Ah = 125Wh) and still have the same total power capacity, but at a lower voltage so you no longer have the need for a voltage regulator.
     
    mwalden824 likes this.
  16. MrSoftware

    Active Member

    Oct 29, 2013
    504
    124
    I don't think anyone was trying to make you feel stupid, it's difficult in black and white text to read emotion. I think he was genuinely trying to help.

    Sometimes the best way to learn is to build it, test it, figure out what went wrong and try it again! :D
     
    mwalden824 likes this.
  17. mwalden824

    Thread Starter Member

    Mar 6, 2011
    51
    2
    That's very true. My professor used to tell us that he hopes we screw up. Which sounds kind of awful, but he said you learn much more from screwing up than you do from getting it right.

    Yeah, I know he probably wasn't trying to make me feel stupid. Just having a bad day I guess. I apologize tcmtech.

    Well thanks again,
    Michael Walden
     
  18. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    Here are some batteries has 16V and enough current, it may match what you need, but the price is high, if you can afford it then you will no need the regulator, 2N3055 or mosfets and heatsinks.
     
  19. tcmtech

    Well-Known Member

    Nov 4, 2013
    2,039
    1,667
    I read all of your posts and took what info was given then extrapolated from that to what you were trying to do and made my advice based on that.

    The going with four 7500 mah cells VS six 5000 mah cells and reducing voltage was exactly what I was trying to point out.

    Sorry if expecting you to do a bit of filling in the blanks from there was insulting to you.
     
    mwalden824 likes this.
  20. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    1) Yes, a switcher can do this. It is roughly 2 kW. We designed 2kW rack mount switchers of various voltage/current combinations. It is possible. You did not say what internal power source was available?

    2) No, you can't just add a couple of transistors to a linear regulator.

    3) You should contact a power supply house to see if they have a product. Building this yourself from baseline starting point is virtually impossible.
     
    mwalden824 likes this.
Loading...