DAC7760: DAC7760 overheating issue

Hi Rohidas,

What is the condition of the output when the device is overheating? Does the device get damaged?

The thermal pad should be connected to AVSS as your schematic indicates.

I also noticed that you are tying Vout and Iout together. When tying the outputs a buffer is required on +VSENSE to prevent current flow into the pin when Iout is enabled. I have linked a TI Design that shows how to do this below.

http://www.ti.com/tool/TIPD119

Thanks,

Garrett

Dear Garrett,

Thanks. Output is configured for 20mA output with 0ohm load connected at connector (outside the plastic enclosure). With this DAC7760 does not damage but its body temperature goes upto 100 C at 27C room temperature and upto 126 C at 60C chamber temperature.

Following is brief description of problem and our DAC7760 section design details:

Subject: DAC7760 IC body temperature goes upto approx. 110C at 27C room temperature (with 20mA output or 0V output)

Design detail: Design contains 2 Channel output. For this two no. DAC7760 IC used.

PCB details: Single 4 layer PCB design

Enclosure: PCB design enclosed in plastic enclosure.

Supply voltages to DAC7760 ICs:

For DAC7760, we have generated +15V, -15V and +5V supply from External 24V DC supply. To generate these voltages, we have used following power supplies in our design.

   1. 24V DC to +15V PSU DC to DC converter (TI make MC34063 based design) :

            Input supply range: 18V to 40V DC;  

            Output voltage: +15V DC ( range: 14.90 V to 15.15V); 

            Output  maximum load: 70mA

            Output Ripple: Approx. 180mV;  

            PSU efficiency: Approx. 75%;

            PSU switching frequency: approx. 130KHz

2. 24V DC to -15V PSU DC to DC converter (TI make MC34063 based design) :

           Input supply range: 18V to 40V DC;   

           Output voltage: -15V DC ( range: -14.90 V to -15.15V);  

           Output maximum load: 25mA

           Output Ripple: Approx. 30mV

           PSU efficiency: Approx. 40%

           PSU switching frequency: approx. 130KHz  

3. 24V DC to +5V PSU (LM317MDT linear voltage regulator based design) :

           Input supply range: 18V to 40V DC

           Output voltage: +5V DC ( range: 4.90 V to 5.15V)

           Output maximum load: 60mA

           Output Ripple: Approx. 20mV

Second point, we connected DAC7760 Thermal PAD to AVSS pin, but same thermal behaviour observed.

Should we check thermal behaviour with adding buffer at +VSENSE pin ? Does it improve thermal performance or Functional, EMIEMC performance?

Pls. write if above information is sufficient  or you require any more information for problem analysis.

Thanks & Regards

Rohidas Sawant 

Hi Rohidas,

Let me ask a few additional questions to understand your situation.

1. Is the output at a stable 20mA when the device is overheating?

2. Did you solder the thermal pad of the device to the PCB? This is the main way the device rejects heat. I know you said it was unconnected electrically before.

3. What is the current flowing from each supply when the device is overheating? This will help determine if the power dissipation is reasonable.

Thanks,

Garrett

Dear Garrett,

Thanks.

Following are the answers to your questions:

1. Output is stable at 20mA when the device is overheating.

2. We have soldered the thermal pad of the device to the PCB. Below IC there is Copper area. This area is approx. same as mentioned 

    in  datasheet This Copper is not connected to AVSS electrical signal i.e. it is left open.

3. We measured currents at Room  25 C temperature. We have measured currents when both DAC ICs are configured for 20mA output.

    At this time power supply currents are as follows. This current is of both DAC7760 ICs.:

     a) +15V supply current : 52.81mA

     b) -15V supply current : 9.51mA     

We measured this currents immediately within  3 to 5 minutes after Module is powered ON. We did not measured current, when DAC7760 IC body temperature goes upto 100C. DAC7760 IC body temperature goes upto 100C within 45minutes to 1 hour.

We will measure currents again when DAC7760 body temperature is high i.e. approx. 100 C and share to you.

4. We measured body temperature of DAC7760 IC as well as MC34063 ICs, LM317 IC. The temperatures are as follows: 

     a)  DAC7760 IC:  106 C

     b)  24V to -15V PSU TI make MC34063A IC : 96 C

     c) 24V to  +15V PSU TI make MC34063A IC : 90 C  (approx.)

     d) 24V to +5V PSU LM317 IC ( SMD 3Pin TO-252 Package): 91 (approx.)

Thanks

Rohidas Sawant

Dear Garrett,

  a) +15V supply current : 52.81mA  &   b) -15V supply current : 9.51mA 

These currents are for both DAC7760s not of single IC.

We are taking temperature, current, voltage readings again with +/-15V supply for DAC7760 ICs. Also we are taking these readings with +/-12V supply for DAC7760 ICs. We are getting improvement with +/-12V supply. We will also observe the supply voltage waveforms at start up. We will provide the waveforms and readings to you for analysis and suggestions.

Thanks & Regards

Rohidas Sawant

Hi Garrett,

We have taken readings with +/-12V DC Analog Power supply to DAC7760 ICs. We observed that with this change, component case temperatures ( for e.g. of DAC7760 ICs, MC34063 ICs, LM317, Inductors ) gets reduced by approx.. 12 to 15 C.

Now at 60C atmospheric temperatures and module with plastic enclosure, component case temperatures are as follows:  

 1. DAC7760 IC case temperature is approx. 99 C,

 2. MC34063 IC case temperature is approx. 90 C

 3. LM317 IC case temperature is approx. 99 C.

In 24V DC to +!2V DC PSU design (Step down), MC34063 IC is used & its other component selected are as follows:

   CT = 100pf;   L = 56uH/0.6Amp;   RSC = 0.5 Ohm / 1 WATT;   CO = 40 uF;  

In 24V DC to -12V DC PSU design (voltage inverting converter), MC34063 IC is used & its other component selected are as follows:

   CT = 100pf;   L = 56uH/0.6Amp;   RSC = 0.5 Ohm / 1 WATT;   CO = 40 uF;  

We also measured actual current consumption by module. Readings are as follows:

1.  +12V supply: 52mA. ( for 2 nos. DAC7760 + 6 mA bleeder resistor current)

2. -12V supply: 18mA. ( for 2 nos. DAC7760 + 7.5 mA bleeder resistor current).

Our analysis is, with above component power dissipation and its maximum acceptable Tj value, observed component case temperature is within limit.

By changing analog power supply form +/-15V to +/-12V, we are getting improvement. But still we want more improvement. So our queries are as follows:

1. For DAC7760, for better performance,  is +/-12V supply OK ?

2. For MC34063 PSU design, are component selected OK ?

3. Is there any PCB layout design guideline documents for DAC7760 IC and MC34063 IC 

Pls. guide us for above queries.

Thanks & Regards

Rohidas Sawant

Hi Rohidas,

The current level for the 2 DAC7760 ICs seems fairly reasonable and I would not expect for it to be causing the majority of the on board heating. At 12V this looks like too much temperature increase for the device at 60C ambient but this could be due to the plastic enclosure. I have attached a thermal image of the EVM driving a 10 ohm load at 24V to show the typical thermal performance. This is not enclosed though as your board is and the DAC7760 is the main device causing the heating.

Find my response to your questions below.

1. Using +/-12V supplies is fine and will reduce power dissipation but will also reduce the load resistance that can be driven. The maximum load that can be driven will be (12V-2V)/.02 = 500ohm due to compliance voltage requirements.

2. I am not sure about the power supply components you have used. The datasheet has design procedures for step-down converters that show how to calculate component values. If you have additional questions you can post to our power management E2E forum.

3. There are layout guidelines in each device datasheet.

Can you provide waveforms of AVDD, AVSS, voltage at Iout, and a thermal image if possible during operation? 

Can you measure the current flowing from the 24V supply that powers the step down converters?

It would also be helpful if you repeated the thermal measurements outside of the enclosure to see if the temperature rise matches the expected value.

Thanks,

Garrett