• 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!

Review and Measurements of Hypex NC400 DIY Amp

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
You can self-oscillate both half and full-bridge topologies. Both BD and AD approaches can be combined with self-oscillation:

https://core.ac.uk/download/pdf/9703959.pdf
On paper, yes. Any source for a practical implementation? With measurements?

As for my other claim about bridging before the filter, it is based on common sense and power converters therapy, which you can read here:

D. G. Holmes, Thomas A. Lipo, Pulse Width Modulation for Power Converters Principles and Practice, IEEE Press., 2003.
Not following you on common sense. I guess I miss some. Question is proving that bridging is better. Sorry, I don't have a IEEE account.

The higher the switching frequency, the lower the load's current THD. The higher the switching frequency fed to the low pass output filter, the lower the harmonic distortion at its output because the cutoff frequency of the filter shifted to a lower frequency with respect to the switching frequency. You can easily verify this claim by simulation. You can self-oscillate a full-bridge class-d amp by using two separate control feedback loops, one for each half-bridge leg with two input source signals (positive and negative balanced input).

The full-bridge class-d, when implemented correctly, should hade better performance compared to a half-bridge class-d with the same control strategy, same switching frequency range, and the same output filter.
Why doubling the complexity of something conceptually simple and working perfectly?
Moreover, since you have a PhD in engineering, you should know that each solution is a tradeoff and that no perfect solution exists.
In example, the higher the frequency, the higher the thermal losses. The higher the thermal losses, the hotter the filter. The hotter the filter, the higher distortion.

BD-modulation (full-bridge only) could be either filter-free or filtered, while AD-modulation (half-bridge only) must be filtered. This is also an indication that full-bridge is alway superior, regardless of the modulation technique used (constant open loop PWM, or self-oscillating closed-loop).
Again, you can full bridge with AD modulation. Do you have an example of a filterless BD modulated amplifier with 1kW output passing conducted and radiated emissions certification?

Out of curiosity, did you compare measurements between Hypex and it's competitors? Did you find amplifiers implementing what you consider as being a technically superior solutions?
 

6speed

Active Member
Joined
Nov 7, 2018
Messages
128
Likes
84
Location
Virginia, USA
Are there any retails products that allow us to compare the half vs full bridge difference in question (e.g. UcD700 vs UcD2k, or are there too many other variables that differentiate them? There aren't a lot of people who can completely follow the academic discussion, and of those that can, most will trivialize some portion of it because they don't have BP's level of experience to speak to every real world trade off. If Hypex, ICEpower, or someone else has 2 modules that can serve as reference points that can also be measured, that would be helpful. Otherwise, you risk comparing very different products from different manufacturers and have no hope of attributing measured performance differences to academic considerations.
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands
Are there any retails products that allow us to compare the half vs full bridge difference in question (e.g. UcD700 vs UcD2k, or are there too many other variables that differentiate them? There aren't a lot of people who can completely follow the academic discussion, and of those that can, most will trivialize some portion of it because they don't have BP's level of experience to speak to every real world trade off. If Hypex, ICEpower, or someone else has 2 modules that can serve as reference points that can also be measured, that would be helpful. Otherwise, you risk comparing very different products from different manufacturers and have no hope of attributing measured performance differences to academic considerations.

Stupid question, but who is "BP"?
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
Are there any retails products that allow us to compare the half vs full bridge difference in question (e.g. UcD700 vs UcD2k, or are there too many other variables that differentiate them? There aren't a lot of people who can completely follow the academic discussion, and of those that can, most will trivialize some portion of it because they don't have BP's level of experience to speak to every real world trade off. If Hypex, ICEpower, or someone else has 2 modules that can serve as reference points that can also be measured, that would be helpful. Otherwise, you risk comparing very different products from different manufacturers and have no hope of attributing measured performance differences to academic considerations.
NC252MP, in both single and BTL. Ok, not a true full bridge, but can give an idea.
NC252MP 4 Ohms.png
NC252MP 8 Ohms.png
NC252MP 4 and 8 Ohms BTL.png
 

Julf

Major Contributor
Forum Donor
Joined
Mar 1, 2016
Messages
3,004
Likes
3,998
Location
Amsterdam, The Netherlands

6speed

Active Member
Joined
Nov 7, 2018
Messages
128
Likes
84
Location
Virginia, USA
I read the NC252MP graphs as
  1. Distortion goes up with BTL at low power levels (<~2W).
  2. Lower impedances create more distortion with BTL.
  3. Distortion with BTL may go down at moderate power levels and higher impedances.
  4. The distortion at max power does not make sense to me with 8ohm. Why does 4ohm not behave worse?
 

HammerSandwich

Major Contributor
Forum Donor
Joined
Nov 22, 2018
Messages
1,137
Likes
1,497
@EB1000, I'm no EE. You've mentioned common sense, so I hope I might ask questions without giving offense.
  1. Do you believe that instrumented testing can describe an electronic component's performance? If not, how do you determine which amplifiers perform better than others?
  2. Have you examined the relevant datasheets? The bench testing I've seen verified specifications from Hypex, ICE & Purifi. And it indicates that the 700AS1 offers the lowest performance, except for output power. (OTOH, the additional watts might be more useful than the lower distortion in a given system...)
  3. Does full-bridge inherently have theoretical disadvantages as well? For example:
    1. Are there potential issues matching the 2 sides of the signal that half-bridge naturally eliminates?
    2. Can summing to single-ended at the input work better than amplifying signals which will be nulled at the output?
  4. Is it possible for a talented designer to produce a better final result with a theoretically lower-performance process? IOW, have you possibly confused the ingredients with the meal?
FWIW, I believe we should look at an amplifier/DAC/etc. as a black box for performance analysis. When we find issues, it's appropriate to analyze the problem's source & consider how to redesign for improvement. Weighing an amplifier's topology above its actual output confuses me.
 

hfavandepas

New Member
Joined
Jan 10, 2020
Messages
3
Likes
3
Location
The Netherlands
‘The filtering must be passive’ Not any longer…..

‘The filtering must be passive (e.g. like the one in your speakers)’. But there are other ways too.
These guys from Axign developed a digital way too filter and control a Class D amplifiers.
See https://www.axign.nl
Or see this article ‘Axign of Things to Come’ in the December 2017-issue from Audio-express.
https://audioxpress.com/files/attachment/2659

I first heard an Axign demo-amplifier at the X-fi-audio show in 2017 and was very, very impressed.
But since there are no axign-based pre-amps yet, I can not use them with my active monitors

But a few weeks ago a friend of mine wanted to replace his Naim-amplifier. And he also considered to buy a mains-filter because when the microwave-oven in the kitchen was used, it could be heard through the speakers in the living room.

I persuaded him NOT too buy a mains-filter yet, but too first have a listen to an Axign-based amplifier. Axign targets the B-to-B market (audio-industry) and has no plans for producing amplifiers for the consumer-market. However, the Axign guys have a Philips Semiconductors (NXP) back-ground and so has mr Abbe de Groot from Vintage Audio Repair.
In close co-operation with Axign mr Abbe de Groot builds Axign-based amplifiers for the consumer-market.
See https://truebluebox.nl/

Long story short: 2 weeks ago my friend listened to the axign-based amp at Vintage Audio Repair , took it home for listening in his living room (digital source: Bleu-Sound Node 2i and streaming Tidal Lossless Flac) and: ….. SOLD ! :).
And the microwave-oven? Not a trace. Speakers (passive) hooked up too an axign-based amp are super, super silent.
Even with my ears a few centimeters from the speaker, I could not even hear the slightest hiss. Just absolute silence.

I do realise that ‘good sound’ is a personal matter. Lovers of a tube-style sound, will probably not like this axign-based amp. But people who like fast, punchy, precise, controlled, detailed, spot on imaging, will like this amp very much (I do !) For comparison, I like the sound from brands like: Barefoot, PMC, Klein &Hummel (Neumann), ATC, PSI, Dutch&Dutch, Dynaudio, Geithain. DAC’s: Merging, Crane Song, Weiss and recent designs of RME (again).

Unfortunately it’s not possible too use the True Blue Box as an pre-amp for my active PMC TwoTwo.6 or my active Genelec’s 8351. I very much like the Axign-design because it's immune too noisy mains and power supply ripple, etc.
But unfortunately there are no plans for an Axign-based pre-amp yet. :-(

So anyone having the opportunity too listen too an Axign-based amp, do try!
Or just wait till the audio-industry starts building axign-based amps for the consumer market.
 

Attachments

  • Dec2017AxignAmpSeries.pdf
    5.8 MB · Views: 218

landco

Member
Joined
May 7, 2019
Messages
69
Likes
7
Location
Russia
I happen to have a PhD degree in electrical engineering
Can you confirm my opinion that Hypex Electronics amplifiers made by engineers of low-skilled and better to buy products of other manufacturers, such as IcePower, Infineon Technologies or Texas?
 

EB1000

Senior Member
Joined
Jan 8, 2020
Messages
484
Likes
579
Location
Israel
@EB1000, I'm no EE. You've mentioned common sense, so I hope I might ask questions without giving offense.
  1. Do you believe that instrumented testing can describe an electronic component's performance? If not, how do you determine which amplifiers perform better than others?
  2. Have you examined the relevant datasheets? The bench testing I've seen verified specifications from Hypex, ICE & Purifi. And it indicates that the 700AS1 offers the lowest performance, except for output power. (OTOH, the additional watts might be more useful than the lower distortion in a given system...)
  3. Does full-bridge inherently have theoretical disadvantages as well? For example:
    1. Are there potential issues matching the 2 sides of the signal that half-bridge naturally eliminates?
    2. Can summing to single-ended at the input work better than amplifying signals which will be nulled at the output?
  4. Is it possible for a talented designer to produce a better final result with a theoretically lower-performance process? IOW, have you possibly confused the ingredients with the meal?
FWIW, I believe we should look at an amplifier/DAC/etc. as a black box for performance analysis. When we find issues, it's appropriate to analyze the problem's source & consider how to redesign for improvement. Weighing an amplifier's topology above its actual output confuses me.




  1. Do you believe that instrumented testing can describe an electronic component's performance? If not, how do you determine which amplifiers perform better than others?
Yes. Actually, when I shop for an audio component, I only rely on instrumental testing and compare all figures of merit. Not that I don't trust my own years, but if it measures good, it will always sound good.


  1. Have you examined the relevant datasheets? The bench testing I've seen verified specifications from Hypex, ICE & Purifi. And it indicates that the 700AS1 offers the lowest performance, except for output power. (OTOH, the additional watts might be more useful than the lower distortion in a given system...)

Yes, I did, and I did notice that on paper the NC400 and the pufrifi look better. I did ask one engineer that works for PS Audio, why did they choose Icepower modules over Hypex, and he told me that they did evaluate many different class-d brands, including Hypex, and concluded that the Icepower ASC series sounded best with their analog cell (I dod not pwn any PS Audio products, I have a NAD C658 Preamp/DAC).

Anyway, I still haven't decided if I'll go with 700AS1 or NC400, but so far the 700AS1's true balanced topology seems to be a more logical choice.
  1. Does full-bridge inherently have theoretical disadvantages as well? For example:
    1. Are there potential issues matching the 2 sides of the signal that half-bridge naturally eliminates?
    2. Can summing to single-ended at the input work better than amplifying signals which will be nulled at the output?
1 - Yes, BD-based full-bridge tend to have more 5th harmonic distortion at low modulation depths (low-level listening)
2 - I can't answer that with 100% confidence for class-d amps. As for my field of expertise, which is power conversion (mostly grid-connected three-phase DC to AC), then the answer is yes. There are many academic papers to support that claim, but they only discuss fixed frequency PWM, not self-oscillation as in moist class-d.


Meanwhile, I've contacted both Bruno Putzeys from Hypex and Karsten Nielsen from Icepower and asked them some very technical questions about their choice of topologies and control schemes. I really hope to get some replies. I'll keep you updated.
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
Can you confirm my opinion that Hypex Electronics amplifiers made by engineers of low-skilled and better to buy products of other manufacturers, such as IcePower, Infineon Technologies or Texas?
LOL, calling Bruno a low skilled engineer is quite innovative!
Can you explain why companies like NAD prefer buying their hardware to a 20 people firm than integrating whatever chip from a giant like TI? Are they also a bunch of idiots?
Denon uses a "full bridge" + AD-modulation, see DRA-F109 class D amplifier.
Well, they used whatever chip from ST or TI
 

Voxy

Member
Joined
Jan 9, 2020
Messages
20
Likes
2
As I said, NC400 - is redundant solution. When you are in search of minimal distortion from your amplifier, you must not forget that the main source of distortions - your speaker system - 1% 2% 3% ... even if you have good speakers.

But this fact prefer to forget brands like HYPEX. All these results of laboratory measurements - no more than a marketing ploy.

If you are not satisfied with ready-made solution and you want to build your own power - turn your attention to IRAUDAMP7S from Infineon Technologies, this solution is significantly cheaper, so, I repeat, you will not hear any difference in the sound.

As for my personal preference, HYPEX I do not like, because, in addition to amplifiers, they trade audiophile cables 200-300 dollars. It seems to me a serious chip manufacturer will not trade audiophile cables :)

Therefore, in my personal taste, Infineon Technologies - this is much more of a company in terms of technology than HYPEX. Earlier Infineon Technologies company was a division of Siemens AG Semiconductor Group.

P.S. More information about the amplifiers based on IRAUDAMP7S you can look at the channel https://www.youtube.com/channel/UC2yDSO46soRXRvQyhOfKr9Q/videos

P.P.S. If you want the finished product for a low price, I recommend to pay your attention to the second-hand market of amplifiers class D, eg, Denon DRA-F107, Denon DRA-F109. On paper, the characteristics of the chip is lower than the NC400, but in practice you hardly hear any difference :) for $ 100-150 is a very good offer

I am interested in those Infineon amps. Are there completed ones with the mentioned power supply in the market?
 

EB1000

Senior Member
Joined
Jan 8, 2020
Messages
484
Likes
579
Location
Israel
As I said, NC400 - is redundant solution. When you are in search of minimal distortion from your amplifier, you must not forget that the main source of distortions - your speaker system - 1% 2% 3% ... even if you have good speakers.

But this fact prefer to forget brands like HYPEX. All these results of laboratory measurements - no more than a marketing ploy.

If you are not satisfied with ready-made solution and you want to build your own power - turn your attention to IRAUDAMP7S from Infineon Technologies, this solution is significantly cheaper, so, I repeat, you will not hear any difference in the sound.

As for my personal preference, HYPEX I do not like, because, in addition to amplifiers, they trade audiophile cables 200-300 dollars. It seems to me a serious chip manufacturer will not trade audiophile cables :)

Therefore, in my personal taste, Infineon Technologies - this is much more of a company in terms of technology than HYPEX. Earlier Infineon Technologies company was a division of Siemens AG Semiconductor Group.

P.S. More information about the amplifiers based on IRAUDAMP7S you can look at the channel https://www.youtube.com/channel/UC2yDSO46soRXRvQyhOfKr9Q/videos

P.P.S. If you want the finished product for a low price, I recommend to pay your attention to the second-hand market of amplifiers class D, eg, Denon DRA-F107, Denon DRA-F109. On paper, the characteristics of the chip is lower than the NC400, but in practice you hardly hear any difference :) for $ 100-150 is a very good offer

I've looked at the specs of the Iraudamp7s, its a dual-channel half-bridge with relatively low self-oscillating freq of 400KHz. It can do full-bridge BTL, but it will limit the speaker impedance to 8 ohms. This is a major drawback. You'll be better off with Icepower modules. The newer AS series, such as the 300AS1 or the 700AS1 are full-bridge mono amps with self-oscillating freq of 550kHz per each bridge-leg, this means an effective 1.2MHz switching frequency at the differential output.

You are correct about the required amount of distortion, but don't forget that dynamic range also plays an important role.
 

Koeitje

Major Contributor
Joined
Oct 10, 2019
Messages
2,292
Likes
3,880
Can you confirm my opinion that Hypex Electronics amplifiers made by engineers of low-skilled and better to buy products of other manufacturers, such as IcePower, Infineon Technologies or Texas?
Bruno is one of the best audio engineers in the world.
 

EB1000

Senior Member
Joined
Jan 8, 2020
Messages
484
Likes
579
Location
Israel
LOL, calling Bruno a low skilled engineer is quite innovative!
Can you explain why companies like NAD prefer buying their hardware to a 20 people firm than integrating whatever chip from a giant like TI? Are they also a bunch of idiots?

Well, they used whatever chip from ST or TI

I wouldn't call him a low skilled, checking his academic background, he only holds a bachelor's degree in EE, and according to IEEE xplore, he was published 4 times. This is actually considered a lot for someone with only a BSc degree, When I received my BSc, I did not have any publications, except for one conference paper. By the time I received my Ph.D., I had only 4 publications...


He is definitely very skilled in his field of expertise. But since Karsten Nielsen - the founder of Icepower have a Ph.D. degree, I decided to go with an Icepower module for my DIY project, and not a Hypex module (just kidding.. :) )
BP.jpg
 

boXem

Major Contributor
Audio Company
Joined
Jun 19, 2019
Messages
2,014
Likes
4,852
Location
Europe
I wouldn't call him a low skilled, checking his academic background, he only holds a bachelor's degree in EE, and according to IEEE xplore, he was published 4 times. This is actually considered a lot for someone with only a BSc degree, When I received my BSc, I did not have any publications, except for one conference paper. By the time I received my Ph.D., I had only 4 publications...


He is definitely very skilled in his field of expertise. But since Karsten Nielsen - the founder of Icepower have a Ph.D. degree, I decided to go with an Icepower module for my DIY project, and not a Hypex module (just kidding.. :) )View attachment 45412
Sheldon Cooper, is it you?

OT: the different approaches of diploma pending the culture is interesting. When I got my master in engineering, I was considering people targeting a PhD as lazy asses who didn't want to work. Then I worked with Germans and discovered how being a "Doktor" could help your career.
More than 20 years later, I just focus on achievements and don't even look at the diploma.
 
Top Bottom