Class A Aleph 3 clone Measurement

[dumpingg]

My test setup which is based on a Windows 10 laptop, REW software, Focusrite Solo 3rd Gen and DIY attenuation/connection box is described here:

https://www.audiosciencereview.com/...ement-setup-panasonic-xr57-avr-example.10589/

I recently purchased a China built Aleph 3 Class A power amplifier clone. It is a clone and it is based on a Class A amp that I knew in advance does not measure well and is definately not going to win any low noise/distortion awards. But, just for the heck of it, I wanted to try out the Class A amp with my bi-amped Magnepan 1.6 speakers. I am using the Class A amp on the tweeters and a second different Class-D based amp on the woofers. The Pass approach to amplifier design and management of distortion products (especially 2nd order) in the spectrum is described in a whitepaper about distortion in amplifiers here:
http://www.firstwatt.com/pdf/art_dist_fdbk.pdf.

Here are some photos of the Clone unit:

Looking at the internal photo you can see that the signal cables are routed close to the power wires and transformer. I did try rerouting the signal wires but that did not seem to improve the my measured results.

Anyway, I must say that despite the fact that this amp does not measure well, it really sounds good on the tweeters of my magnepans. Especially on live vocal and jazz type music giving a live air type feeling to the music.

The real non-clone version of the amplifier was previously measured by stereophile here:
https://www.stereophile.com/content/pass-labs-aleph-3-power-amplifier-measurements
From my measurements, it appears that the clone is similar in performance (specifically figure 7 in the referenced link above but the y axis scales/ranges are different).

Here are my measurements: ***(UPDATED MEASUREMENTS later in this thread)

 

[dumpingg]

Well I also have a Behringer UCA202 USB sound device. the Focusrite is a balanced sound device that I use RCA to 1/4in adapters on. Measuring with the UCA gives better results. Looks like I have some setup debugging to do.

 

[dumpingg]

Attached is an updated 19k/20k dual tone -4dBFS signal which was the maximum REW would allow me to set in the signal generator utility. I used a TOSLINK connected DAC as the source to try and remove some of the apparent ground issues. This is also with the input sensitivity raised to the max (@-5.49dBFS) on my soundcard before clipping made the measurements wacky. I think these measurements compare favorably to what was posted on the stereophile review linked above. As this is a dual tone signal, my voltmeter trueRMS did not sync to it and I don't know the exact wattage but I had the amp set at max volume.

 

[dumpingg]

And just for completeness, attached are additional measurements taken with my Focusrite Solo but driven with the TOSLINK optical attached DAC generating the 1khz tone. While these measurements look worse than the 1khz one taken above with the UCA202, take a close look at the y axis and noise floor. The Focusrite is digging deeper into the noise floor to show lower level details. Also the Focusrite sound card allows me to adjust the input sensativity to pretty close to 0dBFS before clipping. Unlike the UCA202 that maxes out at -4dBFS.

1khz, 5watts at 4ohms

1khz, 5watts at 8ohms

 

[pma]

Good, but please do not forget that you are still tuning measuring method errors rather than measuring amplifiers.

 

[dumpingg]

Good, but please do not forget that you are still tuning measuring method errors rather than measuring amplifiers.

Could you clarify? One of the reasons I started with measuring my Panasonic XR-57 (posted here https://www.audiosciencereview.com/...ement-setup-panasonic-xr57-avr-example.10589/) was so that I would have a known amirm measurement case to compare my technique against.