Shop OBEX P1 Docs P2 Docs Learn Events
Using a MAX232 ic to drive Mosfets — Parallax Forums

Using a MAX232 ic to drive Mosfets

Zap-oZap-o Posts: 452
edited 2011-03-30 15:20 in General Discussion
Has anyone used or considered using a MAX232 to drive 4 mosfets in an Hbridge configuration? let me know what you think and if you have had any luck doing so. Thanks in advance.

Comments

  • Mike GMike G Posts: 2,702
    edited 2011-03-28 08:37
    The MAX232 is a RS 232 driver. Why would you even consider the MAX232 to drive an H-Brdige?
  • Zap-oZap-o Posts: 452
    edited 2011-03-28 08:45
    I consider this because i want to run my H-bridge off a 5Volt 20Amp power supply and it provides up to +/- 15V output (driving mosfets hince the need for a higher voltage) with a 3.3V in (signal from propeller). And its very easy to implement using a few capacitors.

    I also seen this:
    http://www.pololu.com/file/0J59
  • jverivejverive Posts: 4
    edited 2011-03-28 09:01
    It's an interesting idea, but beware that the MAX232 has integral slew rate limiting circuitry. The output impedance is optimized for driving cables with inherently significant capacitance, and in an H-bridge configuration, slew-rate limiting is generally a bad idea; you don't want high-side and low-side drivers on the same side on at the same time. If you do use the MAX232, minimize load capacitance to maximize output slew-rate. You may also want to look at the MAX232A, which has faster slew-rates.
  • Zap-oZap-o Posts: 452
    edited 2011-03-28 09:13
    So the slew rate may be limited but If I am not going to PWM the mosfets this should be okay. Good point about the High side and low side on at the same time.
  • Mike GMike G Posts: 2,702
    edited 2011-03-28 09:25
    I guess I don't understand. I would be looking at a transistor switch, simple and cheap. The Pololu device is an actual serial device. Internally the device is using PWM control with a uController. I'm not seeing the relationship.
  • jverivejverive Posts: 4
    edited 2011-03-28 09:38
    Slew-rate limiting can be a problem if you're trying to switch any of the MOSFETS at the same time or one right after another, since unless you allow sufficient dead-time between switch transitions. If you're turning one device off, slew-rate limiting can mean it's still partially on while another device is turning on. If the two devices are on the same side (i.e. high-side and low-side of one side of the bridge,) you can have a large current spike from the upper power rail to the other. I consulted with a manufacturer who had a spike of over 1000 amps because of a similar problem (only with IGBTs.) It caused huge problems with the control circuitry. Adequate dead-time was the key. Different motors/servos can be more/less tolerant to the effects of dead-time, so some experimenting may be required.
  • Zap-oZap-o Posts: 452
    edited 2011-03-28 09:49
    Mike G wrote: »
    I guess I don't understand. I would be looking at a transistor switch, simple and cheap. The Pololu device is an actual serial device. Internally the device is using PWM control with a uController. I'm not seeing the relationship.

    Perhaps its me that is confused. I was looking at a different way to drive the gate of a mosfet using a low Voltage power supply.

    Thanks for the input people. Looks like Ill need to design a charge pump.
  • FranklinFranklin Posts: 4,747
    edited 2011-03-28 10:41
    Charge pumps are good at what they do but are not designed to provide much current or a continous signal.
  • jverivejverive Posts: 4
    edited 2011-03-28 10:44
    Make sure you account for the charge requirements of switching a MOSFET, as its gate capacitance is not constant (or even linear) throughout the switching interval. Sometimes it's easier to think of MOSFETs as charge-gated devices instead of simple voltage-gated switches.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-28 13:48
    Franklin wrote:
    Charge pumps are good at what they do but are not designed to provide much current or a continous signal.
    With a large-enough output storage cap and a low-enough duty cycle, a charge pump should be able to excite a MOSFET gate efficiently. Moreover, once the gate is charged, a continuous current is not required, except for leakage and any external gate-to-ground resistance that might be present.

    -Phil
  • TubularTubular Posts: 4,726
    edited 2011-03-28 13:55
    I think this is a very interesting idea. Worth a try.

    If you only needed to drive one fet, you could have a single RS232 TX and RX, plus single FET drive, out of the one MAX3232

    There are 8ch RS232 chips in 5/3 configuration, and isolated ones too.
  • Zap-oZap-o Posts: 452
    edited 2011-03-28 19:27
    Okay I want to stop banging my head against the wall and ask...

    ... Is there an easy way to turn on an off a MOSFET? It seems to be much more complex than anything I ever wanted to use in a design yet so simple I must be over thinking this. The gate is difficult to power with the propeller because you need a in-between device. A mosfet driver, but which one? There are so many, why cant they make a mosfet that fully turns on at 3.3Volts hell its 2011?

    So what is your favorite way to turn on a MOSFET using the propeller? I need some insight and not the male enhancement pill please.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-28 20:27
    Zap-o wrote:
    ...why cant they make a mosfet that fully turns on at 3.3Volts...?
    Maybe that should have been your first question. How about an IRF3708? But, of course, that was answered in your previous thread, here:

    -Phil
  • Zap-oZap-o Posts: 452
    edited 2011-03-29 06:39
    Actually Phil this is not true. Logic gates wont fully turn on with 3.3V at least as far as I know. I needed to apply some 12Volts to the gate. So This is why I ask again my friend. This is why I am struggling with the MOSFET. Its not simple IMO.

    I am searching for a device that will turn on a MOSFET(n-chan) and Turn off a MOSFET (p-chan). This is why I figured a device like the MAX232 would be close. Its got -15Volts for the p-chan and +15Volts for the n-chan.
  • User NameUser Name Posts: 1,451
    edited 2011-03-29 07:29
    Zap-o, I've used the charge pump circuitry of a MAX232 many times to provide power for op-amps and MOSFET drivers. But I've never used the RS-232 amplifiers in the chip to drive MOSFETs for the exact reason jverive gives. Slow gate switching leads to higher device dissipation and the possibility of parasitic oscillation during switching. For a home project, there's little harm in giving it a try. Depending on what you're driving with the H-bridge, how fast you're doing it, and how over-specified your MOSFETs are, it might work well for you. In a commercial design it would look pretty dodgy.
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2011-03-29 08:30
    I think that the "AVAILABLE OUTPUT CURRENT vs. DATA RATE" as well as the "OUTPUT VOLTAGE vs. LOAD CURRENT" is what's going to kill the idea. In this case, the data rate would be the rate at which the mosfets are turned on/off.

    (See page #4)
    http://www.datasheetcatalog.org/datasheet/maxim/MAX220-MAX249.pdf
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-29 09:07
    Zap-o,

    Have you even looked at the IRF3708 datasheet? Yes, the drain-source resistance goes lower with a higher gate voltage, but it's still a very low and quite useful 0.011 ohms with 3.3V on the gate. It's even characterized for gate voltages as low a 2.8V.

    -Phil
  • Zap-oZap-o Posts: 452
    edited 2011-03-29 17:14
    Phil,

    Sure I have looked at the data sheet, perhaps I am incorrect in my interpretation. I see on the first page under Absolute maximum ratings:

    Continuous Drain Current, VGS @ 10V

    Does this mean that I need 10Volts to open this mosfet all the way? I want to run up around 20A through my H bridge maybe up to 35A.

    Please correct me if I am wrong (now that I type this I think I will be wrong, otherwise you would not be scalding me). :)

    My next concern is even if that mosfet works around 3.3V. What happens if I need to use other mosfets that require much more gate voltage? Is it as simple as a multi voltage power supply or do I have to examine driver IC's.
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2011-03-29 18:45
    Zap-o,

    "I want to run up around 20A through my H bridge maybe up to 35A" - Can you borrow any parasitic power from the H-Bridge supply? A high-side driver is not that difficult, either with discrete components or an IC. This thread is way more complicated than it really needs to be.

    Let me ask this... what is the supply voltage to your H-Bridge?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-29 19:35
    Zap-o,

    'Sorry if my demeanor seemed scalding (or even scolding). :) I just wanted to make sure you got that a MOSFET doesn't have to be "all the way on" to be useful as a switch. It's all a matter of degree and picking a suitable part and a suitable operating point.

    With the IRF3708 at IDS = 20A and RDS = 0.011 ohms (resulting from VGS = 3.3V), VDS will equal 0.22V and the IRF3708 will dissipate 4.4 watts, requiring a heatsink. Even "all the way open", RDS drops only another 0.003 ohms; so for most practical purposes, it's fully "on" when driven at 3.3V.

    At 35 amps, this device will drop 0.385V and dissipate 13.5 watts. You will have to determine if this is tolerable, as you would for any other MOSFET-and-driver combination.

    -Phil
    ____________

    'Got interrupted during this post by a major hubbub outside my house:
    2 fire engines
    3 police cars
    2 aid cars
    3 "emergency command vehicles"
    4 volunteer fire department personal vehicles

    'Turns out it was for a minor chimney fire at the neighbor's (which was already extinguished when the FD got here). 'Can't imagine what the response would've been for a major conflagration! (And I wonder why my property taxes are so high. Sigh!)
  • Zap-oZap-o Posts: 452
    edited 2011-03-30 06:46
    Beau Schwabe (Parallax)
    - My current project uses a power supply of 5volts 20Amps.

    Phil,
    - What can I say, you are a life saver and thanks. I think the bulb went off in my head and now it needs replacing.

    - Let me ask another question that may seem stupid, but hell here goes.
    - What effect, if there is one, does theVoltage at gate to source have on the Voltage from drain to source. I suspect that it should not matter.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-30 07:53
    The gate-to-source voltage controls the drain-to-source resistance, RDS. Using Ohm's law, you can then compute the drain-to-source voltage for a given operating current:
    VDS = IDS * RDS

    BTW, since MOSFETs are basically resistive, they can be paralleled (unlike bipolar transistors) to provide more current capacity.

    -Phil
  • tonyp12tonyp12 Posts: 1,951
    edited 2011-03-30 08:26
    As you have 5V available, run the 3.3v through a buffer ic like a 74AHC or 74VHC

    Maybe redusing the number of mcu pins from 4 to 2,
    with some cleaver arrangment using this chip that would allow the 3rd state of braking.
    http://www.fairchildsemi.com/ds/74/74VHC139.pdf
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-03-30 08:32
    The buffer idea is interesting. I would use something from the 74AHCT or 74VHCT family, though, to accommodate the 3.3V input level more reliably.

    -Phil
Sign In or Register to comment.