For this next set of measurements I’ll be looking at another amplifier designed by Uffe at Slowdiyer. There’s no particular reason that I chose another of these designs straight after looking at his recent offering beyond the fact that it’s conveniently available on the rack next to my computer. In addition, it gives me a chance to actually listen to this again, as it’s been a while. The full build article is here and power measurements can be found on this page. The original article, including Gerbers, can be found on the Slowdiyer website.
I mentioned previously the particular setup of the measurement chain can impact on the measurements and this proved to be the case here. As I had good results using the internal generator of the Focusrite Scarlett 2i2 last time, I repeated that setup, but there appeared to be quite a significant amount of AC noise present. I switched to the Akitika external oscillator and achieved better results so the images below are from that series of measurements.
The measurement chain for these readings is:
Akitika 1kHz 2ppm external oscillator>1m RCA cable>OPABUF HPA>1m TRS cable>25/330Ω dummy load>30cm RCA/TRS cable>Scarlett 2i2
The dummy load is temporarily connected to a Siglent SDS 1202X-E oscilloscope to set the voltage. This is disconnected for the distortion measurements. The Akitika is set at maximum output (1.5V RMS according to the spec sheet) and all voltages on the graphs are V RMS. The desktop PC is running REW 5.20.4 on Windows 11 and I’m using ASIO at 48kHz sample rate. It should be remembered that it is the amp and my choice of power supply that is tested here (in this case dual filtered Meanwell IRM 20-12).
The noise and distortion graphs are included below, click on an image to get the full size, and the main figures are tabulated for quick reference. A brief explanation about the numbers and what they mean for audibility can be found in a thread at ASR, and the relevant information is reproduced in the extract below. This distils the guidance from NWAVGuy from an article written originally for Inner Fidelity (but now sadly lost in the ether of the stereophile website).
In this article, NwAvGuy provides some guidelines for good amplifier design. He is “an electrical engineer by education (BSEE) and career”, not a psychoacoustician, and he does not cite his sources, so his limits will be treated as lenient.
NwAvGuy says that “noise needs to be -85dB below the signal to be inaudible which works out to only 0.005% THD+N. But music masks distortion so 0.01% (-80dB) is considered acceptable.” Since I make no assumptions about the listener or listening material, I’ll disregard the -80 dB threshold and go with -85 dB.
His guideline for distortion is <0.05% equivalent to -66 dBFS.
FlipFlop (2018) Audibility thresholds of amp and DAC measurements
| Load Ω | V RMS | THD dB | THD+N dB |
|---|---|---|---|
| 25 | 0.5 | -109 | -82 |
| 25 | 1.0 | -116 | -88 |
| 25 | 1.5 | -119 | -88 |
| 25 | 2.0 | -122 | -85 |
| 330 | 0.5 | -110 | -82 |
| 330 | 1.0 | -116 | -89 |
| 330 | 1.5 | -119 | -92 |
| 330 | 2.0 | -101 | -94 |
| 330 | 2.5 | -99 | -94 |
A note of caution with some REW measurements is needed here. There appears to be a glitch sometimes with REW whereby the measurements in the distortion panel give erroneous results such that the panel figures fail to take into account the level of the fundamental. In essence, the harmonic measurements will read lower than they should. This is resolved by shutting down REW and restarting.
As I’m typing this up, I’m listening on my HD600 to Anja Lechner (FLAC 16/44.1) from my phone into this amp. There’s plenty of power for these headphones and to my ears this amp does what it’s supposed to – deliver a nice, clear sound.