so it's the pre-amp that's clipping?
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Hegel H95 Review (Streaming Amplifier)
- Thread starter amirm
- Start date
- Joined
- Jan 20, 2020
- Messages
- 263
- Likes
- 391
It's damage control: we agree that manual may be misleading, even the label LINE OUT may be uninspired but the product... the product is perfect. An errata to the manual is so cheap.
Agreed, Hegel is firmly in the old-school HIFI mindset, which sites such as this are very effective to expose. Lazy engineering is a thing of the past.
Schiit is cleaning up their act, Hegel needs to do the same. In the meantime, we should take our business elsewhere.
I see no proof that the dac was clipping in coaxial tests even if volume was set at 95. In fact, the 72 SINAD result is expected of 80db DAC into a 82db amplifier, which are best case scenarios for DAC and AMP and indicative of real world performance.
> When the power amplifier starts clipping, this will reflect in the RCA pre out as well
Which is brings me back to the block diagram, how? Dodgy power circuitry?
I missed that, thank you for clairifying. It's quite the big deal. Did @amirm explain why he did that?
Actually, that is an interesting question
Given the fact that there is no global feedback and there are multiple chained gain stages all with local feedback, the question then is why clipping at the end makes the start also distorted? Possibly one of the electronics guru's can shed some more detailed light on this?
DSJR
Major Contributor
A pal suggested a Quad Vena 2 amp as a better priced alternative (also to the Topping PA5/E50 combination looking at facilities offered if not chasing SINAD). I gather the Leak also-retro unit is similar inside...
Basic measurements here and again, 500mV for full output
Basic measurements here and again, 500mV for full output
Last edited:
solderdude
Grand Contributor
It wasn't clipping in that test nor did I say it was.
The 72dB SINAD of the amp was caused by the stimulus being -24dB so the noise floor of the DAC (which arguably is already high) is amplified causing the noise level to go up in ways that in normal usage would not occur unless the recording was digitally attenuated by 24dB or recorded with peaks reaching -24dB.
The plot is not a real world situation for 5W (in 'normal usage') but the result of the standardized test method used by Amir (which makes sense) but in that particular plot made it look worse than it is when you were using it 'normal' and SINAD would be amp determined in real world situations.
It basically shows the higher noise floor of the DAC (24dB increased) compared to the line-in a bit higher.
KSTR
Major Contributor
OK, one last try.
Soundstage's measurements:
The important info is this section from the Soundstage report:
"We attempted to optimize the volume position to achieve the best signal-to-noise ratio (SNR) and THD+N measurements at the speaker outputs (8 ohms) for both analog and digital inputs. For all analog input primary measurements (unless otherwise specified), the best measurements were achieved with the volume was set to 99. At a volume setting of 99, 240mVrms was required at the input for 10W (8 ohms) at the output. For the digital input, we found a volume setting of 59 yielded 10W (8 ohms) at the output for a 0dBFS input. We found essentially no differences between the optical and coaxial S/PDIF inputs, as well as the USB input, in terms of THD+N."
Line Input:
"Shown above is an FFT for a 1kHz input sinewave stimulus at the analog input, measured at the output across an 8-ohm load at 10W. Here we see that the second harmonic, at 2kHz, is just below -90dBrA, or 0.003%, while the third harmonic, at 3kHz, is just above -90dBrA. The other harmonics measured lower, below -100dBrA, or ?0.001%, down to about -125dBRa, or 0.00006%, at 20kHz. Below 1kHz, we see noise artifacts, with the 60Hz peak due to power-supply noise reaching -100dBrA (left), or 0.001%, and -110dBrA (right), or 0.0003%, and the 120Hz peak reaching just below -90dBrA. The harmonics of these two peaks can also be seen at lower levels."
DAC Input:
"Shown above are FFTs for a 1kHz 0dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load at 10W."
(Bold mine in all cases).
What we can infer from these plots that the DAC path is NOT significantly nosier (in terms of wideband noise) than than the line input and the distortion is very similar which tells us the DAC intrinsic distortion is NOT dominating either, for that very reasonable test conditions (0dBFS or equivalent analog input voltage, volume dialed in to achieve 10W/8R output and mentioned different pot setting for the analog input).
The difference in mains hum indicates PCB layout / cabling problems for the DAC lines (which we already noted). In general the measured performance sure is not spectacular, middle of the road at best.
Now, what we wanted to know about suitability for the occasional DAC- or PreAmp-only use-case is the behavior of DAC vs line input at the line output, which neither SS nor Amir measured. I'd volunteer to do so if I gained access to an H95, as well as I would open the amp (when permitted) and try to measure the DAC chip's true direct output to see if any analog stages in between do degrade the signal, etc.
Soundstage's measurements:
The important info is this section from the Soundstage report:
"We attempted to optimize the volume position to achieve the best signal-to-noise ratio (SNR) and THD+N measurements at the speaker outputs (8 ohms) for both analog and digital inputs. For all analog input primary measurements (unless otherwise specified), the best measurements were achieved with the volume was set to 99. At a volume setting of 99, 240mVrms was required at the input for 10W (8 ohms) at the output. For the digital input, we found a volume setting of 59 yielded 10W (8 ohms) at the output for a 0dBFS input. We found essentially no differences between the optical and coaxial S/PDIF inputs, as well as the USB input, in terms of THD+N."
Line Input:
"Shown above is an FFT for a 1kHz input sinewave stimulus at the analog input, measured at the output across an 8-ohm load at 10W. Here we see that the second harmonic, at 2kHz, is just below -90dBrA, or 0.003%, while the third harmonic, at 3kHz, is just above -90dBrA. The other harmonics measured lower, below -100dBrA, or ?0.001%, down to about -125dBRa, or 0.00006%, at 20kHz. Below 1kHz, we see noise artifacts, with the 60Hz peak due to power-supply noise reaching -100dBrA (left), or 0.001%, and -110dBrA (right), or 0.0003%, and the 120Hz peak reaching just below -90dBrA. The harmonics of these two peaks can also be seen at lower levels."
DAC Input:
"Shown above are FFTs for a 1kHz 0dBFS dithered 24/96 input sinewave stimulus at the coaxial digital input, measured at the output across an 8-ohm load at 10W."
(Bold mine in all cases).
What we can infer from these plots that the DAC path is NOT significantly nosier (in terms of wideband noise) than than the line input and the distortion is very similar which tells us the DAC intrinsic distortion is NOT dominating either, for that very reasonable test conditions (0dBFS or equivalent analog input voltage, volume dialed in to achieve 10W/8R output and mentioned different pot setting for the analog input).
The difference in mains hum indicates PCB layout / cabling problems for the DAC lines (which we already noted). In general the measured performance sure is not spectacular, middle of the road at best.
Now, what we wanted to know about suitability for the occasional DAC- or PreAmp-only use-case is the behavior of DAC vs line input at the line output, which neither SS nor Amir measured. I'd volunteer to do so if I gained access to an H95, as well as I would open the amp (when permitted) and try to measure the DAC chip's true direct output to see if any analog stages in between do degrade the signal, etc.
I still don't really get it, he can still get 5W output by having full amplitude digital signal and bring the volume knob on the amplifier down.
It's very counterintuitive and frankly kind of borderline malicious tbh.
tw99
Senior Member
Malicious ? Seems a bit strong.
It's just an effect of the way he's attempting to standardise the measurements between integrated, AV, and power amps (using the volume control to set the gain to a "standard" level). If he did it a different way, people would probably just complain that the differing gain levels invalidates the comparison between the different amp types.
"Perfect" is not the way I will describe an amplifier that measures and behaves that badly.
OP
- Thread Starter
- #915
I have. Volume control was kept the same as analog test and I adjusted the digital value to get the same 5 watts. It is the only way to get apples vs apples comparison of digital versus analog. The analog value was also quite small so digital had to be as well (given the 29 dB gain set on the amplifier).
@amirm was feeding a digital signal and measuring the analogue outputs. How else should we supposed to use the amp???
Are you telling us that the digital input should never be fed full scale signal?
OP
- Thread Starter
- #917
Once again: we are using standardized THX gain level for an amplifier which is 29 dB. That is the only gain setting I test an amplifier at. Whether input is digital or analog, it is tested at the same gain level.
I'm one of those people who complain about this, so i get you.
Personally i would make an exception once the analogue input of the Audio precision is no longer used, as the low noise floor of the AP is what makes designing a low gain amplifier attractive for reviews.
Count De Money
Member
- Joined
- May 30, 2020
- Messages
- 49
- Likes
- 77
He's telling you not to do that while cranking the analogue volume pot of the amplifier all the way up to 95%.
Something has to give here.
Similar threads
- Poll
- Poll
