Post Go back to editing

ADP8140 Analog dimming

Hi,

I am using ADP8140 for driver of two string of white led. Both string has different light color so I'm using FET for controlling bot lines so that I can shut down or light up any string I want. I have DAC connected to DIM pin for adjust LED current. I have two LED string, each string has parallel 4+4 LED. I'm feeding LED straight from my power supply, +24V. 

Here is my current schematics.

I already add PU for FAULT as I was missing that and also route EN to VREG.

At COMP, I have RC network, C10 0.47uF and R9 1K. 

When I turn on my power +24, led light up immediately and driver start to sink current as much as Rset (R11) resistor allow. Changing voltage at DIM pin does not effect at all.

Any Idea, what is wrong at my design?

-Risto

  • Hi risto.t,

    What voltage are you forcing to the DIM pin? It should be between 0V to 2V for it to linearly scale the LED current. If DIM is set to 2V or higher, then it would set the current sink to as much as the Rset would allow:

    Let me know if this works, and just in case it doesn't can i know what the SINKx voltages are? Thanks!

    Regards,

    Doc

  • Hi Jon,

    Even that DIM pin is 0V, ADP8140 is sinking current as much as the Rset allow.

    Do I need to care about MIN and VT pins if I'm using DIM pin for dimming? Right now MIN is tied to GND and VT is floating.

    SINKx voltage is 11.8V. I'm feeding +24V to the LEDs from external power supply.

    Thanks,

    -Risto

  • Hi Jon,

    I did more testing today.

    Funny thing is that even EN pin is low, driver start to sink current as soon as I enable FET which drives LED. I could not find any short to the ground at driver. 

    What I found out was that VT pin has 3.0V even that its not connected VREG or other power supplies. Is that normal?

    Here is FET circuitry. T2 and T4 are controlled by IO-expander.

    Here are LED string.

    Could you check if IC what we are using is ok. Code on top of the IC is ADP8149 #1447 7319

    Thanks for help.

    -Risto

  • Hi Risto,

    I had a quick look at the lab with your setup and that is the normal response of the chip. How did you connect FB_OUT to control T1 and T3? Based on the result on the lab, i think(please confirm) what you are lacking is the connection of FB_OUT to control the MOSFETs. You could add a switch to connect the gate of T1/T3 to either FB_OUT or to the 24V supply. That way you could give the control to ADP8140 once it is enabled.

    Regards,

    Doc

  • Hi Risto,

    I tried a similar setup but with the gate of the PMOS connected to FB_OUT and it works fine, so i think it is necessary. You'll need to do something similar to the image below(per string) to make your application work:

    When open, it disables the PMOS.

    When short, it gives the control to ADP8140.

    You might be interested in LT3475 or LT3964 if it fits your application or use the selection guide to find a better fit to your application.

    Regards,

    Doc

  • Hi Jon,

    I have not connect FB_OUT anywhere right now.

    T1 and T3 and controlled by T2 and T4. T2 and T4 are controlled by external IO-expander. This is because I need to be able to shut down both LED string separetly. 

    Is it mandatory to connect FB_OUT, does the driver work without it?

    Thanks for your support.

    -Risto

  • Hi Jon,

    I got my circuit working after I connect feedback to the switch FET as you suggested.

    Thanks for your good support.

    -Risto

  • Hi,

    We have now the problem that both LED string cannot turn on same time. Current circuit below. How do you see, is there errors at circuit? Is there problem for IC if other string is on and other off by FET? 

    Thanks,

    -Risto

  • Hi Risto,

    Sorry this did not notify in my email. The image is a bit blurry so i'll need to confirm with you some points:
    1. can you confirm that you used an NFET?
    2. what is the "high" logic that you are using to turn-on the LED string. You might need a bit high level for this as FB_OUT can be quite high and you'll need some extra voltage for Vgsth of the FET.
    3. can you try using a PMOS device? try using an open-drain gate drive so you won't need to have a very high logic level. You could just float it to your pull-up voltage or pull it down to ground.
    4. What voltage are you measuring at FB_OUT?
    5. What is the difference of this circuit to your previously working one?

    Regards,
    Doc