Any class D amps tested with RMAA yet?

[gustep12]

Hello,

Yesterday I started playing a bit with the newest version of the RMAA software on my PC. I found that my Blugears B-enspirer sound card has roughly 95dB spurious free dynamic range and a ca. -120dB noise floor when I loop the front out to the line in with an analog cable.

Sounds good so far....

I then decided to include my class D receiver, the JVC RX-D301, in the loop. There are five ways of doing this:

1.) b-Enspirer (via S/PDIF PCM) to RX-D301 (back via headphone out)

2.) b-Enspirer (via S/PDIF DTS) to RX-D301 (back via headphone out)

3.) b-Enspirer (via S/PDIF DDL) to RX-D301 (back via headphone out)

4.) b-Enspirer (via USB) to RX-D301 (back via headphone out)

5.) b-Enspirer (via analog out) to RX-D301 (back via headphone out)



As you see, the first four choices are digital feeds for the amp, and the last is analog. Of the five possible connections, I tested 1, 2, 3, and 5 so far.

The first observation was that of the two means to transfer 5.1 channels digitally - namely Dolby Digital Live and DTS - the DTS signal was better than DDL, with better flatness and less drop-off in the high frequencies. But everything else, including analog, was pretty similar.

Also, both DTS and DDL had poor stereo separation compared to everything else - I'm not sure if this is by design or because of some surround effect that I overlooked and didn't disable.

Sadly, all routes resulted in a pretty poor -70dB spurious-free dynamic range. Mostly, this was due to a 60Hz hum and its harmonics.

I like this amplifier because it gives me many inputs and 7 discrete amplification paths, one of which I re-routed to power my passive subwoofer.

But still, I am wondering if someone else has tested their class D and has obtained better values here?

Let me know if there's a good way to share the RMAA report online, I'd be happy to post the results.

[TheMG]

Your test may be inconclusive. The receiver may have a seperate amplifier for the headphone output. Some receivers use a resistor divider on the main speaker output for the headphone output, but some will have a completely seperate low-power amplifier for the headphones. If this is the case, then your test did not include the class D amps.

[gustep12]

I agree, it's most likely not including the class D amps - to test that, I think I would need to wire a transformer to the speaker output to get a line level signal from the low voltage, very low impedance output - or is there a better way?

I think the headphone out would be the best possible output, though. It's unlikely (but certainly possible) that the headphone amp should be worse than the power stage. Also, this is only meant to be a starting point, I am basically interested to see how much better the headphone out on other comparable systems can be. Also, if we agree on a simple testing method for the speaker out, I'll be happy to repeat the measurement.


Simple testing is better than no testing. It's a starting point.


I'll try to post the results below.

Test conditions: 16-bit, 48kHz DAC / ADC.
Sound card: Bluegears b-Enspirer (CMI 8788)
Receiver: JVC RX-D301

Analog: Analog to receiver, analog (headphone) back to PC
PCM: 48kHz PCM to receiver, analog (headphone) back to PC
DDL: Dolby Digital Live to receiver, analog (headphone) back to PC
DTS: DTS Connect to receiver, analog (headphone) back to PC

Frequency Response

[gustep12]

Noise Level

[gustep12]

Dynamic range

[gustep12]

THD + Noise

[gustep12]

Intermodulation distortion

[gustep12]

Stereo crosstalk

[gustep12]

By the way, I just found my past writeup of how I modded the class D amp to provide an output for my passive subwoofer (I did this because I already had a passive, and because I prefer to not have yet another amp involved - affordable active subwoofers have very bad amps, typically).

Main thing is, my writeup provides some details (block diagram) into the architecture of this amp, and maybe it can point towards a good method of testing it with RMAA.

Find it archived here:

http://archive2.avsforum.com/avs-vb/...d.php?t=736254

[Tekko]

I always use rmaa to test my diy class d amps.