Shop OBEX P1 Docs P2 Docs Learn Events
H-bridge get's boiling hot! What can i do? — Parallax Forums

H-bridge get's boiling hot! What can i do?

BotdocterBotdocter Posts: 271
edited 2011-01-23 00:23 in Propeller 1
My parallax propeller robot control board has an onboard h-bridge.
But when connected to my 12v motors (2), it get's really hot. I put my finger on there and it whas not enjoyable....
I'm afraid that i will burn it. Is there a way to restrict the outgoing power?

What can i do to cool it? A fan maybe? I thought about a aluminium cooling plate, but i can't mount it on the chip.
«1

Comments

  • LeonLeon Posts: 7,620
    edited 2011-01-07 07:39
    Are you using suitable motors?
  • BotdocterBotdocter Posts: 271
    edited 2011-01-07 07:49
    They're the motors of an rc car. Just like a powerdrillmotor.
    I believe the h-bridge can handle 2,5 amps per channel. I only tried slow speeds. Can that be the problem?
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2011-01-07 07:51
    You can find the datasheet for the H-Bridge driver chip here:

    http://www.st.com/stonline/books/pdf/docs/7616.pdf

    The motors shouldn't be drawing more that 2.8A each or you're asking for trouble. Can you connect one of the motors directly to the battery and measure the current? See how it acts under no load and also if there is a load on it. Worst cases will be when the motor starts and under a stall condition (no movement). It sounds like you don't have things sized quite right and may be using motors larger than are recommended for that H-bridge.

    Robert
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-07 11:43
    Hi Botdocter,

    RC-car motors are available in a wide range of amperes. Even up to 20Amperes but still small in size!
    If you haven't bought a Multimeter with a 10A range then it is the right time to do so now. And then simply measuring the amperes.

    In my Opinion a Multimeter that can measure AC and DC millivolts to 400 Volt and AC/DC microampere up to 10A is a must if
    you work with electronics


    I'm using this one
    http://www.dealexcel.com/ut61b-digital-multimeters-acdc-r-f-c-t-duty-cycle-usb_p367.html

    beside two other multimeters.
    Lot's of measuring functions,
    Volts, AMperes
    Ohm-Meter
    Capacity-Meter
    Even Dutycycle
    backlight, and even a USB-connection to your PC to log measurings

    If you can afford $54 I would recommend this one

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-07 14:39
    I have a multimeter wich can do the trick. It's just gonna be hard to put it in between because of the soldered connectors.

    I'm thinking about buying external motor controllers. Wich one should i use? I think 10 amps should be enough right?
  • LeonLeon Posts: 7,620
    edited 2011-01-07 15:45
    You need to find how much current the motors take.
  • bsnutbsnut Posts: 521
    edited 2011-01-07 16:37
    Leon wrote: »
    You need to find how much current the motors take.

    Hi Botdocter,

    I agree with Leon on this. You need to look on the side of the the motors you are using, to see what the nameplate rating is and then markup that number up 150%. By doing this, then you can size what motor controllers are needed for what you need to do. This prevents any overcurrent problems and I do this as an electrician with all types of motors.
  • BotdocterBotdocter Posts: 271
    edited 2011-01-07 16:51
    I know, and have been trying to reach the motors but they're so deep build in that i have to take apart the entire thing. So that's gonna take a while i think.

    I found these brushed ESC's. Apart from the current, Are they usable with the propeller? And will they turn the motor both ways? Or just forward and no back?

    http://hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=9090
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-07 17:03
    I took a look at the propeller Motor control board
    http://www.parallax.com/StoreSearchResults/tabid/768/List/0/SortField/4/ProductID/584/Default.aspx?txtSearch=robot+control+board

    You got a pluggable connector for Vin, for Motor A and Motor B

    At all three connectors you can insert probing cables to measure the current.
    It doesn't even matter which probe is connected to what as long as your meter is switched to measuring amperes and you
    connect the amperemeter as "INLINE"

    Examples
    Vin-gnd----black probe of meter ---meter itself
    red probe of meter ---- gnd of Powersupply

    Vin-Plus----black probe of meter ---meter itself
    red probe of meter ---- lus of Powersupply

    the only difference that makes is that the sign of the current flowing is inverted.

    Or you cut or unsolder one of the cables
    from
    - powersupply
    - motor
    and INSERT the multimeter there. This means the whole current have to flow through the multimeter

    If all of this is should be new to you I highly recommend to learn the basic concepts of electricity
    (I mean a bit beyond "electricity makes motors spin and LED light up") especially how to measure voltages and how to measure currents.

    The current consumed by the board itself will be 0,05A or 0,1A. The motor will take 2, 3, or 6 A so the current to the board doesn't matter

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-07 17:16
    Thanks for your reply.

    I will come to it in the next few days.
    I accually gratuated in electronics ; ) Allthough that has been years ago...

    Thanks again!

    PS: I found these brushed ESC's. Apart from the current, Are they usable with the propeller? And will they turn the motor both ways? Or just forward and no back?

    http://hobbyking.com/hobbyking/store...idProduct=9090
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-08 01:24
    Hi BotDoctor,

    the link doesn't work. If it is a electronic speed control for R/C-vehicles it expects a standard servo-signal on the inputside.
    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-08 03:14
    Try this link: http://t.co/ntsba66

    So if the esc receives a normal servo signal, it will work right? It will turn both ways?
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-08 07:43
    Not sure.

    turning the motor forward baclward depends NOT on th einputsignal but on the inner circuitry of the ESC.
    For Airplanes with E-Motor or helicopters one direction is enough.

    It needs to have a H-bridge to change the voltage's polarity. A Single MOS-FET that does PWM is not enough.

    I would like to comment on your style to build projects
    You bought the controllerboard and connected to possibly much to big motors. That's why the H-bridge gets so hot.
    Now you searched in the internet for a controller that can do a lot of ampere and is still cheap (this turnigy thing)

    Putting together components in electronics is a bit more complicated than plugin in a USB-device to a PC.
    USB has a few standards and EVERY USB-device has to fullfil one or all of these standards (USB 1.0 USB2.0 USB3.0)

    That's why it works plug and play.

    Microelectronics is almost NOT standardised or let's say there are standards but a lot of different ones.
    For example a serial interface can be TTL-level or V24 = +- 12V) RS232, RS422, RS485 variing baudrate, parity bits, stopbits, inverted logic levels, hardware flow-control.

    So you have to check if components that should be interfaced can be configurated to equal parameters and if they work on the same electrical standards.

    In your case you have to measure the current of the motor under maximum load.
    If you know this value you can choose a suitable motorcontrolling H-bridge.

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-08 09:43
    Ok i will have to measure the current. I will...

    Is it possible to boost my h-bridge? With a few mosfets maybe or can i use the existing h-bridge in another way?
  • LeonLeon Posts: 7,620
    edited 2011-01-08 09:47
    Not easily. You will need a different one.
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-08 10:59
    for using the existing H-bridge on the robot control board means to choose a E-Motor
    with a maximum current (under mechanical load of the motor) of 2A. Depending on the weight
    of your vehicle this might be suitable or too small.

    There might be still a chance if your actual motors are really oversized and you do NOT
    need a high rpm on the wheels of your robot. Is it a robot? How big? weight? How fast should it move?

    If you can tell these details it might be possible to use a smaller motor with a gearbox to
    translate the high rpm of the motor to something usable for wheels.

    If you take a look into RC-cars - even 50 mph fast R/C-racers have two tooth-wheels to translate
    the rpm DOWN (not up)

    What I can see from the picture of the board in the manual is that the propeller-chip
    is connected directly with the H-bridge.

    boosting an electronic device itself is almost impossible. Except maybe installing a liquid cooling-system
    which is much more effort than using a bigger H-bridge. Adding big MOS-Fets would mean to add
    extra circuitry to get back from a high voltage PWM signal with changing polarity to a signal
    motor enable and a signal direction, to drive the input of a bigger H-bridge.

    by the way
    It is always good to give an overview about:

    - Your project: what do you want to do IN THE END

    - your programming skills

    - your knowledge about electronics.(maybee by describing your most comlicated circuit that you have finished successfully)

    Doesn't have to be a 10 page report but 2-5 sentences per item
    with a rough classification like

    expert,
    advanced hobbyist,
    moderate experienced hobbyist,
    newbee

    This will help the forum-members to give you understandable advice.

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-08 14:05
    Thank you stefan.

    I tried to measure the current taken by the motors but somehow when i put a ampere meter in between the motors won't turn. @?!??
    The board is powered and does what it has to but somehow when the motors start turning the program crashes and the propeller is non-responsive.

    When i apply the power without the meter in between everything works fine!
  • LeonLeon Posts: 7,620
    edited 2011-01-08 14:28
    Check the fuse in the DVM.
  • BotdocterBotdocter Posts: 271
    edited 2011-01-08 14:45
    There is no fuse. There is only a fuse for low currents.
  • BotdocterBotdocter Posts: 271
    edited 2011-01-08 18:42
    Ok i got it to work. Ive been able to measure the drawn current.
    At it's max. It uses 2.4A. When starting it draws less. It doesn't come over the 2.4 ampere ever.
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-09 01:13
    Hi BotDoctor,

    congrats ! Well done that you measured the maximum current.
    that is the most important and guiding to the solution value we need.

    To check the maximum current is so common for this kind of application that
    nobody even mentioned it at the beginning of this thread.

    I want to introduce you to a semi-professional way of working on your project.

    Therefore I have some guiding questions that you should answer:

    Where can the usermanual to for the Propeller Robot Control Board (PRCB) be found?
    if found - download it

    Which is the chip on the propeller controller board that controls the current of the motors?

    Do a search on the internet for the datasheet of this chip.
    read the feature section of the datasheet.

    Then give an answer to the question: "Can these 2.4 Ampere Motors be driven with the propeller controller board?"
    What happens?

    I have done exact all these steps. I could tell it to you directly. But I don't want you to beg me for my fish. I want you to learn fishing yourself.

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-09 06:04
    So i read the h-bridge datasheet (posted before)
    And it should be able to take 2.8A continuous. And 7A peak.
    So why is it getting so hot?
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-09 09:27
    There is still some resistance in the semiconductable material (R-on-resistance)

    Guiding questions: How is electric power calculated? Which values do you need therefore?

    What says the datasheet about these values?

    What says the datasheet about the thermal behaviour of the chip?

    best regards

    Stefan
  • BotdocterBotdocter Posts: 271
    edited 2011-01-09 09:43
    In case no-one noticed, i gratuated in electronics. Allthough it has been some years.

    I=U:R

    For the answers in you post: we both know wat the datasheet says.
    Do i reallly have to type it out? Oh wait... I did allready in my former post.

    It does't answer my question though!

    I want yo thank you for your help but is it nescecairy to treat me as a kid?
    I know my way around electronics. If there is something i don't know i will ask.
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-09 10:30
    right formula in this case.
    Power: P = R * I²

    Ron = 0.3ohm

    P = 0.3Ohm * (2.4A)² = 1.73 W

    Datasheet Thermal Data

    PowerSOI20 Package:
    62°C/W Maximum Thermal Resistance Junction-Ambient Mounted on a multi-layer FR4 PCB without any heat sinking surface on the board.

    62°C/W * 1.73W = 107.3°C

    best regards

    Stefan
  • LeonLeon Posts: 7,620
    edited 2011-01-09 11:01
    If the motor current was being measured with the H-bridge in circuit, it needs to be measured when powered directly.
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2011-01-09 11:46
    If you were using the wheelkit from the Stringray this wouldn't be an issue since those have already been tested with that control board and works within the limits of the board. In cases like this you can probably get away without testing everything first since it has been done for you by Parallax. Otherwise when you are doing something new it falls on you to do the testing. If that part is skipped you can get lucky. If not, you'll run into issues like the one you are seeing here. Although testing can add some additional work it can often save you a lot more time in the long run.

    Leon is right on about measuring the current of the motor 'itself' when connected to a power source. The power source should be just like the one you want to use, probably a fully charged battery pack. With that you need to test it several different ways: No load, during startup, under load, startup under load, etc. to find out exactly what it does draw.

    If you are making these measurements while connected to the H-bridge without do the basic motor tests first then you're pretty much just wasting time and effort.

    I posted a link to the datasheet for the H-Bridge chip on the board. It is supposed to have thermal limiting built-in (although I prefer not to push it) and what may be happening is that you are exceeding the capacity of the chip which is causing it to heat up too much and start to shut down. This may limit what you are pulling through it so you might only be seeing what it is letting through instead of what it actually wants to use....

    When you asked about beefing up the H-bridge it is easier to bypass what is on the robot control board. At that point you are better off just using a USB Propeller Proto board and connecting that to a different H-bridge driver. You can either use a standard H-Bridge where you can control the direction and provide the PWN for the speed in each direction. Or, you can get one with some smarts on board that has an input the same as a continuous rotation server where pulses of about 1ms are full speed one way, 1.5ms are off, and 2ms are full speed the other direction. Values in between make the motor go in the appropriate direction at a slower speed. Those pulses should be repeated approximately every 20ms. You're still controlling an H-Bridge but instead of driving it directly you use use one pin.

    If you really want some good information on testing motors for use with H-bridge drivers and also the R/C controllers start looking up information on Battlebots. For that we did push things to the limits (and well beyond) but that was only meant for 3 minute matches. With so many motor options out there everyone had to test the motors to see what they would draw and figure out if they would work with each controller. I personally witnessed several cases where that wasn't done and watched motor controllers, wiring, and motors all go up in smoke and flames....

    With a bit more research and testing I'm sure it will all fall into place for you!

    Good luck on your project.

    Robert
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-01-09 11:57
    Hi, Leon,

    this is new to me. And you catched me on not following my own advice.
    Read the datasheet carefully and completely.

    Do you think the resistance of the H-Bridge keeps the current down
    or does the non dissipative protection of the H-bridge limited the current already?

    Do you know if the H-bridge can be driven a long time when ithe overcurrent protection is always active?

    best regards

    Stefan
  • LeonLeon Posts: 7,620
    edited 2011-01-09 12:17
    With the H-bridge in circuit (which is how I think he measured the current), the motor current measurement will be meaningless.
  • zoopydogsitzoopydogsit Posts: 174
    edited 2011-01-09 12:49
    And it should be measured under the expected mechanical load.

    It should have the lowest current free running without a wheel attached. It should draw more with the weight of the robot running on a flat surface. It'll draw even more going up an incline. It will draw even more pushing against a wall. It will draw most with the wheel unable to turn. Also ensure the robot has the weight that is expected for it to run, if this includes batteries, cameras, and other items then either bolt them on or place equivilent expected weight on so you will be measuring usual conditions.

    And as Leon said, motor to battery with the meter in series to provide the least resistive path.

    This next comment may be for Leon or anyone else that has experience. I haven't done this, but I'd assume he could deliver less power, and thus current to the motors if he used PWM to drive the duty cycle to the motors. Sure this would equate to less torque, but should mean less current passing through the H-Bridge. Though the downside is that it may be a bit tougher on the H-Bridge (depending on the implementation), and switching frequency could make it even hotter or cause it to premeturely fail. If it would work then it would be relatively easy to implement in the propeller. Anyone tried this?
Sign In or Register to comment.