Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Temperature measurement on power mosfets during the power consumption in motor contro

Status
Not open for further replies.

vishjay

Member level 2
Joined
Feb 1, 2019
Messages
42
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
306
Hi,

I am measuring case temperature placing the thermocouples on mosfets in the bridge configuration of motor control board. During the motor rotation or power consumption the six power mosfets are getting warm in my experiment and to monitor if the case temperature exceeds rated junction temperature of power mosfet's i placed thermocouples to get the data in degree centigrades. The question is how long we have to let it in running condition? How long we have to wait to check if case temp exceeds jn temp of power mosfet?
 

There are a lot of factors at play here, specifically, the physical aspects of your thermocouple, how it's bonded to the MOSFET, etc. Look at this: https://www.omega.com/techref/ThermocoupleResponseTime.html

But if it were me, I'd just let it run continuously until the temperature stabilizes. Aren't you expecting it to run continuously in normal operation?
 

Hi,

Mosfets include a body diode. This is the same silicon as the mosfet's switch, obviously they have the same temperature.
A silicon diode has a temperature coefficient of about 2.1mV/°C in forward voltage.
Thus a temperature rise from 25°C to 150°C = 125°C results in about 260mV.

If you can find a way to measure the diode's forward voltage - eg in times when the PWM switches the Mosfet OFF - you durectly measure the junction temperature. ...almost perfect without adding time delay or added thermal mass.

This is not very easy, but If you want us to help you with this solution you should give us the half bridge schematic and the PWM timing.

*******
The theromcouple solution is more simple, but includes delays and can only read the case temperature instead of junction temperature.
The thermal model from junction to case may be simplyfied as a simple RC delay.
To get a clue about the junction temperature you need to do a multiplication (R part) and you need to take care about the rise rate of the temperature.
This can simply made in an EXCEL chart.

Klaus
 
If you can find a way to measure the diode's forward voltage ...

A very elegant solution indeed.

But perhaps you need to add /modify the circuit extensively; forward drops are best measured at constant current.

I was about to suggest a silicon small signal diode pasted on the body instead of the thermocouple.

Diode forward drops depend on band gap in some complex way and can give the absolute temp without the need of a reference junction.

Measuring the temp using the built in body diode can become messy (but I am fully ignorant of the intricacies involved). But I feel that replacing the thermocouple with a simple diode may be simpler.
 

The question is how long we have to let it in running condition? How long we have to wait to check if case temp exceeds jn temp of power mosfet?

Best test this is at environmental temperature 36C for 48 hours. It will pass the test if your Heat-sink it does qualify too.
 

Datalog the thermocouple readings and look for an
asymptote. Which may change with ambient temp,
airflow and load, but you will find one of you keep
conditions constant.

RTDs are maybe a better signal in noisy environments.
Or the diode approach.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top