AD5686 output current & headroom

Dear Analog,

I'd like to drive four resistive heaters with the AD5686. The application requires sourcing up to 35mA continuously on all four channels at voltages that vary between 1.5 V and 2.5 V.

Fig. 24 suggests that this is possible with a supply voltage of 5V, but there is not a lot of (head)room for error. On the other hand current limiting above 40 mA is a desirable safety feature.

- Would you recommend the AD5686 for this application?

- Should I expect the shoulder of the curve on fig. 24 to shift significantly with temperature or from chip to chip?

- For a given voltage Vout < Vref, does the choice of gain = 1 or gain = 2 impact the current sourcing capabilities at Vout?

Kind regards

Parents
  • +1
    •  Analog Employees 
    •  Super User 
    on Aug 31, 2021 2:26 AM

    Hi, 

    First, to clear up the headroom graph, this just means that you need to provide sufficient headroom on your supply for your required voltage out and current load. In your case, with output ranging up to 2.5V, having a 5V supply should be sufficient and would cause no voltage drop on the output at 35mA. 

    - Should I expect the shoulder of the curve on fig. 24 to shift significantly with temperature or from chip to chip?

    Yes, it could be, but if your max out is 2.5V, you already have a 2.5V headroom from VDD, which should be enough gap to cover for the headroom. 

    - For a given voltage Vout < Vref, does the choice of gain = 1 or gain = 2 impact the current sourcing capabilities at Vout?

    the gain setting doesn't actually affect the current sourcing capability of this chip. 

    - Would you recommend the AD5686 for this application?

    To be honest, 35mA continuous current is quite close to the overcurrent limit which is 40mA. As designers, we generally don't want to play near the limits of our devices. I could say that the AD5686 would work just need to look at the power dissipation on the device, make sure that the ambient doesn't get too hot. you're looking at (5V-1.5V)*35mA*4=0.49W of power on the device, assuming 0.1W system consumption, your already working at (0.49W+0.1W) * 112.6C/W= 66.43C for TSSOP package. The abs max Tj is 125C, so the ambient shouldn't reach, 125C-66.43C=58.57C on your board to make sure the device doesn't break. 

    Best regards,

    Ian

  • Thank you for the thorough reply, that answers my questions. The current limit being so close is in fact a desirable feature in this application, as long as it is well specified at 40mA — the load itself must be protected against overcurrent and is driven relatively close to its maximum power dissipation in normal operation. As regards the power dissipation in the DAC I have selected the LFCSP package and I will be adding some series resistance on the lower voltage channels to bring them up to 2.5V @ 35 mA.

Reply
  • Thank you for the thorough reply, that answers my questions. The current limit being so close is in fact a desirable feature in this application, as long as it is well specified at 40mA — the load itself must be protected against overcurrent and is driven relatively close to its maximum power dissipation in normal operation. As regards the power dissipation in the DAC I have selected the LFCSP package and I will be adding some series resistance on the lower voltage channels to bring them up to 2.5V @ 35 mA.

Children
No Data