Variation of NTC in HM3

Hiroshi

Hi Wolfspeed Expert,

I have a customer purchased HM3 modules to develop inverter application.

When they read NTC valve, they found there's a module which has 2°C higher even at room temperature (no load applied) by comparing to other modules.

When operating inverter with HM3 modules, the temperature difference rises up to 6°C. This would be due to variation between SiC MOSFET chips, however customer can’t get why NTC value is different at room temp. 

 

Question 1: Regarding the thermal resistance Rthjc value described in the datasheet, is this the value when only high-side or low-side is heated? Or the both high-side and low-side are heated? Since NTC is placed near low-side, the customer wants to clarify how Wolfspeed heated the device during measurement.

 

Question 2: Could you please provide the document describing the method how Wolfspeed measured the thermal junction?

 

Question 3: Is the NTC value difference (2°C @ RT) within the spec? Also, please advise the reasons for the NTC variance from your perspective.

Thank you for your support.

Best Regards,

Hiroshi

TBhatia

Hello, thank you for your message. A member of our technical team is looking into your questions and we will respond shortly. 

PaulKanatzar

Hiroshi,

Question 1: We characterize the Rthjc with the industry standard one switch position at a time. We are not heating both switch positions simultaneously.

Question 2: I do not have a document available that I can share at this time. We use the Siemens T3ster thermal characterization system to characterize the thermal impedance of the module. The equipment calibrates to each module and measures the junction temperature through the device itself. We are not using the NTC for any of that characterization.

Question 3: The NTC has a base tolerance of ±1% and a ±1% range for the beta value from 25-85 °C. Those tolerances in conjunction with the measurements not being taken in a controlled environment immediately indicate an issue to me. It's likely a combination of part variation and environment.

Do we have an NDA in place with this customer?

Thanks,

Paul Kanatzar

Hiroshi

Hi Paul,

 

Thank you for the answers. Yes we have a NDA with the customers in place. Please allow me to ask further clarification.

Question 1: Was low-side heated during Rthjc measurement? Or high-side?

Question 3: Besides variation of NTC and environment, are there any factor causing such variation, for instance, assembly quality? What’s the pass/fail criteria during EOL testing?

Thank you. 

Best Regards,

Hiroshi

PaulKanatzar

Hiroshi,

Question 1: We heat the high side and the low side individually multiple times and use the the worst case Rthjc measurement for the datasheet.

Question 2: It is possible that the modules that agree are higher performance than the datasheet specification and the third meets the datasheet specification. The pass fail criteria is set to ensure the module complies with the datasheet. I do not believe we share the pass fail criteria externally.

Given that we have an NDA in place I would highly suggest utilizing the thermal models to evaluate the customers system.

-Paul

TBhatia

Hi, I hope that this answered your question. I will close this discussion for now but if you have a follow up question, please "Start a New Discussion" and we would be glad to support you further.