deadwood83
Member
- Joined
- Sep 1, 2022
- Messages
- 91
- Likes
- 160
I am not aware of anybody measuring Marantz' 2020-2021 slimline series receivers with the updated PCm5102A DACs after the AK fire. Supposedly the NR series does not have Marantz' infamous HDAM/JDAM/AMRAAM/NSAM/AMTRAK or whatever they are called.
I recently came into ownership of a Cosmos ADC and decided it was about time (and I am curious). First of all, I want to thank Ivan for bringing the Cosmos ADC product to market (and dealing with my questions) as well as pkane for bringing the multitone software out and making it quite accessible.
I wanted to start off with a baseline just to put into perspective my measurements vs Amir's. I have one device which we have both measured; the Loxjie D40 (DAC section). You can see his numbers here: https://www.audiosciencereview.com/forum/index.php?threads/loxjie-d40-review-dac-hp-amp.31040/. My measurements of the same device can be found at the end.
My NR1711 is basically used as an HDMI switch and I don't run speakers off it. I do not have the proper cables made at this time to use the 43V input of the Cosmos to measure the onboard amplifier. Previous versions of the Marantz series which had the AK DAC chip would use PCM5100 as Zone2. Models after May 2021 have the 8k resolution fix as well as the main AK DAC swapped for the PCM5102A.
My measurements will be with the Cosmos ADC set to 4.5V sensitivity fed by the Marantz pre-outs for L+R using an SE-BAL RCA-XLR cable for both channels. Output is slightly 'spicy' for an RCA signal at about 2.5Vpp.
Connections: Lenovo M75q Gen 2 -> HDMI -> Marantz RCA Pre-Out -> XLR -> Cosmos ADC -> USB -> Lenovo M75q Gen2
Marantz Settings: Speakers are set to 'Large', Subwoofer is disabled, amp assign is set to 'Bi-Amp Mains' ECO mode is set to Auto (reduces amp power when not needed, and since I have no speakers connected, it is very much not needed), Audyssey is disabled, graphic EQ is flat at 0 on all six bands, all level adjustments are set to +0.0dB.
Multitone settings: -4dB Play Gain. L+R channels on both output and input. 1048756 point FFT with 32 averages. All interfaces in WASAPI exclusive at 24/192kHz. Frequency range 20-20000. One peculiarity I noticed is that any 0dbFS signal fed to the receiver will result in an incredible amount of noise floor increase. This happened regardless of ADC sensitivity setting. I believe the AVR itself cannot handle a full 'hot' signal being fed into it via HDMI. All signals were set to -4dB in multitone to get a clean picture. All traces in BLUE are the left channel unless otherwise specified.
To probably nobody;s surprise, the Marantz is slightly noisy, but the numbers actually surprised me (in a good way... until later).
One of Marantz' marketing points is the lack of ground pin in the IEC socket. This is a Class 2 device. No ground pin, no ground noise, right? Well, we'll see about that. The mighty 60Hz is strong. Never underestimate the 60Hz.
NR1711 1kHz FFT:
Marantz is not doing the PCM5102A justice. Less than 16 bits usable range. For reference, here is the PCM5102A eval board tested by a gentleman on the TI E2E boards:
NR1711 Multitone:
There may be no ground plug, but that has never stopped the mighty 60Hz!
BUT! What if we feed an analog signal in from the D40 to the Marantz' CD input? Is it the amp circuitry, or the DAC implementation?
NR1711 Analog to Analog 1kHz FFT:
Oof.
NR1711 Analog to Analog Multitone:
Uhhhhhhhh. The analog inputs are truly awful This AVR should ONLY be used for digital input. I have no idea what the analog signal goes through but it is awful.
I spent way too much time measuring this sad device. Perhaps I will revisit it as a hacked unit with LVDS I2S output into a proper DAC. At least that way it won't feel like a huge waste of cash and effort.The digital section is passable
As a yardstick, I present my Loxjie D40 measurements so you can compare them with Amir's to come to your own conclusions. All my measurements show lower performance than Amir's by a few dBr.
D40 1kHz fft:
D40 Multitone (32-tone AP version):
On the D40 I got a TD+N of -116.4dB which is roughly equal to 116.4dB SINAD. Amir measured 122dB SINAD, most likely due to a much more controlled environment and better gear. If we look at my 116 vs his 122 there is a delta of ~6dB SINAD. I have no idea if this is linear/applicable when examining across devices. If it is applicable, then the 84 SINAD of the AVR can be brought down to 90 SINAD which would place it near the AV8805 and AV7705.
Edits: spelling, Zone 2 DAC correction.
I recently came into ownership of a Cosmos ADC and decided it was about time (and I am curious). First of all, I want to thank Ivan for bringing the Cosmos ADC product to market (and dealing with my questions) as well as pkane for bringing the multitone software out and making it quite accessible.
I wanted to start off with a baseline just to put into perspective my measurements vs Amir's. I have one device which we have both measured; the Loxjie D40 (DAC section). You can see his numbers here: https://www.audiosciencereview.com/forum/index.php?threads/loxjie-d40-review-dac-hp-amp.31040/. My measurements of the same device can be found at the end.
My NR1711 is basically used as an HDMI switch and I don't run speakers off it. I do not have the proper cables made at this time to use the 43V input of the Cosmos to measure the onboard amplifier. Previous versions of the Marantz series which had the AK DAC chip would use PCM5100 as Zone2. Models after May 2021 have the 8k resolution fix as well as the main AK DAC swapped for the PCM5102A.
My measurements will be with the Cosmos ADC set to 4.5V sensitivity fed by the Marantz pre-outs for L+R using an SE-BAL RCA-XLR cable for both channels. Output is slightly 'spicy' for an RCA signal at about 2.5Vpp.
Connections: Lenovo M75q Gen 2 -> HDMI -> Marantz RCA Pre-Out -> XLR -> Cosmos ADC -> USB -> Lenovo M75q Gen2
Marantz Settings: Speakers are set to 'Large', Subwoofer is disabled, amp assign is set to 'Bi-Amp Mains' ECO mode is set to Auto (reduces amp power when not needed, and since I have no speakers connected, it is very much not needed), Audyssey is disabled, graphic EQ is flat at 0 on all six bands, all level adjustments are set to +0.0dB.
Multitone settings: -4dB Play Gain. L+R channels on both output and input. 1048756 point FFT with 32 averages. All interfaces in WASAPI exclusive at 24/192kHz. Frequency range 20-20000. One peculiarity I noticed is that any 0dbFS signal fed to the receiver will result in an incredible amount of noise floor increase. This happened regardless of ADC sensitivity setting. I believe the AVR itself cannot handle a full 'hot' signal being fed into it via HDMI. All signals were set to -4dB in multitone to get a clean picture. All traces in BLUE are the left channel unless otherwise specified.
To probably nobody;s surprise, the Marantz is slightly noisy, but the numbers actually surprised me (in a good way... until later).
One of Marantz' marketing points is the lack of ground pin in the IEC socket. This is a Class 2 device. No ground pin, no ground noise, right? Well, we'll see about that. The mighty 60Hz is strong. Never underestimate the 60Hz.
NR1711 1kHz FFT:
Marantz is not doing the PCM5102A justice. Less than 16 bits usable range. For reference, here is the PCM5102A eval board tested by a gentleman on the TI E2E boards:
NR1711 Multitone:
There may be no ground plug, but that has never stopped the mighty 60Hz!
BUT! What if we feed an analog signal in from the D40 to the Marantz' CD input? Is it the amp circuitry, or the DAC implementation?
NR1711 Analog to Analog 1kHz FFT:
Oof.
NR1711 Analog to Analog Multitone:
Uhhhhhhhh. The analog inputs are truly awful This AVR should ONLY be used for digital input. I have no idea what the analog signal goes through but it is awful.
I spent way too much time measuring this sad device. Perhaps I will revisit it as a hacked unit with LVDS I2S output into a proper DAC. At least that way it won't feel like a huge waste of cash and effort.The digital section is passable
As a yardstick, I present my Loxjie D40 measurements so you can compare them with Amir's to come to your own conclusions. All my measurements show lower performance than Amir's by a few dBr.
D40 1kHz fft:
D40 Multitone (32-tone AP version):
On the D40 I got a TD+N of -116.4dB which is roughly equal to 116.4dB SINAD. Amir measured 122dB SINAD, most likely due to a much more controlled environment and better gear. If we look at my 116 vs his 122 there is a delta of ~6dB SINAD. I have no idea if this is linear/applicable when examining across devices. If it is applicable, then the 84 SINAD of the AVR can be brought down to 90 SINAD which would place it near the AV8805 and AV7705.
Edits: spelling, Zone 2 DAC correction.
Last edited: