Shop OBEX P1 Docs P2 Docs Learn Events
Running a L293D Vss voltage at 13V? — Parallax Forums

Running a L293D Vss voltage at 13V?

BitsBits Posts: 414
edited 2012-06-23 08:15 in General Discussion
The data sheet clearly states that Vss can be above 5V but I remain not convinced. Has anyone ran Vss the same as Vs, or in my case I want to run the IC at 13V for both Vss and Vs since I have limited board space and putting in a regulator for just one IC is not elegant.

Comments

  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2012-06-20 13:08
    It may also be worth comparing it to the TI version of that chip the SN754410

    http://www.ti.com/lit/ds/symlink/sn754410.pdf

    This datasheet also suggests that the logic voltage can be as high as the supply for the outputs although the recommended setting is the 5V range. I've never tried running the logic side higher than 5V but if you do I wonder if that will affect what ends up being a valid high signal for the chip. Is it still a 0-5V input or will that scale according to what you're using as the logic supply for the chip? It would be a worth doing some tests to find out.

    I am using the TI754410 chip with both the logic and output at 5V which works ok. It has a separate 5V 3A supply for a large motor. These chips were one of the few H-bridges where I could run the motor down to 5V when the logic was at 5V too. Most of the others expected the motor voltage to be at least a couple volts higher than the logic supply.
  • BitsBits Posts: 414
    edited 2012-06-20 13:14
    Robot,
    This datasheet also suggests that the logic voltage can be as high as the supply for the outputs although the recommended setting is the 5V range. I've never tried running the logic side higher than 5V but if you do I wonder if that will affect what ends up being a valid high signal for the chip. Is it still a 0-5V input or will that scale according to what you're using as the logic supply for the chip? It would be a worth doing some tests to find out.

    That is exactly what I was thinking. I wonder if it has a embedded voltage regulator as the data sheet is not clear on this. I only need a few hundred m-amps for a stepper motor. Oh well I might have to breadboard this first.
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2012-06-20 13:18
    If you find out for sure please let us know!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-06-20 15:26
    I believe the datasheet. But I would definitely not connect Vs and Vss (what an unfortunate designation) together. It would be better to isolate Vss from Vs with a diode and cap, maybe even a choke, to prevent glitches on Vs from causing logic hiccups.

    BTW, there's nothing at all inelegant about providing a regulator for one chip, especially if it prevents problems by providing a little extra isolation from noise on the primary supply.

    -Phil
  • kwinnkwinn Posts: 8,697
    edited 2012-06-21 06:53
    I don't think running both at 13V would work. If you look at the equivalent input schematic (TI version) it consists of a constant current source between Vcc1 and the emitter of a pnp transistor. A 5V signal would not be able to turn that transistor off if Vcc1 was more than a little bit higher than 5V.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-06-21 07:46
    kwinn,

    It seems to me that the constant current source on the emitter would make the switching insensitive to the supply voltage, as opposed to just a pull-up resistor. IOW, the current stays the same regardless of Vcc1. OTOH, you will notice on the TI datasheet that the recommended operating voltage for Vcc1 is between 4.5V and 5.5V, even though AbsMax goes up to 36V. The ST datasheet makes no such recommendation.

    -Phil
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-06-21 08:30
    Clearly the TI part and the L293D are not interchangeable where Vs is concerned.

    The L293D says Vss can be from 4.5V to 36V. Just make sure Vss is not higher than Vs.

    The logic Hi and Low levels, on the L293D. are independent of the voltage applied to Vs.

    I would connect Vss and Vs together. This chip is designed to be used this way and still have normal TTL input logic levels,
    The Prop will drive this properly no mater what the Vss voltage is.

    Duane J
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-06-21 09:02
    I would connect Vs and Vss together. This chip is designed to be used this way and still have normal TTL input logic levels,
    I think that would be asking for trouble, due to glitches on Vs resulting from load switching. The two need some kind of isolation from each other, so that such glitches do not affect the chip's logic circuitry.

    -Phil
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-06-21 09:38
    Hi Phil;

    The L293D was designed for the applications Bits is doing.
    I've used it myself and seen no problems.

    There are no flip flops or other logic except for the level translators.
    Even if there were some transient conditions they seem to cause no problems.

    Duane J
  • kwinnkwinn Posts: 8,697
    edited 2012-06-21 13:06
    kwinn,

    It seems to me that the constant current source on the emitter would make the switching insensitive to the supply voltage, as opposed to just a pull-up resistor. IOW, the current stays the same regardless of Vcc1. OTOH, you will notice on the TI datasheet that the recommended operating voltage for Vcc1 is between 4.5V and 5.5V, even though AbsMax goes up to 36V. The ST datasheet makes no such recommendation.

    -Phil


    I did say "I think" since it was based on the input schematic. No opinion on the ST part since there is no input schematic on the data sheet so I will take Duane's word for that.

    For the TI part I will stick to my original opinion. For it to oerate with a 5V Vcc1 the constant current source has to operate with less than 5V across it. This implies that there would be enough voltage on the emitter of the pnp (13V - 5V leaves 8V) that it could conduct with 5V on the base.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-06-21 13:32
    kwinn,

    No, you're thinking in terms of a common-emitter circuit. But the PNP is connected as an emitter follower. So the voltage on the emitter -- regardless of the supply voltage -- will always equal the voltage on the base, plus 0.7V. It doesn't matter whether the transistor turns off or not. The only thing that's important is that the transistor's output ranges from (VIL + 0.7V) to (VIH + 0.7V <# VCC1), regardless of the value of VCC1. Downstream, this range is effectively thresholded to perform the logic switching.

    -Phil
  • kwinnkwinn Posts: 8,697
    edited 2012-06-21 20:14
    You'e right, I was thinking of it as a common emitter with the output signal on the collector. The transistor can never be turned off with a 13V Vcc1 but as you say the emitter will be at Vin + 0.7 so that does not matter.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-06-23 08:15
    There are no flip flops or other logic except for the level translators.
    Even if there were some transient conditions they seem to cause no problems.
    Even without flip-flops, you don't think a condition could be set up where the part oscillates due to power supply noise? IOW, switch turns on, load engages and glitches power rail, causing switch to turn off, power rail comes back, switch turns on again, etc.

    -Phil
Sign In or Register to comment.