For space-constrained, battery-powered products, the humble LDO regulator is a
simple and reliable solution to a challenging engineering problem.
By Steven Keeping
Climbers have several ways of reaching the peak of Snowdon, which at 1,085 metres
is the highest point in Wales (UK) and offers views of five countries, 24 counties, 29 lakes and 17 islands. The hardy can choose several walking routes to the summit of which
the easiest, but also the longest, is the 7km Llanberis Path. The hardier still can traverse the Llanberis Path as part of the annual Snowdon Race. (You’ll need to post a timing that’s
less than 40 minutes if you hope to win the thing.) Unsurprisingly, most day-trippers head for the Snowdon Mountain Railway, a 7.6km, singletrack, 800mm gauge railway running alongside the Llanberis Path.
Running a train up an average gradient of 12.7 per cent with sections as steep as 18.2 per cent is a formidable mechanical engineering challenge. Conventional trains struggle to cope
with gradients over 2 per cent due to the lack of adhesion between the driving wheels and the rails. One popular solution is to limit the track gradient by zigzagging the line up the mountain. The downsides are dramatically increased civil engineering costs and a much longer journey for passengers.
With no precedent for such a steep railway in the United Kingdom, the Welsh engineers turned to the Swiss, the world experts in getting trains up mountains. The Swiss solution is as ingenious as it is simple; the train is pulled up the slope by attaching a cogwheel to the driven axle that engages with a toothed rack rail set between the running rails. Such a rack and pinion arrangement ensures that the driven wheels can’t slip, no matter how steep the gradient. It’s a cheap, effective and low-maintenance solution that’s still working well over 120 years since it was first installed.
Beware the trade-offs
Electronics engineers face technical challenges as tough as those overcome by the Snowdon railway builders. And the temptation is to look to the latest technology for the answer. One such example is of building a power converter for a battery-powered device such as an IoT sensor or a wearable. A major consideration for the designer is to select a power converter that extends battery life so that maintenance technicians don’t have to swap out sensor cells too often or consumers aren’t inconvenienced by frequent smartwatch recharges.
Designers need a power converter to generate the required output voltage and current for a product’s electronics from a given input power source. And this needs to happen during both steady-state and transient conditions. In an IoT sensor or wearable, the input power source is the battery, and the regulator needs to maintain the constant voltage demanded by the product’s electronics even as the cell discharges and its voltage tails off.
An obvious answer to this problem is a switching voltage converter, a technology that first appeared in the 1960s but continues to evolve today. When a switching voltage converter’s internal transistor is ‘on’ and conducting current, the voltage drop across its power path is minimal. When the transistor is ‘off’ and blocking high voltage, there is almost no current through its power path. Consequently, the transistor performs like an ideal switch and dissipates very little power, boosting the efficiency of a switching regulator to over 90 per
cent in some applications.
But switching regulators, like zigzagging railway tracks, comes with trade-offs. They are complex, requiring external feedback loops; they take up a lot of space because they require peripheral components such as energy storing inductors, and capacitors and resistors for filter circuits; they are expensive; and the transistor switching can generate electrical noise that can disturb the sensitive downstream silicon.
Keep it simple
An alternative is the low-dropout (LDO) linear regulator, a device that traces its history back to an article written by Robert Dobkin (then an engineer at National Semiconductor, and later founder and CTO of Linear Technology) published in Electronic Design back in 1977. One key difference between the switching regulator and the linear regulator can be deduced from their respective names; where the former employs a switching transistor at its heart, the latter uses a transistor operating in its linear range.
In the linear regulator, the transistor works as a variable resistor operating in series with the output load. The regulator employs an integrated feedback loop using an error amplifier to sense the output voltage via a sampling resistor network, which is then compared with a reference voltage.
LDO regulators work in the same way as conventional linear voltage regulators, but feature an important tweak in their topology. Unlike conventional linear regulators, LDO regulators use an opencollector or drain topology, enabling the minimum voltage drop from the input to output voltage (while still ensuring proper operation), which is as low as the saturation voltage across the transistor—plus a small safety overhead.
As efficiency is high on the list of attributes for a power converter in a battery-powered device, care must be taken to make the most of this ‘lowdropout voltage’. An LDO regulator
always burns some power in order to regulate the output voltage and, because the device is essentially a variable resistor, its power dissipation is equal to the voltage difference across the device times its output current. Its efficiency is the ratio of output voltage to input voltage. Part of the engineer’s job then becomes to match battery input voltage as closely as possible to the desired output to make the most of the power budget. For example, for a given current and a battery input voltage of 3 volts, an LDO regulator supplying 2.5 volts will be about 16 per cent more efficient than one supplying 2 volts. Because of its topology, the difference between the input and output for an LDO will be lower than that of a linear regulator. If the engineer selects carefully, an LDO power converter can approach an efficiency of 90 per cent, getting close to that of a switching regulator but without the complexity, expense and electrical noise challenges of that device.
LDO regulators are less suitable for high current applications because the power dissipation climbs with the current for a given voltage drop. And the devices are unable to boost voltages. It is for these applications that the switching supply truly comes into its own. But for space-constrained, battery-powered products, the humble LDO, like the cog railway, is a simple and reliable solution to a challenging engineering problem.
The author is a contributing writer for Mouser Electronics. He is based in Sydney, Australia.