Hi ,
I am currently looking to calculate the junction temperature of the diode under different operating conditions . this could be a operation where the current is flowing continuously or at intervals it could have some high value of current . I understand that most of the losses will be conduction and very less switching losses , I am not sure how to quantify these losses , and get a equivalent temperature , Any info is greatly appreciated

 

Find the average current through the diode and multiply it by the voltage across the diode.

 

Hello fitch,

Calculating junction temperatures for the various options you state will require some work. I'm not sure why you want to know the junction temperatures specifically - other than ensuring you don't exceed the thermal rating of the devices.

You might consider some or all of the following:

1. Start by consulting the diode manufacturer's data sheet.
2. Are they single diodes or part of a 3-phase rectifier module?
3. The diode thermal resistance parameters, junction-to-case etc.
4. Is there or do you require a heat sink? What is its thermal resistance?
5. Is there forced air cooling? Maximum enclosure ambient temp expected?
6. Is the diode current flow continuous or discontinuous - if the latter, is there time for cooling to occur between current cycles?

Have fun anyway!

 

thanks for the info , The reason I am specifically , looking is to understand the temperature rise within the module for a given condition . This would be a three phase diode bridge. is there any papers/info that I could use to gain insight into how losses can be calculated ?

Thanks

 

The reverse leakage current in a diode or PN junction doubles for every 10 degree rise in temperature-which means that the device temperature can be approximated within its operating specs.

I=Is*2^(T-20/dT)
I=measured reverse leakage current at operating state
Is= reverse leakage current at room temperature given in spec sheet
T=Temperature of PN junction in degrees celcius
dT=10 degrees celcius

Solving for T in degrees celcius in the above equation;

T= [ln (x)/ln (2)]*10+20 where x = I/Is

Example: I = 100 uA (must measure this value), Is =15 uA at room temperature, given in spec sheet;

T= [ln(6.67)/ln(2)]*10+20 = 27.37+20 = 47.37 degrees celcius

 

If you are really interested in power losses in the rectifier module at a given operating condition, you need to go back to mik3's advice about calculating the diode losses - in reality you'll have to do the work and carefully consider each operating case you have in mind.

Most power rectifier manufacturers produce excellent design / application literature on their devices and the particular circuit parameters that apply for several design scenarios. Perhaps have a look at some of the power semiconductor manufacturer's websites.

Don't know of any papers on this topic - although any semiconductor textbook suitable for the serious power electronic designer's bookshelf would (no doubt) offer some well reasoned help.

Nice ideas from alwayslearning!

 

By way of example here is a link to a semiconductor manufacturer's web page which includes some on-line thermal design software. I'm not promoting their particular products, just the wealth of design tools that some manufacturer's provide to assist the designer. Obviously, they would like us to choose their products!

http://www.semikron.com/internet/index.jsp?sekId=311

 

This would be a three phase diode bridge. is there any papers/info that I could use to gain insight into how losses can be calculated ?

Yes, see the following paper and book. I've attached a copy of the paper which applies if the loads are approximately constant current (i.e. inductive). Analyzing 3-phase bridges is not that easy. The rectifier bridge is nonlinear and time-dependent, due to commutation/conduction states of the diodes. These references provide an average value model which is easier to work with. However, it takes a while to understand, and sometimes it's easier to make a brute-force numerical model. According to these references and a few others, in the general case, numerical models are needed if the load is not approximately constant current.

1. S.D. Sudhoff, H.J. Hegner and K.A. Corzine, “Transient and dynamic average-value modeling of synchronous machine fed load-commutated converters”, IEEE Trans. on Energy Conversion, Vol. 11, No. 3, September 1996.
2. P.C. Krause, O. Wasynczuk and S.D. Sudhoff, “Analysis of Electric Machinery and Drive Systems”, 2nd edition, IEEE Series on Power Engineering, Mohamed E. El-Hawary, Series Editor, Wiley Inter-science, ISBN 0-471-14326-X, 2002.