When You Really Need Low-Dropout Regulation: A New LDO from Diodes Incorporated
By Robert Keim, Director of Engineering at All About Circuits
Linear regulators such as the AP7380 series can offer convenience, simplicity, and flexibility in terms of input voltage. Nowadays we need to make a distinction between high-dropout low-dropout regulators and low-dropout low-dropout regulators. This curious situation results from the fact that “low-dropout regulator” (abbreviated LDO) has become essentially synonymous with “linear regulator”. In my experience, it is perfectly acceptable to refer to any linear regulator as an LDO, regardless of whether we have so much as glanced at the dropout-voltage spec. Why exactly this occurred I cannot say—I suppose partially because “LDO” is easier to say than “linear regulator”.
Both ground current and load current contribute to LDO power dissipation. You can find more information here.
This little bit of terminological imprecision is nothing to stress about, but it has evolved into a slight problem with respect to the actual dropout performance of linear regulators. I assume that at some point all the newer regulators were rather low-dropout relative to older devices, and thus the adoption of “LDO” as a generic term seemed reasonable. But now we’re in a situation where we need to make a distinction between regulators that are low-dropout by older standards and regulators that are low-dropout by current standards.
Apparently, someone at Linear Technology has at least indirectly recognized this issue; on page 4 of this app note you’ll find the abbreviation “VLDO”, i.e., very-low-dropout regulator. So if we adopt this scheme, a normal linear regulator is an LDO, and a low-dropout linear regulator is a VLDO. (Am I the only one who finds this whole thing amusing?)
The Efficiency Question
By the way, while we’re on the topic of LDOs, I have expressed elsewhere my general lack of interest in switching regulators. For me the improved efficiency does not compensate for the noisy output voltage and the complicated circuit design. Obviously I use them when I have to, but not without bitterness. Anyways, since this article focuses on a new linear regulator, it seems to me a good opportunity to mention the following excellent point presented in the same Linear Tech app note: if the difference between your input voltage and output voltage is small, a linear regulator can actually be more efficient than a switcher.
However, in addition to input-to-output differential, we have to consider load current. If you’re familiar with efficiency plots for switching regulators, you know that the situation deteriorates rapidly as load current decreases past a certain point. This is evident in the following plot for a series of DC/DC converter modules discussed in this article:
Image courtesy of Murata.
So switcher efficiency can be highly dependent on load current, whereas with a linear regulator it depends only on the input voltage and the output voltage:
So if the input-to-output differential is small and you are expecting extended periods of operation at low load current, you might be better off with a linear regulator—even in terms of efficiency.
The question that confronts us now is, How low does the dropout voltage need to be in order for a device to be identified as a VLDO? A quick Digi-Key search indicates that there are still numerous linear regulators with dropout voltage greater than or equal to 1 V, so I propose that as the threshold. I can’t back this up with data, but I have a vague impression that there is something impressive about a device that can maintain solid voltage regulation with less than 1 V between input and output.
The AP7380 from Diodes Inc. qualifies as a VLDO under most conditions. The details are shown in the following excerpt from the datasheet, which also serves as a reminder that you need to carefully check the dropout specs before designing an entire PCB around your linear regulator.
As you can see, changes in output current and output voltage lead to nontrivial differences in dropout performance—from a minimum of 250 mV (typical) to a worst-case spec of 1500 mV. The following plot gives you a clearer idea of the relationship between output current and dropout voltage, and it also reminds us of the ubiquitous (though easily forgotten) effects of temperature:
Plot taken from the AP7380 datasheet.
Another AP7380 feature that Diodes Inc. wants to emphasize is the wide input voltage range—from 3.5 V to 24 V. This is undoubtedly convenient, because one device can be used in multiple power-supply situations, and also because the device will be more resistant to supply transients. Of course, you still have to worry about the increased power dissipation that occurs at higher input-to-output differentials.