TMS320F28335: Questions on thermal metrics in the datasheet

Dengke,
You should look at the JESD51 standards. JESD51-2, JESD51-3, JESD51-7, and JESD51-9 apply to this device(as well as some others), but I think you will find JESD51-2 helpful.

Regards,
Cody

Cody，
I have read the document of JUSD51-2.The conditions for the temperature test in the document are very strict.
Is 85 degree in DSP datasheet tested under such conditions?
This test method is not applicable in our application.

Can you provide me with a workable way to evaluate the thermal reliability of DSP? The method does not need to be very precise, just a rough range. However, the selection of each parameter must correspond to the actual application environment and is convincing.

Look forward to your good news .

Regards,
Dengke

Dengke,

if you're just trying to get a good estimate, use a small gauge thermal couple and follow the instructions of the manufacturer for affixing it to the device. They typically have some recommendations and guidelines of how to get a good measurement. 

Regards,

Cody 

cody,

In fact, the reason why I want to assess the junction temperature of DSP is that during the thermal test of an outdoor product, I found that the case temperature of DSP has reached 85 degrees Celsius. I asked the engineer of TI, they said the device's temperature should not exceed 85 degrees Celsius. The reason why they say so is that DSP datasheet indicates that the maximum ambient temperature of DSP is 85 degrees Celsius . According to them, I have to modify the thermal design of the product, which is very difficult. I have looked at the document semiconductor and IC package thermal metrics, which refers to the junction temperature that can really affect the thermal reliability of the device, so I want to use the case temperature measured by the thermocouple line to evaluate the junction temperature. However, it is very difficult to select the thermal resistance parameters corresponding to my actual use environment. I can not make a reasonable assessment of the junction temperature of DSP. It annoyed me.

I did not see the instructions you mentioned that could guide me in thermal design assessment. Can you share it with me?

Dengke,

What you have been told is true, you should not exceed 85C with a "A" part. One option you may have is to upgrade to a "S" or "Q" part with some minor PCB changes, these have higher temperature ranges.

If you would like to continue to evaluate your junction temperature using your case temperature, Section 4 of SLUA844 describes what kind of thermocouple to use and how to attach it.

Hope it helps,
Cody 

cody,

As you have told me before, the case temperature is very close to the junction temperature. That is to say, the case temperature can also be close to 120C. 85C refers to the working environment temperature, environmental temperature and case temperature is not the same concept, right?

My question is not how to measure the case temperature, but how to evaluate the junction temperature after case temperature has been measured.

In the thermal assessment of products, I need to set a maximum safety value for case temperature, rather than just knowing that it is similar to the junction temperature. I want you to give me some advice on what I should do after I measure the case temperature.

regards,

dengke