Thursday, October 24, 2013

Single Phase Bridge Thermal Paths

Single-phase bridges are hot!  Especially high voltage ones.  Bordering on power modules, and highly application specific, high voltage (1kV and higher) 1P bridges are the thermal divas of the rectifier world.  Why is that?  Well, partly due to the magnitude of high reverse voltages, high leakage currents, and forward conduction losses.

During the recovery phase of a diode, it is blocking high voltage.  In a less-than-ideal world, the high reverse voltage generates a ‘leakage’ current in the diode.  This happens in low voltage systems too, but because the reverse voltage is much lower, it is generally ignored.  Ignore it in high voltage systems at your own risk!  Here’s why….

Peak Reverse Power

Peak reverse power is defined as the product of peak reverse voltage and leakage current, and is expressed as


P(reverse) = Vrwm * Ir

At Vrwm = 10kV, Ir = 500nA, P(reverse) = 5mW.  No a big deal, right? 


Forward Dissipated Power

Okay, so now let’s add in the forward power expressed as 


P(forward) = Vf * If

Where Vf = Forward Voltage drop, and If = Forward Current.  If Vf = 8V, and If = 1A, then P(forward) = 8W.


Total Dissipated Power

Total power, Ptot, is the sum of P(reverse) and P(forward) and is expressed as  

P(tot) = P(reverse) + P(forward)

In this example, Ptot is mostly Vf * If, and is slightly more than 8W (8.005W to be exact).  That’s at room temp (25°C).  But what if the base plate temperature is 50°C, and the thermal impedance of the device is 3°C/Watt?  That means the diode junction temperature is approximately

T(diode junction) = T(base plate) + T(Thermal Imp) °C/W * W

Or, T(diode junction) = 50°C + 3C/Watt * 8 Watt = 74°C.


The Impact of Reverse Recovery Time and Thermal Runaway

Did you know that for every 25-degree rise in temperature, leakage current triples?  It’s partly explained by temperature-sensitive reverse recovery times.  At temperatures go up, reverse recovery time gets longer which means the time to dissipate reverse power increases too.   At a diode junction temp of 74°C, the leakage current will be approximately six times the leakage measured at room temp.  In our example, at T(base plate) = 50°C, leakage current of the diode will be approx. 6 * 500nA = 3uA.  Now reverse power becomes 30mW, compared to 5mW at 25°C. 

As the temperature goes up, so does reverse power, which means the diode junction gets hotter, which means the reverse recovery time gets longer resulting in even more reverse power dissipation.  It’s easy to see how and why thermal runaway happens. 

The good news is thermal runaway can be prevented using one or more techniques.  Keep operating temperatures as low as possible, add heat-sinking capabilities, and derate components as much as as you can.  

Of course, your numbers will vary from this example.  There are other mitigating factors such as frequency and duty cycle, input signal, and the absence or presence of heat sinks.

Don't hesitate to call us if you have any questions about our single phase bridges or diodes.  We're here to help.

No comments:

Post a Comment