Shop OBEX P1 Docs P2 Docs Learn Events
suitable higher current replacement for ULN2803 — Parallax Forums

suitable higher current replacement for ULN2803

jeromelabjeromelab Posts: 31
edited 2009-01-05 06:21 in Propeller 1
Hello,

I need to switch ledstrips circuits from propeller, each circuit is 1.2A at 12V, ULN2803 or ULN2003 can output 0.5A at most (and that is with paralleling all inputs/outputs). I am looking for something similar to a transistor array as I have 40 ledstrips circuits to switch. The board size is limited too, so I am looking for a space saving solution (so no TIP mosfets). I cannot find anything that would fit the bill, could someone point me in the right direction?

Thanks

Comments

  • BTXBTX Posts: 674
    edited 2008-12-30 14:40
    Maybe you'll need to do some more space to put all that, If you have 40 ledstrips and 1.2A each, is not easy to find what you want, look for some setepper motor controllers (although I think you´ll need more space than using SMD MOSFETS). L298, L6203, etc, or IRFZ's MOSFET series in SMD.

    You have a lot of power to drive -> You´should have space for work, and to refrigerate them.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Regards.

    Alberto.
  • Mike GreenMike Green Posts: 23,101
    edited 2008-12-30 14:48
    Here's a dual driver capable of handling 1.5A per channel. It comes in DIP packaging for breadboarding as well as surface mount to save space.
    www.micrel.com/page.do?page=/product-info/products/mic4426.shtml
  • LeonLeon Posts: 7,620
    edited 2008-12-30 16:18
    They are really intended for driving high-capacitance loads like MOSFETs (I'm using one of them for that), I'm not sure if they would be all that suitable for LEDs. The spec. would need careful checking.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
    Suzuki SV1000S motorcycle
  • jeromelabjeromelab Posts: 31
    edited 2008-12-30 18:08
    I will try the mic4426, stepper motor controllers tend to be very expensive. I may give a shoot at ULN2064/75 as well.

    Thanks!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2008-12-30 19:37
    I agree with Leon: MOSFET drivers, like the MIC4426, are designed to provide high peak currents, but not high sustained currents. As such, they are perfect for driving transient loads, such as MOSFET gates, piezo devices, etc., but not LEDs, motor coils, and the like.

    -Phil
  • jeromelabjeromelab Posts: 31
    edited 2008-12-31 10:03
    A darlington array is better then? or the IRFZ's MOSFET series in SMD? The footprint of 20 SOIC8 chips is fitting and is nicely spaced in my 190 * 65 mm PCB.
    I am not familiar with the drawback of most transistors but this one seems suitable www.zetex.com/3.0/product_portfolio.asp?pno=FMMT617, any opinion?
    There will be a ventilator in the enclosure, blowing directly on the chips, and if necessary some heatsinks.

    Post Edited (jeromelab) : 12/31/2008 10:33:43 AM GMT
  • Fred HawkinsFred Hawkins Posts: 997
    edited 2008-12-31 10:56
    worth a look?



    TI sez...
    16-Channel, Constant-Current LED Driver with Schmitt-Trigger Input: The TLC5925,
    designed for LED displays and LED lighting applications with constant-current
    control and thermal shut-down protection, is capable of driving up to 45mA per
    channel.
    - 16-bit shift register and data latches convert serial input data into parallel
    ·output format
    - At output stage, 16 regulated current ports are designed to provide uniform
    ·and constant current for driving LEDs within a wide range of VF variations
    - Applications: Gambling/gaming machines, video billboards, general full-color
    ·LED displays, highway signage, information signs
    TLC5925I Price (1ku): $0.55
    http://sdm3.rm04.net/ctt?kn=87&m=2287328&r=MTUxMjAzMTgwNjAS1&b=2&j=NjIwODg0MjAS1&mt=1&rt=0
    and
    • 16 Constant-Current Output Channels
    • Constant Output Current Invariant to Load Voltage Change
    • Excellent Output Current Accuracy:
      • Between Channels: < ±4% (Max)
      • Between ICs: < ±6% (Max)
    • Constant Output Current Range:
      3 mA to 45 mA
    • Output Current Adjusted By External Resistor
    • Fast Response of Output Current, OE (Min): 100 ns
    • 30-MHz Clock Frequency
    • Schmitt-Trigger Inputs
    • 3.3-V to 5-V Supply Voltage
    • Thermal Shutdown for Overtemperature Protection
    • ESD Performance: 1-kV HBM
    • APPLICATIONS
      • Gaming Machine/Entertainment
      • General LED Applications
      • LED Display Systems
      • Signs LED Lighting
      • White Goods

    The TLC5925 is designed for LED displays and LED lighting applications. The TLC5925 contains a 16-bit shift register and data latches, which convert serial input data into parallel output format. At the TLC5925 output stage, 16 regulated-current ports provide uniform and constant current for driving LEDs within a wide range of VF variations. Used in system design for LED display applications (e.g., LED panels), the TLC5925 provides great flexibility and device performance. Users can adjust the output current from 3 mA to 45 mA through an external resistor, Rext, which gives flexibility in controlling the light intensity of LEDs. TLC5925 is designed for up to 17 V at the output port. The high clock frequency, 30 MHz, also satisfies the system requirements of high-volume data transmission.

    The serial data is transferred into TLC5925 via SDI, shifted in the shift register, and transferred out via SDO. LE can latch the serial data in the shift register to the output latch. OE enables the output drivers to sink current.
    Mouser has the 5926 flavor (w/ 256 gray) for $1.57 each. Only 120ma though.


    Post Edited (Fred Hawkins) : 12/31/2008 5:43:18 PM GMT
  • jeromelabjeromelab Posts: 31
    edited 2008-12-31 11:05
    45mA instead of the required 1200mA, it will not make it. I just need simple on/off switch, no pwm or correction.But thanks for the info, price is definitevely right! I have been using the TLC5930 in the past(much more expensive), it seems pretty similar, it worked fine but failed always when limiting current (and slightly overheating).
  • BTXBTX Posts: 674
    edited 2008-12-31 11:37
    IRFZ´s have a better performance that common darlingtons, in all ways.
    Just use two boards in sandwich if you need more space, or three.
    Try a first assembly with only one IRF and test it for many time, check temp, time ressponse, drop voltage, etc.
    Warning, it is reccomended that you use a resistor in paralel with the mosfet gate and gnd.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Regards.

    Alberto.
  • Fred HawkinsFred Hawkins Posts: 997
    edited 2008-12-31 11:48
    Oops, sorry about the ma mistake. (Even with the TI suggested kludge of using multiple channels to hit a current mark, 10 per in this instance, using the 5926 won't work.)

    This one hits 1.2a but only one output. Comes in two bug-on-the-plate arrangements -- how's your surface mount handling?

    http://focus.ti.com/docs/prod/folders/print/tps61050.html

    Post Edited (Fred Hawkins) : 12/31/2008 11:55:27 AM GMT
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2008-12-31 13:51
    The continuous current on multiple outputs would rule out most ICs I reckon. However you could go for dual logic level MOSFETs in tiny 6-pin SOT packages (3mm sq) that can handle 3 amps per transistor. One of these devices is the FDC6401N.

    For 12V non-inductive loads you can just run these directly from 3.3V logic and although you will need 20 of these to drive your 40 ledstrips the devices take up less space than the connectors. Just think of a little black dot next to every two outputs.

    Obviously the Prop doesn't have 40 I/O lines so I guess you would just use five cascaded 74HC164 or 74HC595 shift registers to drive the MOSFETs.

    *Peter*
  • Fred HawkinsFred Hawkins Posts: 997
    edited 2009-01-05 06:21
    This is a nice puzzle, and alas, no better solutions. But poking around TI did turn up this chip, which although it doesn't hit·jeromelab's heavy amperage requirement, it does more than most: TPIC6A595

    If I read this one right, it's a 595 8bit·shift register (cascadable like usual) that handles 350mA of 50v on each line. There's four flavors of this chip with various ratings:

    tpic6595 45V 250mA continuous, 1.5A pulsed peak (only 2% duty though, 100µs duration)

    tpic6a595 50V 350mA continuous, 1.1A pulsed peak (ditto on limits for this and the rest)

    tpic6b595 50V 150mA continuous, 500mA pulsed peak

    tpic6c595·33V 100mA continuous, 250mA pulsed peak

    Then there's the TPIC6596. I don't get the difference, but the logic block diagram is slightly different.·It's 45V 250ma, but has a 2A pulsed peak.
Sign In or Register to comment.