• Resolved

DAC8718: device heating up (when negative voltage AVss applied)

Expert 3275 points

Replies: 7

Views: 95

Part Number: DAC8718

Team,

this is a continuation of the previous thread about DAC7718 heating up to 45 degrees C when the negative -5V voltage is applied to AVss (through a 1kOhm resistor). 

We now use DAC8718 and the problem with heating persists. We tried not connecting thermal pad to ground as advised, it didn't change anything.

Is this expected behavior (this device can have such temperature) or something is wrong?

Thanks for your help.

  • Hi Bart,

    Thank you for your query. Is it possible to share your schematics? You can share it over email if there is any confidentiality issue. or else, please share the DAC part of the schematics.

    Please also let me know what is the power-supply and reference sequencing you are following.

    Regards,

    Uttam Sahu

    Applications Engineer, Precision DAC

  • Expert 3275 points

    In reply to Uttam Sahu:

  • In reply to Bart:

    Hi Bart,

    There could be a few reasons for this device heating up.  Firstly, what is the load on the outputs of the DAC?  Is it capacitive? How much current is the DAC consuming? 

    The device is rated for above 45°C, but there could be a problem in your circuit.  What is the voltage drop across the 1kΩ resistor?  What is the actual VSS, AVDD, and VREF voltages at the pin of the device?

    Note that this device requires a specific power-on sequence for the supplies - are you complying with that sequence?

    Thanks,

    Paul

  • Expert 3275 points

    In reply to Paul_Frost:

    Hi Paul,

    the outputs were left without any load.

    Our power supply is made in a simple way (on the isolated side) and both AVss and AVdd are turned on at the same time. In this design we have no control over this to make sure one of those is before other. Nontheless we manually tried to power up in a sequence as in the app note, though final effect is the same (device is heating up slightly anyways). 

    The difference in temperature between ambient and the device is not greater than 10 degrees C. The difference in consumption between the case where negative voltage is used and without it is 15mA. Perhaps using negative voltage for supply is causing the heating effect and it needs to be like that?

    Other than this, device is working very well without any issues.

  • In reply to Bart:

    Hi Bart,

    If AVSS is not applied, the device is probably in a POR state.  15mA is a bit higher than expected on this device if it is unloaded - are there other things in the design that could be drawing current? 

    For example, using the electrical characteristics table, we see a max IQ AIDD of 6mA and AISS of 4mA.  Assuming ±10V supplies, we have a total power from those supplies of 10V×6mA + -10V×-4mA = 100mW. If you are using the QFP package, then the θJA spec is 55°C/W.  This means you should be seeing about 5.5°C increase. 

    Can you share your schematic? Is there some other load in the system?

    Thanks,

    Paul

  • Expert 3275 points

    In reply to Paul_Frost:

    Hi Paul,

    I confirm there is no load and output voltage is set to 0V. See the sch.

    Kind regards

  • In reply to Bart:

    Hi Bart,

    I do not think this temperature increase is a major issue, as it is still within the thermal limits of the device.  My only other advice would be to try to confirm where the current is coming from.  This might entail disconnected pins to remove any load or current (such as VMON load), measure the current on each supply (AVDD, AVSS, IOVDD, DVDD), and verify that nothing is oscillating (DAC_REF, V_OUTx, VMON).

    Thanks,

    Paul