How does Belleson Superpower compare with a Dexa NewClassD regulator? We compare to the NewClassD voltage regulator shown here. The NewClassD is a production unit purchased from Parts Connexion, which also sells Belleson regulators. Both devices were tested using identical conditions, both using the same input supply and Vout = 9V for both regulators.

Dynamic performance

First, let us say that the dynamic performance of the NewClassD regulator is excellent, on par with Superpower and better than other regulators we've tested. That said, we move on to input ripple rejection, output impedance and noise.

Ripple Rejection

The Dexa/NCD regulator has, at power line frequencies of 50Hz or 60Hz, decent rejection of input ripple. However performance falls quickly as ripple frequency increases, where at 5KHz input ripple, rejection is down about 30dB from its maximum of -109dB at 50Hz. Superpower is 7dB better at 50Hz and 15dB better at 5KHz. Recall that in voltage measurements (not power), 6dB is a factor of 2 and 12dB is a factor of 4. Superpower measures more than 4 times better than the NewClassD regulator at 5KHz input ripple. Even the maligned LM7805 offers better ripple rejection at 5KHz.

Ripple rejection (PSRR) of 4 regulators, lower values are better. The graph shows the magnitude of Vout(AC) given a Vin of 12V(DC) + 0.5Vpp(AC). Vout(DC)=9V for all regulators except the LM7805, which is of course 5V out. Regulators were all equally loaded with a 220Ω resistor (54mA DC load).

SPJ78 PSRR at 5KHz

Belleson SPJ78 regulator PSRR AC output with ½Vac(pp) input ripple at 5KHz riding on 12Vdc. Peak of fundamental is -95dB, over 5½ times better than the NewClassD's measurement of -80dB. Click the graph for a clear full size spectrum image.

---

NewClassD PSRR at 5KHz

NewClassD regulator PSRR AC output with ½Vac(pp) input ripple at 5KHz riding on 12Vdc. Peak of fundamental is barely -80dB, with a 2nd harmonic as high as Superpower's fundamental. Click the graph for a clear full size spectrum image.

---