My HPA-3B was purchased used and already modded by the previous owner. Before I received it, it was updated with the Sparkos 15v regulators (lowering rail supply from 18v). The opamps were replaced with Burson v5i (list $69 for two), so I have not heard or measured the stock 3B.
Based on a recommendation from @trl I decided to try TI's OPA1652 opamps. This took some doing due to the necessary conversion from the tiny form factor to the 8 pin DIP socket. The good thing is that 1652 is cheap as chips (!) at about $2 each @ quantity 1. Armed with a rework station, soldering paste and a magnifying glass, I spent a few hours making the proper connectors and installing the chips ...only destroying two of them in the process. One by using too much paste and causing all pins to be fused together, the other just didn't work for some reason although everything looked right.
HD650 and HE560 headphones using balanced cables sounded great. Going back to Burson, everything still sounded excellent. I couldn't tell the difference, although OPA1652 do generate a little more heat. Next up, did some quick measurements using Apogee Element24 as a DAC and ADC. Turns out, there was really not much of a measurable difference between these two opamps in the 3B circuit.
Burson v5i, 1kHz:
TI OPA1652, 1kHz. Very slightly higher third harmonic, but for some reason lower 60Hz mains frequency. Probably unrelated to opamp swapping:
Multitone test is also almost exactly the same for both opamps, other than the 60Hz component (blue is the Burson v5i):
Opamp rolling didn't seem to have any effect on the frequency response:
In conclusion, the much less expensive OPA1652 appears to be every bit as good as the Burson v5i when used in Matrix 3B, with no obvious audible or measurable differences. YMMV
Based on a recommendation from @trl I decided to try TI's OPA1652 opamps. This took some doing due to the necessary conversion from the tiny form factor to the 8 pin DIP socket. The good thing is that 1652 is cheap as chips (!) at about $2 each @ quantity 1. Armed with a rework station, soldering paste and a magnifying glass, I spent a few hours making the proper connectors and installing the chips ...only destroying two of them in the process. One by using too much paste and causing all pins to be fused together, the other just didn't work for some reason although everything looked right.
HD650 and HE560 headphones using balanced cables sounded great. Going back to Burson, everything still sounded excellent. I couldn't tell the difference, although OPA1652 do generate a little more heat. Next up, did some quick measurements using Apogee Element24 as a DAC and ADC. Turns out, there was really not much of a measurable difference between these two opamps in the 3B circuit.
Burson v5i, 1kHz:
TI OPA1652, 1kHz. Very slightly higher third harmonic, but for some reason lower 60Hz mains frequency. Probably unrelated to opamp swapping:
Multitone test is also almost exactly the same for both opamps, other than the 60Hz component (blue is the Burson v5i):
Opamp rolling didn't seem to have any effect on the frequency response:
In conclusion, the much less expensive OPA1652 appears to be every bit as good as the Burson v5i when used in Matrix 3B, with no obvious audible or measurable differences. YMMV
Last edited: