Reverting to the post-starter's question, I would say that for efficient speakers and most bookshelf speakers, any Class-D amplifier would compare poorly to something like this:
20W Class A Amplifier Full Kit - Volt Electronics
And the price is for everything including the chassis.
20W Class A Amplifier Full Kit - Volt Electronics
And the price is for everything including the chassis.
Sorry for self-quoting but it probably would not put a 250mV 200KHz potentially tweeter-frying signal into your speakers either!
Are you really claiming that such a small signal would do any damage to the tweeter? At these high frequencies, the tweeter impedance is usually in the range of about 30 Ohm, which limits the tweeter current to about 8 mA, which would generate a miniscule power well below 1 mW on the resistive part of the tweeter impedance. This is completely safe and orders of magnitude below of what is required to damage a tweeter. I suspect that the damages reported are due to some malfunction or improper DIY-mods. You can get any non-class-D amp into oscillation and burn the tweeter as well; sometimes an improper mod is not even required, in some cases it may be suffient to connect expensive, boutique speaker cables with high enough capacitance. In summary, I don't think that it is fair to blame class-D here. I have UcD amps for several years now without any problem. And yes, I did make the above rough calculation before I built them.
Kurt
Maybe not on its own, but what if it were to beat with certain high frequency input signals (as might be present as an artifact on an SACD for example) and produce higher voltage square waves? I have seen it (with a function-generated sine wave input signal of 100kHz) on a scope to which my own UcD amp was attached.
I really don't know if damage would result. That's why I used the word "potentially", but tweeters are the least robust of the drivers and a Zobel network can always be added across the amplifier outputs (if you really want to drive tweeters with Class-D for some reason) to reduce the magnitude even further. And a Zobel network is a lot cheaper than some silly boutique speaker cable.
Yes other problem amps can oscillate, but Class-D amps must oscillate by design. To my conservative mind, one less oscillation in the audio path can only be a good thing.
I'm not anti-Class-D. I use it for bass in a bi-amp arrangemend with great results.
Hi,
I heard a Kharma MP 350 (Kharma International | Amplifiers | Exquisites | Mono Power Amplifier 350 (MP350)) , the basic I think a Hypex Modul like the UCD400, but with Modifications (I think on the power supply) and a very heavy Cabinet.
Very good sound. Did anybody know how you can tune a UCD400 or 700?
regards
Josef
I heard a Kharma MP 350 (Kharma International | Amplifiers | Exquisites | Mono Power Amplifier 350 (MP350)) , the basic I think a Hypex Modul like the UCD400, but with Modifications (I think on the power supply) and a very heavy Cabinet.
Very good sound. Did anybody know how you can tune a UCD400 or 700?
regards
Josef
The test was conducted by a very experienced and unaffiliated technician. It was sufficiently frightening to him that he refused to connect the amp to his speakers. It is therefore sufficiently frightening for me.
He suggested the addition of Zobels (10 Ohm, 100nF) across the amp's outputs if it is to be connected to tweeters (which it isn't but I saw no harm). He said the iron-cored inductor coil on each module was cheap and filtered too little of the carrier frequency. It doesn't matter what you call it. The word "oscillation" is good enough. It just happens to be designed into the thing rather than being caused by a fault condition.
An interesting (and well-informed) article that talks about amplifier output Zobels is here:
Elliott Sound Products - Audio Power Amplifier Design Guidelines
You suggest that Zobels are not robust, but the above article suggests to me that the capacitor short condition is only applicable at very high inaudible frequencies in which case the robustness is determined by the wattage of the resistor. If your Zobels are fragile, maybe you need bigger resistors.
If you like Class-D for tweeters then that's okay. I like Class-A and that's my prerogative and I am free to express it. And I do so without resorting to rudeness, obscure put-downs or rhetorical questions.
He suggested the addition of Zobels (10 Ohm, 100nF) across the amp's outputs if it is to be connected to tweeters (which it isn't but I saw no harm). He said the iron-cored inductor coil on each module was cheap and filtered too little of the carrier frequency. It doesn't matter what you call it. The word "oscillation" is good enough. It just happens to be designed into the thing rather than being caused by a fault condition.
An interesting (and well-informed) article that talks about amplifier output Zobels is here:
Elliott Sound Products - Audio Power Amplifier Design Guidelines
You suggest that Zobels are not robust, but the above article suggests to me that the capacitor short condition is only applicable at very high inaudible frequencies in which case the robustness is determined by the wattage of the resistor. If your Zobels are fragile, maybe you need bigger resistors.
If you like Class-D for tweeters then that's okay. I like Class-A and that's my prerogative and I am free to express it. And I do so without resorting to rudeness, obscure put-downs or rhetorical questions.
rude post deletedThat a potentially destructive test was voluntarily carried out on your own amps, as you stood there watching, further brings into question your own motivation and affiliations. Don't pass the buck by saying "he", italicized no less, nameless expert, did it. Hobbyists do not look for problems that they aren't experiencing, so what was your personal motivation in that test? If he were so expert he'd never have driven the input with such a signal in the first place, and I would very much like to see such a bogus test repeated with an amp of any other class to see what happens to your output and zobel alike. Try it, I dare you, and watch how fast it smokes. I am suggesting zobels as typically designed will blow apart when driven with an improper signal as you have done here.
Now your issue is not what might happen in the presence of an absurd signal; nothing it seems, but that the "iron"-- nameless expert technician term, filters too poorly. Is that all the time or some of the time or only in the presence of an absurd signal?
The word oscillation is apt in proper context and I found that wasn't how you used it, intentionally or not. That you'd defend your improper use that twisted the context for added shock and fear value, even after having been informed, is illuminating. Coupled with "other problem amps" implies something on your part, like ("but class d") is problem by design, which is true and that's what makes a good one so difficult.
Saying you like it for bi-amping is just repeating the common two decade old "they're only good for bass reproduction" in so many words. That's a false compliment like "expert technician" which is something of an oxymoron in itself, but according to you, you're above all these tactics, right. Even though your final statement was exactly what you said it isn't, but truth seems to be a problem here.
I mean, that's a whole lot of what ifs and pseudo expert, pseudo third party, pseudo conclusions, simply to state your preference, isn't it. Matter of fact the preference you stated wasn't so much "class A", but class A "like this one", which happens to also be an Aussie company, like where you and your expert friend are from.
You can add a zobel if you like, btw, but don't pretend there's some sort of instability issue that it has to tame in this amplifier.
You are free to express an opinion and others are free to call you on your obvious "errors", many of which went well and far beyond what one would consider an "opinion". What's the point in whining about a two way street if only have to have the moderators debate me where you failed to. Surely that's another tactic you're above.
Now your issue is not what might happen in the presence of an absurd signal; nothing it seems, but that the "iron"-- nameless expert technician term, filters too poorly. Is that all the time or some of the time or only in the presence of an absurd signal?
The word oscillation is apt in proper context and I found that wasn't how you used it, intentionally or not. That you'd defend your improper use that twisted the context for added shock and fear value, even after having been informed, is illuminating. Coupled with "other problem amps" implies something on your part, like ("but class d") is problem by design, which is true and that's what makes a good one so difficult.
Saying you like it for bi-amping is just repeating the common two decade old "they're only good for bass reproduction" in so many words. That's a false compliment like "expert technician" which is something of an oxymoron in itself, but according to you, you're above all these tactics, right. Even though your final statement was exactly what you said it isn't, but truth seems to be a problem here.
I mean, that's a whole lot of what ifs and pseudo expert, pseudo third party, pseudo conclusions, simply to state your preference, isn't it. Matter of fact the preference you stated wasn't so much "class A", but class A "like this one", which happens to also be an Aussie company, like where you and your expert friend are from.
You can add a zobel if you like, btw, but don't pretend there's some sort of instability issue that it has to tame in this amplifier.
You are free to express an opinion and others are free to call you on your obvious "errors", many of which went well and far beyond what one would consider an "opinion". What's the point in whining about a two way street if only have to have the moderators debate me where you failed to. Surely that's another tactic you're above.
I have the tentlabs DIY hybrid amplifier (TentLabs). It uses the UCD400 as power stage with the OPAMP's replaced by a valve stage (double triode) with step-down signal transformers.
I have compared this amplifier with the NAIM Supernait with a supercap powersupply (over 4400 euro's). The NAIM is no match to the Tentlabs, the Tentlabs outplays the NAIM easily if you look at the amount a detail you hear in the music.
For me is this proof that the UcD400, combined with a good input stage is a top class D amplifier.
I have compared this amplifier with the NAIM Supernait with a supercap powersupply (over 4400 euro's). The NAIM is no match to the Tentlabs, the Tentlabs outplays the NAIM easily if you look at the amount a detail you hear in the music.
For me is this proof that the UcD400, combined with a good input stage is a top class D amplifier.
I'd like to say that I have no commercial affiliations with any amplifier manufacturers, hi-fi stores or anyone in electronics. I have never heard that particular Class-A kit amp. It was chosen purely at random as a kit that may be of interest to readers of this forum becuse it is DIY. It is now available even more cheaply elsewhere online.
To have my words quoted, twisted, embellished and then analysed by an anonymous subscriber is quite to be expected, but can hardly be considered constructive.
My UcD400 modules leave a high frequency sine wave in the output. No matter what you do to the input buffer and power supply, that will always be there as it is generated onboard and downstream. If people are happy to attach tweeters with short, low capacitance cables then that's their choice.
To have my words quoted, twisted, embellished and then analysed by an anonymous subscriber is quite to be expected, but can hardly be considered constructive.
My UcD400 modules leave a high frequency sine wave in the output. No matter what you do to the input buffer and power supply, that will always be there as it is generated onboard and downstream. If people are happy to attach tweeters with short, low capacitance cables then that's their choice.
Ive had some commercial dealings with Class D amps so you might be interested in my
experiences.
Some years ago I was involved with the manufacture of a power amp based on the Tripath TA1040A module. While we had reliability problems with the module I dont
recall any problems with burnt tweeters. This module had from memory about 2V
of 660Khz idle noise of its o/p.
Since then I built many amps using the Hypex Ucd400 module which has been much more reliable with again no reported problems with burnt tweeters. This module also has some idle noise on its o/p
But we have had a few cases of burnouts with our ribbon supertweeter until we put in better protection. This has a flat mostly resistive impedance out to 100Khz so is more susceptable. This has happened with a wide variety of amp types not specifically Class D. We believe its due to some amps being driven into clipping.
I used to use the Ucd in my personal system quite happily until an colleague brought along a 300B DHT SE amp for me to try . Now I use the Ucd for Bass/Lower mid and the SE for the rest. Works superbly for me.
experiences.
Some years ago I was involved with the manufacture of a power amp based on the Tripath TA1040A module. While we had reliability problems with the module I dont
recall any problems with burnt tweeters. This module had from memory about 2V
of 660Khz idle noise of its o/p.
Since then I built many amps using the Hypex Ucd400 module which has been much more reliable with again no reported problems with burnt tweeters. This module also has some idle noise on its o/p
But we have had a few cases of burnouts with our ribbon supertweeter until we put in better protection. This has a flat mostly resistive impedance out to 100Khz so is more susceptable. This has happened with a wide variety of amp types not specifically Class D. We believe its due to some amps being driven into clipping.
I used to use the Ucd in my personal system quite happily until an colleague brought along a 300B DHT SE amp for me to try . Now I use the Ucd for Bass/Lower mid and the SE for the rest. Works superbly for me.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.