TPS709-Q1: Output voltage is keep incresing with temperature increase

Hi Ganapathi, 

Across temperature, the DC accuracy of the TPS709 is +/-2% and 2% on top of 1.2 is 1.224V. With an ambient temperature of 93C, assuming the load is 10mA, and LDO is dissipating 0.168w of power, and this will result in a junction temperature rise of 0.168 x 210.9 = 35C; depending on the layout, the actual temperature range could be higher/or lower than the estimated 35C. 

With this being said, the junction temperature of the TPS709 with 93C ambient could go over 125C and the output could near or slightly above 2% accuracy. In the meantime, I am not sure how the output voltage was measured, there could be measurement error as well; for accurate measurement, we will need to have a dedicated sense line in connecting right to the DUT of the output of the LDO, the return path for the instrument also needs to have the shortest connection to the DUT ground. 

Since we suspect the LDO is moving slightly out of the accuracy due to higher than 125C recommended junction temperature, what's the main concern you have for the application? 

Regards, 
Jason Song

Hi Jason,

Thanks a lot for quick reply. We are measuring this voltage through ADC. Agree with your point, once Tj>125 Dec, we cant guarantee *** operation. And also, I observed  ADC count was keep increasing while temperature is increasing (from 1.2V towards 1.23V). Usually the voltage shall decrease with increase in temperature. Not sure what causing the voltage to increase.

And, I noticed LDO output voltage became 0 once ambient temperature is moved >93 Deg C. Thermal shutdown limit is 158 Deg C. As per our application, at this ambient temperature , Tj can be around 130 Deg. C only.

Would like to know this LDO behavior w.r.t temperature increase as its output is not stable and keep changing w.r.t ambient temperature. Below is the circuit I am using, VIN is 18V.

Thanks & Regards

Ganapathi

Hi 

Thanks a lot for quick reply. We are measuring this voltage through ADC. Agree with your point, once Tj>125 Dec, we cant guarantee *** operation. And also, I observed  ADC count was keep increasing while temperature is increasing (from 1.2V towards 1.23V). Usually the voltage shall decrease with increase in temperature. Not sure what causing the voltage to increase.

--This is not true for LDO in general and the output at higher temperatures could go higher or lower depending on the internal reference characteristics. You should also check the voltage sensing accuracy with the ADC at high temperatures. 

And, I noticed LDO output voltage became 0 once ambient temperature is moved >93 Deg C. Thermal shutdown limit is 158 Deg C. As per our application, at this ambient temperature , Tj can be around 130 Deg. C only.

--Did you check this with your ADC? I would double-check this with an oscilloscope, even the device is entering the thermal shutdown, once the temperature has dropped, the part will start tunning on again. When the part is entering and existing thermal shutdown, the output voltage will also go up and down. The output will not be latched off for thermal shutdown protection. 

Would like to know this LDO behavior w.r.t temperature increase as its output is not stable and keep changing w.r.t ambient temperature. Below is the circuit I am using, VIN is 18V.

--Figure 7 was taken with a very limited number of units, but we can see that the output tends to be lower. I suggest checking how the ADC is sampling the output: how many samples? any average? what's the aperture time for the sample? what's the resolution when sampling a known good reference, how repeatable on the counts? 

Regards, 
Jason Song