Hello all.
Yesterday I went to a press office and purchased the September'07 issue of ELEKTOR magazine. I was greatly surprised to see that there is a review of 11 amplifier modules. Among them, they compare Coldamp , Hypex and NewClassD and a number of other conventional designs.
Interesting read... although they don't publish any listening impressions, only bare measurements and some comments. (in fact they ask for feedback from readers that test them in ther audio equipment!)
Yesterday I went to a press office and purchased the September'07 issue of ELEKTOR magazine. I was greatly surprised to see that there is a review of 11 amplifier modules. Among them, they compare Coldamp , Hypex and NewClassD and a number of other conventional designs.
Interesting read... although they don't publish any listening impressions, only bare measurements and some comments. (in fact they ask for feedback from readers that test them in ther audio equipment!)
Basically none 
It is a rather aseptic study where they expose some of the main characteristic of each one and a table with some measurements of bandwidth, THD, etc.
They basically say that the three studied Class-D amplifiers are very similar in characteristics and measures, but don't provide any impression about their sound, in fact they haven't listened to all of them, from what I can read.
It is a rather aseptic study where they expose some of the main characteristic of each one and a table with some measurements of bandwidth, THD, etc.
They basically say that the three studied Class-D amplifiers are very similar in characteristics and measures, but don't provide any impression about their sound, in fact they haven't listened to all of them, from what I can read.
http://www.elektor.com/magazines.46742.lynkx i can't find that article Pierre, what matter? Or you told about differ magazine?
ps: sorry, here it is http://www.elektor.com/magazines/2007/september/good-and-occasionally-favourable.227957.lynkx
but, registration procedure very boring thing, maybe anybody post article here?
Blin, they want money for article, i'm very disappointment, damn!
Pierre, what matter? Or you told about differ magazine?ps: sorry, here it is http://www.elektor.com/magazines/2007/september/good-and-occasionally-favourable.227957.lynkx
but, registration procedure very boring thing, maybe anybody post article here?
Blin, they want money for article, i'm very disappointment, damn!
I just read the article (in German). It indeed seems very disappointing in the view of judging and giving "audiophile points". But it is a good overview about several modules, with different technology, and maybe for regular Elektor readers who are no rabid audiophiles probably also partly new.
I found it very interesting - especially the technical data comparision charts. You also must consider that they compared cheap kits in class-ab with chipamps and Class-D amps:
Interestingly the cheapest module (Velleman VM100, Class-ab discreet, 50Euro) has the best signal to noise ratio (100db) compared to the Coldamp BP4078 with mere 79db (120Euro). The Coldamp is much more powerful nevertheless...
The Coldamp has the worst intermodulation (4.5%) compared to the cheap Monacor IPA50 (0.14%, using a TDA7296).
The Hypex UCD400ST measured 226W (8Ohms), and did not even switch down without an extra cooling (the module is very small as you know).
etc etc, and no judgement.
Yes, I found it a very interesting read. And even without somebody telling me how good or bad the modules have sounded. I trust those individual statements very little anyhow.
I found it very interesting - especially the technical data comparision charts. You also must consider that they compared cheap kits in class-ab with chipamps and Class-D amps:
Interestingly the cheapest module (Velleman VM100, Class-ab discreet, 50Euro) has the best signal to noise ratio (100db) compared to the Coldamp BP4078 with mere 79db (120Euro). The Coldamp is much more powerful nevertheless...
The Coldamp has the worst intermodulation (4.5%) compared to the cheap Monacor IPA50 (0.14%, using a TDA7296).
The Hypex UCD400ST measured 226W (8Ohms), and did not even switch down without an extra cooling (the module is very small as you know).
etc etc, and no judgement.
Yes, I found it a very interesting read. And even without somebody telling me how good or bad the modules have sounded. I trust those individual statements very little anyhow.
They measured them as they were delivered, I suppose, which is definitely fair. They did not change feedback and amplifications rates, which maybe seem unfair in technical terms. They also use a very stabile power supply for testing.
Only the Hypex had an appropriate input sensitivity for professional balanced use with studio level (about 2V), most of other were to sensitive, one (Groundamp) was very insensitive. The Coldamp is 2db down already with 20kHz, starting this tilt at 5kHz, the NewClassD NCD1 had almost the same. The Hypex did not show this.
Only the Hypex had an appropriate input sensitivity for professional balanced use with studio level (about 2V), most of other were to sensitive, one (Groundamp) was very insensitive. The Coldamp is 2db down already with 20kHz, starting this tilt at 5kHz, the NewClassD NCD1 had almost the same. The Hypex did not show this.
Hello,
Some comments now that you have given some data about our Coldamp modules.
As you have pointed out, to be fair, all the measurements must be made with the modules as they are received, and all of them under the same measurement conditions.
Some hints...
The gain can be changed. We decided to use 45V/V gain (for each input, so 90V/V balanced) as this allows direct connection to typical unbalanced line sources, such as CD players. In a day or two we are about to open a new version of out www.coldamp.com website, where users can download more documentation. Casually, one of the application notes described how to easily change gain.
About the bandwidth, it fully covers the audio band. In fact, the 2dB drop at 20KHz (4ohm) is of little importance sonically. No customer has complained about that when they have heared the modules, as highs are perfectly balanced and clear. The -3dB point was set at 27KHz aprox as a compromise with ripple rejection, stability, etc. This can be easily increased if desired, too, so no problem. We see no use in setting the -3dB point at 50KHz, to be sincere, but that's a personal decision.
And about the rest of measurements, you know that getting accurate results while measuring Class-D is very difficult, and depend on the setup heavily. The IMD has appeared a bit high, although we have got much better measurements at our labs (below 1% in that conditions), as well as for the 1W THD (other measurements have shown below 0.2%). Proper wiring, power supply, etc. are the key to excellent figures.
In a nutshell, measurements have to be taken with great care, as they have a limited impact in sound quality and cannot be taken as a figure of merit alone. The proof for us is the satisfaction of our customers, some of them with very very critic ears as required by their jobs (studio recorders, etc).
We also take the opportunity to say that the Spanish edition of ELEKTOR has some errata, such as saying that our modules have pre-filter feedback. That's not true and was corrected, but they have inadvertently published the uncorrected version of the article. This edition also doesn't mention the fixed sw. frequency and output notch filter, as the rest of issues publish.
Sorry for the long reading...
Best regards,
Sergio
Some comments now that you have given some data about our Coldamp modules.
As you have pointed out, to be fair, all the measurements must be made with the modules as they are received, and all of them under the same measurement conditions.
Some hints...
The gain can be changed. We decided to use 45V/V gain (for each input, so 90V/V balanced) as this allows direct connection to typical unbalanced line sources, such as CD players. In a day or two we are about to open a new version of out www.coldamp.com website, where users can download more documentation. Casually, one of the application notes described how to easily change gain.
About the bandwidth, it fully covers the audio band. In fact, the 2dB drop at 20KHz (4ohm) is of little importance sonically. No customer has complained about that when they have heared the modules, as highs are perfectly balanced and clear. The -3dB point was set at 27KHz aprox as a compromise with ripple rejection, stability, etc. This can be easily increased if desired, too, so no problem. We see no use in setting the -3dB point at 50KHz, to be sincere, but that's a personal decision.
And about the rest of measurements, you know that getting accurate results while measuring Class-D is very difficult, and depend on the setup heavily. The IMD has appeared a bit high, although we have got much better measurements at our labs (below 1% in that conditions), as well as for the 1W THD (other measurements have shown below 0.2%). Proper wiring, power supply, etc. are the key to excellent figures.
In a nutshell, measurements have to be taken with great care, as they have a limited impact in sound quality and cannot be taken as a figure of merit alone. The proof for us is the satisfaction of our customers, some of them with very very critic ears as required by their jobs (studio recorders, etc).
We also take the opportunity to say that the Spanish edition of ELEKTOR has some errata, such as saying that our modules have pre-filter feedback. That's not true and was corrected, but they have inadvertently published the uncorrected version of the article. This edition also doesn't mention the fixed sw. frequency and output notch filter, as the rest of issues publish.Sorry for the long reading...
Best regards,
Sergio
!- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.