Superpower step response compared to other regulators
How does Superpower compare with other voltage regulators?
See the oscillograms below and judge for yourself. All devices
were tested using datasheet recommended application circuits, all
devices have a 100µF output capacitor and use the same input supply.
All were measured with 12V out except LT1086 (5V), LT1117 (5V),
LT1585 (3.3V). A 1A current step was the test, except for one regulator
that rated to 250mA.
Superpower 12V, 200mA step
Superpower 12V, 1A step
Superpower delivering 1A
into 0.1 Ohms, bottom trace is load voltage. The top trace shows Vout
transients, which are nearly invisible at 0.1V/div. Compare this oscillogram
with all those to the right and decide for yourself which you prefer.
Burson LNR 12V, 200mA step
This Burson device behaves more
like a resonant circuit than a regulator. Even though it's rated to
250mA, it does not stabilize to 200mA within 240msec, and takes over a full
second to settle back to 50mA. Notice the vertical axis for the top trace is 0.1V/div
vs. the other scope photos at 0.02V/div.
Dexa/NewClassD UWB 12V, 1A pulse
The NewClassD is a good regulator with clean, fast dynamics.
Compared to Superpower, notice the slow leading edge of load current.
LT3080 12V, 1A step
The LT3080 is a relatively new adjustable regulator.
It has very good transient response, almost as good as the LT1585. There is a slight
rise after the positive going edge that appears to be
a thermal recovery tail but that may also be due to the test circuit.
See a special note below on LT3080.
LT1086 5V, 1A step
The LT1086 has a
relatively quick transient at the leading edge of applied load and an
ugly settling transient on the trailing recovery edge.
LM1117T 5V, 1A step
The LM1117 shows dynamics
similar to the LT1086, even though an examination of their data sheets
shows a different error amplifier and output schematic. There is a slow
transient at the leading edge of applied load and an ugly settling
transient on the trailing recovery edge. This device also requires a
2.5V overhead for dropout.
LT1585 3.3V, 1A step
The LT1585 is billed as
"low dropout fast response" and it does have the best performance of
the monolithic devices. However, you can
see that the Superpower is faster with cleaner leading and trailing
edges. Note also that the maximum voltage for this regulator is 13V.
LM317 12V, 1A step
The LM317 is still a great design after 40+ years and
does well against most of the modern regulators. With a by–pass capacitor on the adjust pin
it performs well. This one has relatively slow edges on both falling and rising edges of the current step,
and comparatively slow recovery from 1A to 50mA load.
LM7812 12V, 1A step
The LM7812 remains a decent regulator. This one has relatively large overshoot
and slow recovery on both falling and rising edges of the current step.
Special note about the LT3080
Some time in 2012 we had a meeting with a Linear Technology Inc. saleman and application engineer about something unrelated to voltage regulators. They properly did their homework before the meeting by visiting our web site and mentioned at the outset a concern about us having a competing interest. After allaying their concerns the app engineer said something like "Your comparison pages have mostly older regulators, why don't you compare to a more recent regulator design?"
Of course he was fully justified with this remark and we decided to make a comparison to the LT3080, for which Bob Dobkin, Linear Technology's Chief Technology Officer and co-founder, received a Product of the Year award from Power Electronics magazine in 2007.
The LT3080 is a great monolithic regulator. In addition to its good electrical performance, it can be easily paralleled on a PCB to source higher currents. It was with some trepidation that we plugged it into our test sockets to compare with Superpower. Ultimately we were happy with the comparison results, and after the most recent improvement to lower noise, Superpower compares favorably well in this specification too.
After reading our patent, Mr. Dobkin emailed to say "I can see how it [the Belleson regulator design] will have excellent specs and is a neat way to make a regulator."