I have the brother to Shep's iceamp and love it. I find it to be a very neutral amplifier. The bass is strong and clean, perhaps because of the tight integration of the PSU. Some have commented that it is a little "soft on top" and it may be, but that may also be why I like it - much like the old Hiraga 20W Class-A.
The only distortion I've seen coming out is a lone 3rd harmonic. Maybe not ideal, but it does not sound bad. And it never drives me out of the room like amps or other devices with higher order distortion. So is the Ice soft on top, or just clean? I don't know.
I know that many Hypex snobs turn up their noses at Ice, but I found it a very clean, musical, pleasant amp. Alas, I pulled it off the table and it went bang on the floor. Has not worked sense and I can't find the cause. Sniff.... 🙁
It's interesting that you should mention this. Once upon a time I considered myself an 'audiophile' and I read and used to loved reading subjective articles. I still do like reading some subjective articles but that's usually because I enjoy Ken Kesslers humour more then anything else.
Then as time went on, due to the desire to learn how to do it myself and my participation in DIYaudio, my perspective on things completely changed. This isn't to say that before I would ignore objective results, far from it, I always wanted to buy gear that performed well, but it wasn't the specs that had the final say. Now though I find myself designing purely to meet specs and actually wanting to obtain a certain level of objective performance because it shows, more then anything, that my implementation(s) has/have been successful - I also learn a lot while doing it which gives me no end of satisfaction. The thing I find interesting though (and of course not surprising) is that I cannot really stand reading the pages and pages of audiophile subjectivism. If I buy a hifi mag everything has to have a set of measurements or it wont get bought (okay if it's got lots of pretty pictures of the insides of equipment I'd buy it🙄).
I'd say that DIYaudio had a great influence on getting me to the point where I am today, so to say that you cannot expect that kind of behaviour here is perhaps not entirely a fair statement. I think the interesting point to perhaps make is that it's far easier to be a subjective DIY tinkerer etc as it doesn't require much in-depth knowledge. The more you learn, the more you appreciate the objective and then your objectives shift too. DIYaudio is also a perfect place for encouraging learning.
As to class D amps, the only one I've really experienced is the TAS56xx series from TI and now I'm going to get subjective 😛 The amplifier itself had a warm and fuzzy/cozy kind of sound to it, it didn't put a foot wrong, but was it obviously coloured. Now the distortion spectra that this thing produced, is imo far less then ideal so is probably one of the things that could have perhaps contributed to it. Not a bad amp though, but then I'd prefer to listen to my blameless amps simply because they measure better and I trust measurements a whole lot more then I trust my ears.😎
Hi there 5th e
Interesting to read what you have to say.
I used to work as a design engineer in telecoms, there's an audio element but I gravitated more and more to RF. We never worked from anything other than datasheets (and measurements to verify them). Consequently, when I revived my interest in audio (I've been a semi-professional musician too) I was stunned to discover that it bore no resemblance to the professional electronics I had been engaged in, or, for that matter, the Hi-Fi scene of the '70s which was when I last paid attention much to audio.
How is anybody expected to design under these conditions? Poor measured performance is just as likely to be ignored by the pundits. Good measured performance is no guarantee of success. Even the most level-headed of us are likely to have our judgement suborned, so pervasive is the disinformation, and so persuasive is the argument that the tiniest possible improvenent, while almost certainly inaudible (0.0004% THD vs. 0.004%?), can do no harm.
Some of what I write is undoubtedly exaggerated, but I write it as an antidote when I feel that the unreality is gaining too much acceptance. It's intended to be creative, not destructive. I think the main problem is that it's easy to slide over into being pompous, so I'll shut up now.
w
Interesting to read what you have to say.
I used to work as a design engineer in telecoms, there's an audio element but I gravitated more and more to RF. We never worked from anything other than datasheets (and measurements to verify them). Consequently, when I revived my interest in audio (I've been a semi-professional musician too) I was stunned to discover that it bore no resemblance to the professional electronics I had been engaged in, or, for that matter, the Hi-Fi scene of the '70s which was when I last paid attention much to audio.
How is anybody expected to design under these conditions? Poor measured performance is just as likely to be ignored by the pundits. Good measured performance is no guarantee of success. Even the most level-headed of us are likely to have our judgement suborned, so pervasive is the disinformation, and so persuasive is the argument that the tiniest possible improvenent, while almost certainly inaudible (0.0004% THD vs. 0.004%?), can do no harm.
Some of what I write is undoubtedly exaggerated, but I write it as an antidote when I feel that the unreality is gaining too much acceptance. It's intended to be creative, not destructive. I think the main problem is that it's easy to slide over into being pompous, so I'll shut up now.
w
Hi there M. Long time...It makes me sad that little brother is dead 🙁 I know you can fix it! Thanks for stepping in. No one believes me! and you should hear mine! you did me proud.
I 've posted some sims with source on Black's Feedforward, aka "Current dumping" - nothing requres it only be applied to Class C
and the AES paper discussed applying Black's Feedforward to Class D
http://www.diyaudio.com/forums/solid-state/184725-alternative-topologies-3.html#post2506046
Waki and 5thE,,,
I think you guys overstate the disregard that some audiophiles have towards measurements. In no way I am looking to get in an argument with either of you, but I do feel a "measurements only approach" to high end audio design is just as "wrong" as a subjective "listening only" approach would be.
Besides some esoteric designs, or old school tube designs, it appears to me that most high end audio amplifiers measure very well when it comes to the standard set of measurements. I find no point in discriminating between amplifiers based on .001% vs. .0001% distortion figures, as this difference in measured perfromance alone is not above the threshold of audability.
The fact remains that most audiophiles hear differences between amplifiers which have nearly identical measurements (and/or measurements where the differences are small enough to be inaudible). Clearly there is more to the sonic picture than what the standard measurements are able to discern. For an audiophile, who listens to music for pleasure, to choose an amplifier by subjective listening, makes perfect sense; after all, the enjoyment of the music comes from subjective listening as well.
I prefer components which both sound good to me, and exhibit good engineering through a good set of measurements.
I think it is fair to say that we do not currently know exactly how to measure all aspects of an amplifiers performance. Nordost and Chord are sponsoring the development of new measurement techniques, (with the help of English Naval contractors who do development work on sonar systems). These techniques use actual music program material, and analyize the performance in both the time and amplitude domains.
I think keeping an open mind to both the current state of the art in terms of measured performance, and subjective evaluation is healthy for the future of high end audio playback. Remember in the early days of solid state there were a lot of awful sounding amplifiers around (which measured fine by the standard measurements of the day). Then new measurements were developed which discovered the distortions which were (mostly) responsible for the poor sound of these early designs (I believe this was TIM, but I could be remembering my history wrong, please correct me if so, if Mr. Pass is still monitoring this thread he may be able to provide the correct historical perspective).
I think you guys overstate the disregard that some audiophiles have towards measurements. In no way I am looking to get in an argument with either of you, but I do feel a "measurements only approach" to high end audio design is just as "wrong" as a subjective "listening only" approach would be.
Besides some esoteric designs, or old school tube designs, it appears to me that most high end audio amplifiers measure very well when it comes to the standard set of measurements. I find no point in discriminating between amplifiers based on .001% vs. .0001% distortion figures, as this difference in measured perfromance alone is not above the threshold of audability.
The fact remains that most audiophiles hear differences between amplifiers which have nearly identical measurements (and/or measurements where the differences are small enough to be inaudible). Clearly there is more to the sonic picture than what the standard measurements are able to discern. For an audiophile, who listens to music for pleasure, to choose an amplifier by subjective listening, makes perfect sense; after all, the enjoyment of the music comes from subjective listening as well.
I prefer components which both sound good to me, and exhibit good engineering through a good set of measurements.
I think it is fair to say that we do not currently know exactly how to measure all aspects of an amplifiers performance. Nordost and Chord are sponsoring the development of new measurement techniques, (with the help of English Naval contractors who do development work on sonar systems). These techniques use actual music program material, and analyize the performance in both the time and amplitude domains.
I think keeping an open mind to both the current state of the art in terms of measured performance, and subjective evaluation is healthy for the future of high end audio playback. Remember in the early days of solid state there were a lot of awful sounding amplifiers around (which measured fine by the standard measurements of the day). Then new measurements were developed which discovered the distortions which were (mostly) responsible for the poor sound of these early designs (I believe this was TIM, but I could be remembering my history wrong, please correct me if so, if Mr. Pass is still monitoring this thread he may be able to provide the correct historical perspective).
Well, my Onkyo 9.1 receiver, which cannot be criticized as being short on technology in general, definitely uses analog power amplifiers, and probably biased somewhat into class A, judging by the heat output. Judging by the SQ, I believe that is probably the approach I would have been happiest with with high quality Blu Ray 24 bit audio and analog inputs, anyway.
From what I've heard, Class D amplifiers' greatest advantage over conventional solid state amplifiers was in reducing or eliminating crossover distortion, but the Class D I've heard never offered really superior transparency or liquidity, where a competently designed conventional SS amp or most tube amps can outdo them.
Plus, there is the issue with radiated emissions interfering with wireless signals which has to be a significant design problem, especially for higher power Class D amplifiers. Who would want a class D amplifier of any SQ if it brought down your wireless network or trashes OTA reception?
Btw, I wouldn't be misled by those who tout super-tweaked ultra expensive class D amplifiers operating on that technologies' furthest edge as sounding 'practically' as good as analog amplifiers when what matters to 99% of us is whether Class D amplifiers that most of us have access to consistently meet or exceed A & AB SQ across the board. And the answer there is, 'no'. That is the same type of cheerleading that locked audio into 30 years and counting of lousy CD quality (and worse) audio.
From what I've heard, Class D amplifiers' greatest advantage over conventional solid state amplifiers was in reducing or eliminating crossover distortion, but the Class D I've heard never offered really superior transparency or liquidity, where a competently designed conventional SS amp or most tube amps can outdo them.
Plus, there is the issue with radiated emissions interfering with wireless signals which has to be a significant design problem, especially for higher power Class D amplifiers. Who would want a class D amplifier of any SQ if it brought down your wireless network or trashes OTA reception?
Btw, I wouldn't be misled by those who tout super-tweaked ultra expensive class D amplifiers operating on that technologies' furthest edge as sounding 'practically' as good as analog amplifiers when what matters to 99% of us is whether Class D amplifiers that most of us have access to consistently meet or exceed A & AB SQ across the board. And the answer there is, 'no'. That is the same type of cheerleading that locked audio into 30 years and counting of lousy CD quality (and worse) audio.
Last edited:
"From what I've heard, Class D amplifiers' greatest advantage over conventional solid state amplifiers was in reducing or eliminating crossover distortion"
No. Crossover distortions is the price we pay for power efficiency.
No. Crossover distortions is the price we pay for power efficiency.
Hardly the case, then PC power supplies and any switching devices should be discriminated against for interfering with your wireless.
No cross-over distortion in switching amplifiers should therefor be equivalent to class A, and a switching amplifier that has lower harmonic distortion than class A should sound better. Maybe it is audiophiles that are unable to recognize that it sounds better.
For a non Class D amplifier, perhaps....
You aren't seriously proposing that the power switching transitions within a Class D amplifier equate with a Class B amplifier's crossover linearity problems, are you?
If you are, then I must disagree - they're different phenomena in most ways (to put it mildly) with regard to the effects they produce on the signal as seen at the speaker terminals. I write this as a person who has designed both Class D and Class AB amplifiers.
Last edited:
Not at all. The PC power supply you refer to is shielded and RFI filtered within its own Faraday cage and is also a low cost commodity with huge volume which is not the case for most other comparable electronics. I say it's not particularly practical or easy to develop 1000 + watts of switching Class D power amplification within the SAME chassis in a cost competitive receiver on a short design cycle that also is packed with high resolution analog and digital audio and video processing, supports Blue Tooth, WiFi, AM and FM reception and has literally dozens of I/O ports.
Which of these high quality features exist in a RFI noise hell PC anyway? I have never heard *anybody* accuse PC expansion slot boards of having particularly good analog quality, largely because of RF spray. To sum up, I don't see the sense in pretending that these issues are negligible when Onkyo clearly doesn't. Besides, I've designed a number of telecom and power circuits that have obtained international approval, so I can claim a real world basis for my opinion.
Last edited:
I think you perhaps read too much into what I was saying. I don't disregard anyone's right to choose and listen/review equipment based on subjective involvement, it's just that I will take their opinion with a grain of salt - meaning that it is only what they think rather then what actually is.
You are implying somewhat that these audiophiles are actually able to discern a real difference between amp A and amp B. I have no doubt about peoples claims when these things are done in an environment in which they are in control of. I would expect to see these differences pretty much disappear mind you when double blind tests are used.
As an example I know full well that a system can sound glorious one day, but horrible the next. What changed in the system between then and now? Nothing, the only thing that changed was myself and I know that the way I am feeling both physically and mentally can dramatically affect my perception of the way something sounds.
If when the system remains completely unchanged, I am capable of swinging between such extremes, then it is very easy to see how people can think that one bit of kit sounds ever so slightly different to the next.
For one I don't think that I can actually hear the difference between 0.0001% and 0.001% distortion. I do know however that my mind is satisfied if all the components of a system are implemented in the best way I know how. If this isn't the case then there is a general sense of unease in my mind and I know this has an extremely strong chance of being able to influence the way in which I perceive the system to sound.
It will be interesting to see what Chord etc can come up with. There have for decades now, been people saying, we can't measure why this sounds better, but it does. I for one am not happy with this as we should be able to measure and then explain why it sounds better. If we can't then a double blind test must be invoked because the only logical explanation left is that we're imagining it.
All of this only really applies to electronics mind you, loudspeakers being a completely different kettle of fish for a number of reasons.
There is another thing to be said about this though. Peoples quest for the perfect system is most likely done because they find the pursuit enjoyable. My hunting for the lowest THD say, some would consider a useless pursuit, because they'd say you won't hear it. I don't care, I'm doing it because I enjoy the challenge and enjoy the (most likely not aural) reward(s) it brings.
This I think is probably the most important thing, people like to feel as if they've made the right decisions and that they've brought a worthwhile contribution towards the system and when they do they are content. The system will sound good because they believe it will do because they made the right decision to buy that piece of equipment that everyone says sounds amazing.
I personally have a huge issue with someone modding something and totally destroying it's amazing technical specifications. In what way did you better the product? Everything went from good to bad. Put it back how it was and be happy with it, you don't need to tweak something to feel happy with it. No doubt this is like telling a bird not to fly because the individual quite clearlyneeded to involve themselves somehow with the product before they felt content with it.
The sound aside though, class D can easily bring a lot of psychological satisfaction for a number of very real/beneficial reasons.
Frankly, as a diy'er, Class D for me is of no interest until I can build something for myself that isn't just black box ICs on a near reference PCB. If someone could come up with a P3A or Symasym equivalent project that I could build for myself from discretes, then I might have a go. Until then, I'll stick with my huge mountain of BJTs. 😉
I mean both comparator threshold and switching time. They can't be avoided absolutely, only minimized. I write this as both an engineer that studied possibilities of this designs before they ever got "Class D" name, and an experimenter who did own homework. As I wrote before, my tape recorder worked very nice, though neighbors heard what I record on their AM receivers. 😉
- Status
- Not open for further replies.