Mar 10, 2020
06:21 AM
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Mar 10, 2020
06:21 AM
Hi guys,
I’m using an IKW40N120H3 in a boost converter.
What is the best way to measure the junction temperature of a device?
Thanks!
I’m using an IKW40N120H3 in a boost converter.
What is the best way to measure the junction temperature of a device?
Thanks!
Solved! Go to Solution.
Labels
- Labels:
-
ispn:3078:1:0
-
l1:144:1:0
-
l2:219:1:0
-
l3:220:1:0
- Tags:
- IFX
1 Solution
Mar 11, 2020
03:52 AM
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Mar 11, 2020
03:52 AM
Hi Walter,
there are several ways of measure the temperature of a device in the application.
However, please let me highlight that the 'real' junction temperature is very hard to be measured by experiments. The best way is generally by means of simulation using the equivalent thermal model provided in the datasheet (you can have a look at this application note: https://www.infineon.com/dgdl/Thermal+Modeling.pdf?fileId=db3a30431441fb5d011472fd33c70aa3)
Alternatively, a mathematical approach based on the estimation of the power losses can be used, together with the thermal impedance provided in the datasheet of the device. Depending on the accuracy of the power losses estimation and the knowledge of the thermal system (e.g. including isolation foils, heatsink, etc.) this calculation may be quite precise or pretty rough.
If you intend to perform an experimental measurement of the device temperature, the quickest approach is to use a thermocouple glued on the package. If you can afford to use an isolated thermocuple, the best position to place the probe is directly on the collector pin, in the proximity of the package body.
Alternatively, you can use a thermal camera to get an estimation of the hottest device temperature, that is usually in correspondence of where the chip is located within the package.
Regards,
Giuseppe
there are several ways of measure the temperature of a device in the application.
However, please let me highlight that the 'real' junction temperature is very hard to be measured by experiments. The best way is generally by means of simulation using the equivalent thermal model provided in the datasheet (you can have a look at this application note: https://www.infineon.com/dgdl/Thermal+Modeling.pdf?fileId=db3a30431441fb5d011472fd33c70aa3)
Alternatively, a mathematical approach based on the estimation of the power losses can be used, together with the thermal impedance provided in the datasheet of the device. Depending on the accuracy of the power losses estimation and the knowledge of the thermal system (e.g. including isolation foils, heatsink, etc.) this calculation may be quite precise or pretty rough.
If you intend to perform an experimental measurement of the device temperature, the quickest approach is to use a thermocouple glued on the package. If you can afford to use an isolated thermocuple, the best position to place the probe is directly on the collector pin, in the proximity of the package body.
Alternatively, you can use a thermal camera to get an estimation of the hottest device temperature, that is usually in correspondence of where the chip is located within the package.
Regards,
Giuseppe
1 Reply
Mar 11, 2020
03:52 AM
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Mar 11, 2020
03:52 AM
Hi Walter,
there are several ways of measure the temperature of a device in the application.
However, please let me highlight that the 'real' junction temperature is very hard to be measured by experiments. The best way is generally by means of simulation using the equivalent thermal model provided in the datasheet (you can have a look at this application note: https://www.infineon.com/dgdl/Thermal+Modeling.pdf?fileId=db3a30431441fb5d011472fd33c70aa3)
Alternatively, a mathematical approach based on the estimation of the power losses can be used, together with the thermal impedance provided in the datasheet of the device. Depending on the accuracy of the power losses estimation and the knowledge of the thermal system (e.g. including isolation foils, heatsink, etc.) this calculation may be quite precise or pretty rough.
If you intend to perform an experimental measurement of the device temperature, the quickest approach is to use a thermocouple glued on the package. If you can afford to use an isolated thermocuple, the best position to place the probe is directly on the collector pin, in the proximity of the package body.
Alternatively, you can use a thermal camera to get an estimation of the hottest device temperature, that is usually in correspondence of where the chip is located within the package.
Regards,
Giuseppe
there are several ways of measure the temperature of a device in the application.
However, please let me highlight that the 'real' junction temperature is very hard to be measured by experiments. The best way is generally by means of simulation using the equivalent thermal model provided in the datasheet (you can have a look at this application note: https://www.infineon.com/dgdl/Thermal+Modeling.pdf?fileId=db3a30431441fb5d011472fd33c70aa3)
Alternatively, a mathematical approach based on the estimation of the power losses can be used, together with the thermal impedance provided in the datasheet of the device. Depending on the accuracy of the power losses estimation and the knowledge of the thermal system (e.g. including isolation foils, heatsink, etc.) this calculation may be quite precise or pretty rough.
If you intend to perform an experimental measurement of the device temperature, the quickest approach is to use a thermocouple glued on the package. If you can afford to use an isolated thermocuple, the best position to place the probe is directly on the collector pin, in the proximity of the package body.
Alternatively, you can use a thermal camera to get an estimation of the hottest device temperature, that is usually in correspondence of where the chip is located within the package.
Regards,
Giuseppe