How to interpret power supply IC specifications: Allowable losses

qwqwqw2088

How to interpret power supply IC specifications: Allowable losses [Copy link]

Regardless of whether or not you use a power IC, you must consider heat issues when using an IC, and never exceed the maximum rated Tjmax (maximum junction temperature/junction temperature), and perform heat dissipation design as appropriate. This is especially necessary for ICs or transistors that handle high power, such as power ICs. This section explains "allowable loss" following "How to interpret power IC specifications."

This is the last section of "How to interpret power supply IC specifications". "Specifications cover", "Block diagram", "Absolute maximum ratings and recommended operating conditions", "Key points of electrical characteristics", "How to read characteristics graphs and waveforms", "Application circuit examples", "Part selection", and "Input equivalent circuit" have already been explained. All of the items should be tips for interpreting specifications and making designs go smoothly, so it is recommended that you read them all.

Allowable loss

The allowable loss of an IC (applicable to most electronic components) refers to the maximum power consumption that does not exceed the temperature at which the IC can maintain its performance. For example, if the allowable loss of a power IC is "1.2W", then simply speaking, the power dissipated under this condition does not exceed 1.2W.

Taking a simple LDO regulator as an example, the calculation is as follows:

When the input is 5V, the output is 3.3V, and the input current (self-consumption + load) is 0.7A: (5V-3.3V)×0.7A=1.19W

Therefore, it can be judged as "barely safe". Of course, the actual situation needs to be further discussed.

The allowable loss is basically often displayed in a graph. The following is an example of a graph.

First, from the 2.66W curve in the graph, we can see that 2.66W is permissible up to 25°C, but the permissible loss decreases linearly above 25°C. The maximum operating temperature of this IC is 85°C, so 1.4W is permissible.

① to ④ indicate the conditions of the substrate and area. The allowable loss varies depending on the mounting conditions. In addition, the thermal resistance θj-a under these conditions is also displayed. In recent years, many power ICs are surface-mounted packages, and the allowable loss varies greatly depending on the mounting conditions, so this kind of detailed condition indication is useful.

If there are these conditions, the design itself only needs to use the allowable loss of the nearest conditions as a "reference". The reason why the word "reference" is framed is to emphasize that it is just a "reference" after all. This will be explained separately.

Let's take an example. If the substrate conditions in ② are quite close, and the upper limit of the Ta of the mounting substrate is 50℃, then it is known that a loss of less than 1.4W can be allowed. In the above LDO regulator example, about 0.8A can be obtained. Although the calculation is a bit troublesome for the switching regulator, if the efficiency (output power/input power) can be measured, 1-efficiency will become the loss. If it is a built-in power transistor type, this can be obtained, but for an external power transistor type, only the loss of the control IC must be calculated (measured). Of course, for the power supply circuit, thermal calculation of the external power transistor is necessary.

So, what do you notice after looking at this chart? Of course, this chart is worth a little deeper reading because it provides generally sufficient information.

(1) Why are the allowable losses the same between 0℃ and 25℃?
(2) At Ta = 150℃, the allowable loss is 0W.

Theoretically (1) is questionable. When the ambient temperature drops, the power loss should be increased in calculation. (2) You will immediately notice that Tjmax is 150°C. Since the power cannot be completely dissipated, the heat generation is zero, which is the condition of Ta = Tj.

In the past, I have been asked question (1). It is rare that I can read Tjmax from the graph and also have θj-a. So I will try to calculate it using condition (1) as an example.

From Tj＝θj-a×PD＋Ta, we get 47.0℃/W×2.66W（@ Ta＝25℃）＋25℃＝150℃ ←Tjmax condition

150℃＝47.0℃/W×?W（@Ta＝0℃）＋0℃, we get 3.19W←0℃, so we can allow for this loss in calculation.

In fact, although there are many different opinions on this, it is probably a common practice that if the room temperature is generally assumed to be 25℃, even if there is no power on, Tj is 25℃. If it is set to 2.66W or more, there is a danger of exceeding Tjmax at the moment of power on, so it is safer to apply the 25℃ value to the allowable loss below 25℃. Of course, in cold areas, it is not above 25℃, which is a very special condition and as long as there is reliable evidence, the allowable loss can be calculated accordingly.

Although this is a bit off track, what I want to say here is that if there is a graph that provides allowable loss, there is no problem in using it, but it must be confirmed according to the basic formula for calculating Tj, and the calculation results should be verified by actual measurement.

I have seen graphs where the values of the allowable loss graphs do not match the calculated results. I think that the derated values are used as the allowable values. On the contrary, if the allowable loss value provided is exactly Tjmax, using it under its loss conditions is equivalent to using it under the worst conditions in terms of reliability, and the service life will be shortened, so it is usually derated, that is, a margin is taken. Therefore, it is necessary to confirm the information provided to a certain extent.

Finally, let me explain why I framed the word "reference" above and emphasized it. In reality, the substrate conditions indicated in the technical specifications rarely match the actual design. Most of the time, the goal is to be "roughly close" or "better than this and worse than that". In addition, heat dissipation, thermal resistance, and Ta become quite uncertain factors, even due to the side-by-side arrangement of power devices or the relationship between cooling conditions (fans, etc.), frames, or setting locations. If the calculation is too complicated, it may not be possible to verify it. Therefore, it is very important to explore the information provided, confirm the calculation, and finally conduct actual measurements. In particular, with the high density of installation in recent years, it is sometimes impossible to use the heat dissipation space or heat sink as desired. When designing a power supply, I hope that everyone will fully understand that thermal discussion and thermal design are items that must be explored.

led2015

The allowable loss of a power supply IC is generally given in the data sheet. Usually, this parameter is given in the form of power or voltage.

cyberlogic

Learned