Superpower Ripple Rejection
compared to newer regulators
How does Superpower compare with the latest generation of high performance voltage regulators? See the FFT spectra below and judge for yourself. Devices were tested using datasheet recommended application circuits, all devices use the same input supply. All were measured with 12V out with no load. A perfect spectrum would show only a noise floor—an ideal regulator never varies from its specified DC output voltage and ripple that transits from input to output appears as a vertical spike out of the noise floor.
This is a somewhat dreary page that shows many hours of measurement taking,
and if you don't want to spend hours reviewing what it says, the data are summarized
in our Ripple Rejection discussion page
and in this summary graph.
120dBV represents 1µV and 100dBV is 10µV of ripple at the output.
Some of the individual spectra for higher frequency measurements have aliased artifacts that are "extra" to the PSRR measurement. The graphs have been modified to change the color of those elements to gray so you can focus on the peak of the ripple (such that it is). Also, some of the really great regulators have such low ripple that it is hidden by regulator noise at some frequencies.
Power Supply Ripple Rejection
PSRR for good regulators is typically very high at low frequency, so to get an accurate measurement we put the regulator output through a gain of 1000 amplifier. Ripple frequency is chosen as 55Hz to make one measurement that's useful for most of the world's power line frequencies. The amplifier has a high-pass corner of 20Hz so it doesn't attenuate at 55Hz.
The choice of 55Hz also allows the ripple rejection to be visible away from the actual power line frequency of 60Hz in our lab. You can see some power line ripple that is either ground feedthrough in the test fixture or introduced in the 1000x amplifier. This can be ignored and you should focus on the voltage peak just to its left.
When comparing, be careful to notice not just the size of the peak but also its maximum on the vertical axis, because its height depends on the noise floor of the regulator. For example, the LT3045 and SPX both show -135dBV but the noise floor of the LT3045 is so low that its ripple appears to be more but it's not.
Test conditions
First, please note that we are measuring a few hundred nanovolts and this is difficult. We solved multiple problems with spurious signals and ground feed-through coming from several sources, including 60Hz added by a nearby oscilloscope, a 185Hz spike from an attached voltmeter (I have to assume that's the frequency of its internal integrator), and ground noise at the measurement frequency caused by inadequate internal grounding in several signal generators. We finally were able to use an old Wavetek Model 166 and a Tiepie HS3 that have less ground signal injection than two other generators on our bench (including the venerable HP model 200CD).
- Vin = 16VDC/sub> + 0.5VACpk
- Sample frequency = 2*max frequency
- 16 bit resolution
- 2k samples per sweep
- 16 sweeps averaged
100Hz Ripple Rejection
New Belleson SPX78
Older Belleson SPZ78
Texas Instruments TPS7A4700
1kHz Ripple Rejection
New Belleson SPX78
Older Belleson SPZ78
Texas Instruments TPS7A4700
10kHz Ripple Rejection
New Belleson SPX78
Older Belleson SPZ78
Texas Instruments TPS7A4700
50kHz Ripple Rejection
80kHz Ripple Rejection
Notes
Measurements are taken in the same test socket, with the same input stimulus and output sense for all devices. Measurements may differ from those you see in manufacturers' data sheets because of different setup, e.g. input or output capacitance, placement of sense device, wire lengths, etc.
The LT3045 and TPS7A4700 are both surface mount monolithic devices that require a PCB to allow them to be plugged into a TO-220 style test socket. The tested devices were, when purchased, already mounted on a PCB with MLCC capacitors connected. Replacing the MLCCs with tantalum on the TPS7A4700 PCB significantly improved performance, and the measurements you see here are with 10µF tantalum.