Hi Ed
Thanks for this very positive review.
Incredible how cheap you got the speakers ....
I just want to ask you about the power supply for the 2.2SE. Did you mean that it used the original predator caps plus two extras in parallel? And all ChemiCon's?
Anyway i think it's a good idea to use a bigger transformer, compared to using bigger caps. In my experience you get a more relaxed mid and top when using an oversized transformer. Whereas the bass doesn't seem to be influenced, as one might expect.
Best regards from
Lars
Thanks for this very positive review.
Incredible how cheap you got the speakers ....I just want to ask you about the power supply for the 2.2SE. Did you mean that it used the original predator caps plus two extras in parallel? And all ChemiCon's?
Anyway i think it's a good idea to use a bigger transformer, compared to using bigger caps. In my experience you get a more relaxed mid and top when using an oversized transformer. Whereas the bass doesn't seem to be influenced, as one might expect.
Best regards from
Lars
http://www.icepower.bang-olufsen.com/sw2025.asp
0.0009% @ 5 kHz / 4 Ohms / 100 mW
Well, 0.0009% is what i call very low THD. But as people who have followed some of my discussions, knows, i do not subscribe to the pitfall that lower THD means better sound.
Not at all!
I do not intend to comment on the sound of the newest icepower modules, i have not heard them.
I should think they are good sounding, anyway i have very good confidence in the technical competences of Bang & Olufsen, i think the THD measurements are correct and good.
As you might know, i do not find THD to be a very reliable scale for sound quality. Not that i am discarding THD as useless, it is a good evaluation of the technical qualities of a circuit. But hunting low THD figures on expense of good sound quality, is a sport i have never embarked. I always found that activity best associated with the japanese audio industry in the 1980's. And we all know how that turned out.
Jan-Peters THD measurements are obviously supposed to show that THD of your amplifier is 10 to 30 times lower than my ZAPpulse modules. I guess most of the commercial (and all of the japanese
joke!) audio industry would agree that your amplifier is 10 to 30 times better than mine. However i also think that most of the high end industry would say: It doesn't matter! Other things influence the sound more!For example i could show you an english made triode amplifier with 7% THD that i bet could beat the socks off most low THD amplifiers in a second!
Of course it is very expensive and volatile, and only sounds good when the triodes are not past expiry date. But still, it shows the sound can be better even if the THD is (very) high.About the 'lively top' of the ZAPpulse amplifier. Actually it does not come from the THD at 5-7 kHz, like you implied earlier. It actually comes from the 4 - 13 dB peaking of the output filter at 150 kHz (Depending on the speaker load). In some speakers it helps to create that airy top, but in cases where you don't want this airy top, (like when your tweeter already has a very airy sound) we demonstate in our documentation how you can easily remove the 150 kHz peak, by lowering the fc to 120 kHz. But in many cases this makes the amplifier sound darkish. Which is preferable depends on the speakers and cables tonal balance.
Now of course nothing would be easier for us, than simply remove the filter peak in the feedback circuit once and for all, and thereby get a nicer looking datasheet. BUT at L C Audio find the sound quality to be of higher priority, and therefore left the option open to those who can benefit from it.
Lars:
Yes I bought one predator powersupply from you plus 4x 15,000uF chemicons. The Zappulse are mounted on opposite sides of the case because I found they modulated and had a whistle (hetrodyned) when placed close together (despite playing with all kinds of earthing designs). The predator with two caps is mounted very near the one module on the left, then the other caps are mounted on the other side of the case very close to the other module (and in parallel to the first caps).
This was the arrangemnt you advised me on and it seems to work well.
I'm curious what you consider to be the "correct" size for the transformers? I still have a pair of zappulse 2.2SE to rebuild with a better design. The first zappulse I built after consulting with you and you advised on a pair of your 200va transformers plus one predator with 10,000uF caps, all powering a stereo pair of zappulse 2.2se placed very close together....
The end result sounds pretty horrible and feels like it has far too much treble (which I assume is some sort of harmonic distortion). There is also a lot of hetrodyning.
The replacement zappulse design (which I described above) sounds HUGELY better. However, I'm not sure if it's the increase capacitance, increased distance between the modules or increased transformer size which is making the difference.
I'm interested in your comments because I have just ordered a pair of 500va transformers and some exotic caps and rectifiers to build a dual mono UCD400. Based on my experience there I want to rebuild the stereo Zappulse as well. However, keeping the 2x 200va transformers would be handy because they fit in my 1U chassis. Possibly though I should bite the bullet and get a bigger chassis and a much larger torroid...?
Punch line seems to be that if the cost is not an object then an "oversized" powersupply might not be a terrible idea even for these efficient Class D amps?
Yes I bought one predator powersupply from you plus 4x 15,000uF chemicons. The Zappulse are mounted on opposite sides of the case because I found they modulated and had a whistle (hetrodyned) when placed close together (despite playing with all kinds of earthing designs). The predator with two caps is mounted very near the one module on the left, then the other caps are mounted on the other side of the case very close to the other module (and in parallel to the first caps).
This was the arrangemnt you advised me on and it seems to work well.
I'm curious what you consider to be the "correct" size for the transformers? I still have a pair of zappulse 2.2SE to rebuild with a better design. The first zappulse I built after consulting with you and you advised on a pair of your 200va transformers plus one predator with 10,000uF caps, all powering a stereo pair of zappulse 2.2se placed very close together....
The end result sounds pretty horrible and feels like it has far too much treble (which I assume is some sort of harmonic distortion). There is also a lot of hetrodyning.
The replacement zappulse design (which I described above) sounds HUGELY better. However, I'm not sure if it's the increase capacitance, increased distance between the modules or increased transformer size which is making the difference.
I'm interested in your comments because I have just ordered a pair of 500va transformers and some exotic caps and rectifiers to build a dual mono UCD400. Based on my experience there I want to rebuild the stereo Zappulse as well. However, keeping the 2x 200va transformers would be handy because they fit in my 1U chassis. Possibly though I should bite the bullet and get a bigger chassis and a much larger torroid...?
Punch line seems to be that if the cost is not an object then an "oversized" powersupply might not be a terrible idea even for these efficient Class D amps?
Ed: I think yes! Even though of course it is not 'necessary' with the oversized transformer, it does seem to give a better sound.
Experiences from several other customers who have the same result as you in the top, suggests that it is connected with the transformer size. Rather unlogical, but that's the word anyway.
I don't think moving the modules apart has any effect on the flow of the treble. Maybe on the heterodyning, but that would be very dependent of the wiring. If you use our suggested wiring (which i do myself on a regular basis) you will not have any heterodyning at all.
See here: http://www.lcaudio.dk/zpactionpackwiring.pdf
Experiences from several other customers who have the same result as you in the top, suggests that it is connected with the transformer size. Rather unlogical, but that's the word anyway.
I don't think moving the modules apart has any effect on the flow of the treble. Maybe on the heterodyning, but that would be very dependent of the wiring. If you use our suggested wiring (which i do myself on a regular basis) you will not have any heterodyning at all.
See here: http://www.lcaudio.dk/zpactionpackwiring.pdf
The only thing I don't have from your diagram is the signal ground tied directly to the mains and chassis ground. However, it whistles quite loudly.
In any case I think moving them further apart has other advantages.
Thanks for the tips
Ed
P.S. You avoided actually specifying a hard number that seems to be the limit beyond which bigger torroids are worthless... Is it a 1Kva for a stereo pair. A 2Kva...? What is too small?
In any case I think moving them further apart has other advantages.
Thanks for the tips
Ed
P.S. You avoided actually specifying a hard number that seems to be the limit beyond which bigger torroids are worthless... Is it a 1Kva for a stereo pair. A 2Kva...? What is too small?
That's at one hundred milliwatts. More complete graphs are available on the data sheet (link on same url), showing a THD of up to 0.07% at 10W, 7kHz.Lars Clausen said:
But hunting low THD figures on expense of good sound quality, is a sport i have never embarked. I always found that activity best associated with the japanese audio industry in the 1980's. And we all know how that turned out.
You seem to be suggesting that achieving low THD necessarily goes at the expense of sound quality. As you know, this depends of whether one relies on large amounts of feedback to do this, or on improving the open loop THD of the circuit. I understand that you prefer low feedback. So do I. It therefore deserves pointing out that UcD uses much less feedback than Zap, Ice or T. Its low THD is achieved by improving the open-loop distortion of the power stage.
Once again, you are fallaciously suggesting that high THD figures are a requirement to good sound.
It is well known that these amplifiers can sound pleasing, but far from correct. One may like the colouration, but it is colouration nevertheless. Such amplifiers can not be said to pass the signal unadulterated. The whole trick of amplifier design is to get the same sense of musicality and lucidity without the colouration.
Whatever you say is the cause of "air" in your amplifier, you will agree with me that this deviates from the normal function of an amplifier, and that this "air" is an added effect which is outside the legislature of amplification. I prefer to entrust the EQ'ing of my CD's to the mastering engineer, who can better make a decision whether HF boost is in order or not, depending on the recording.
On various occasions I have discussed precisely this euphonic peaking effect that is found in all class D amplifiers without output filter control, and which has helped popularise class D amplifiers in the high-end industry. The downside of this is that as listeners grow more experienced, they recognise the effect as a cheap trick, and grow tired of it. These people are then doubly delighted when they find another amplifier that simply provides the "pure and simple truth" (sonically).
So, I prefer to provide the user with an amplifier that always produces correct performance, relying on loudspeaker people to do their part of the job. Apart from that, I am also not going to enter into a competition with anybody by providing more HF boost, only to see the next guy add a few more dB's. Before we know it, the "bathtub curve" times are back.
Well, if you think having an HF boost is about "sound quality"...
Quality test
Hello,
When do we do the next test for the second time in the Netherlands, by speaker & co?
The only test here in the forum is THD, but the quality test of both modules is what I do is listening.
Make sure you have than the right quality of big torroid.
Rudy
Preamp: AVM
Amplifier: Zappulse 2.3SE monoblok 28 Kg:
Amplimo ringkern transformator 2000Va 2 x 42 Volt.
2 x 47.000µF 80 Volt type Mundorf M-Lytic HC.
DVD Player: Marantz DV17
CD Player: Denon DCD 3560 Full modification by Christian Feickert (Lc Audio.)
Speakers: ACR Isostatic RP-400
Hello,
When do we do the next test for the second time in the Netherlands, by speaker & co?
The only test here in the forum is THD, but the quality test of both modules is what I do is listening.
Make sure you have than the right quality of big torroid.
Rudy
Preamp: AVM
Amplifier: Zappulse 2.3SE monoblok 28 Kg:
Amplimo ringkern transformator 2000Va 2 x 42 Volt.
2 x 47.000µF 80 Volt type Mundorf M-Lytic HC.
DVD Player: Marantz DV17
CD Player: Denon DCD 3560 Full modification by Christian Feickert (Lc Audio.)
Speakers: ACR Isostatic RP-400
Sorry, for my interrupting.
Sorry, for my interrupting in the discussion for both modules.
You can only do the listening test of both modules, when the power supply the same is.
When there is a upgrade from the UCD modules I would also listening to the new module, and my Zap amplifier.
Greetings,
Rudy
Sorry, for my interrupting in the discussion for both modules.
You can only do the listening test of both modules, when the power supply the same is.
When there is a upgrade from the UCD modules I would also listening to the new module, and my Zap amplifier.
Greetings,
Rudy
Re: Sorry, for my interrupting.
Actually I already have an aluminium plate with 2 UcD400 and 2 Zap 2.2 modules with a common power supply using 2x10000uF Black Gate tanks (change modules by resoldering leads). If you're ok with this one, we can use this.
I agree on that. If you can provide a setup containing your preferred power supply (vcc<68V) we can swap the modules and take the power supply out of the equation.BMW850 said:
Actually I already have an aluminium plate with 2 UcD400 and 2 Zap 2.2 modules with a common power supply using 2x10000uF Black Gate tanks (change modules by resoldering leads). If you're ok with this one, we can use this.
On a Dutch forum Doede posted some pics of the (old) 2.2SE.
He claimed the effects of the switching frequency (as clearly visable in the pics) cannot be heared (much like in non-oversampling DACs)
http://www.htforum.nl/yabbse/index.php?topic=17055.msg204099#msg204099
He claimed the effects of the switching frequency (as clearly visable in the pics) cannot be heared (much like in non-oversampling DACs)
http://www.htforum.nl/yabbse/index.php?topic=17055.msg204099#msg204099
Jan Peter: Thanks for the measurements. Looks nice
The residual carrier signal is not a defect, it is present on all ZAP modules, but since it is located at around 450 kHz, you can not hear it, and it will not load your treble unit mentionable. (8 Ohms power dissipation is less than 200 mW).
We have done quite a lot of experimentation with different levels of output filtering, to give higher or lower residual carrier signal.
We have weighted a wide audio bandwidth against residual carrier level, and found a carrier level in the magnitude which is found on ZAPpulse to be absolutely ok.
So we designed the output filter to have a Fc of 150 kHz, giving the 12 dB / oct filter an attenuation of around 21,5 dB at 450 kHz. And a nice wide audio bandwidth!
However this means you will have around 3 Vrms residual carrier on the output (as can also be seen on Jan-Peters curves). There are thousands of ZAPpulse modules out there with the same level of carrier, and i have never heard of anybody who had a problem with the carrier signal, except then looking at the scope.
All in all a lower Fc output filter gives a nicer curve on the paper, as you can see on Jan-Peter's curve. But since it also limits bandwidth and relies on feedback to equalize the high frequencies, this means in my opinion, the 'technically nice looking' method works against the sound quality.
The residual carrier signal is not a defect, it is present on all ZAP modules, but since it is located at around 450 kHz, you can not hear it, and it will not load your treble unit mentionable. (8 Ohms power dissipation is less than 200 mW).
We have done quite a lot of experimentation with different levels of output filtering, to give higher or lower residual carrier signal.
We have weighted a wide audio bandwidth against residual carrier level, and found a carrier level in the magnitude which is found on ZAPpulse to be absolutely ok.
So we designed the output filter to have a Fc of 150 kHz, giving the 12 dB / oct filter an attenuation of around 21,5 dB at 450 kHz. And a nice wide audio bandwidth!
However this means you will have around 3 Vrms residual carrier on the output (as can also be seen on Jan-Peters curves). There are thousands of ZAPpulse modules out there with the same level of carrier, and i have never heard of anybody who had a problem with the carrier signal, except then looking at the scope.
All in all a lower Fc output filter gives a nicer curve on the paper, as you can see on Jan-Peter's curve. But since it also limits bandwidth and relies on feedback to equalize the high frequencies, this means in my opinion, the 'technically nice looking' method works against the sound quality.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.