I not sure I will be able to afford 6 NC400 for my Orions
Bruno,
I am too late to influence the power rating decision and in fact I agree that we should have available the highest power possible if there is only one DIY amp.
But I am currently using 2 UcD180HG with HxR and 1 UcD400HG per channel on my Orions. I am concerned that I won't be able to afford to buy 6 NC400 modules to replace these. In fact 8 modules are best because I feel that having the two woofers in parallel don't sound as good, but maybe that won't be the case with NCore.
Biwiring will not help in this case because of the use of an electronic crossover/equalizer/signal-shaper designed by Siegfried Linkwitz. I keep hoping to hear interesting echos from a meeting between him and you if you ever bumped into him at a AES convention - but this hypothetical meeting hasn't happened yet.
I am currently using a linear power supply for each module. Would it be sonically equivalent to use a single Hypex switching supply for three modules, considering that the total power per speaker does not need to exceed 400w?
Considering that I feel the future of Loudspeakers is to get rid of passive crossovers, and the future of Amplification is NCore, there is a mismatch in power/cost. Unless of course the NCore technology is not required when there is an amplifier for each driver.
Guy
Bruno,
I am too late to influence the power rating decision and in fact I agree that we should have available the highest power possible if there is only one DIY amp.
But I am currently using 2 UcD180HG with HxR and 1 UcD400HG per channel on my Orions. I am concerned that I won't be able to afford to buy 6 NC400 modules to replace these. In fact 8 modules are best because I feel that having the two woofers in parallel don't sound as good, but maybe that won't be the case with NCore.
Biwiring will not help in this case because of the use of an electronic crossover/equalizer/signal-shaper designed by Siegfried Linkwitz. I keep hoping to hear interesting echos from a meeting between him and you if you ever bumped into him at a AES convention - but this hypothetical meeting hasn't happened yet.
I am currently using a linear power supply for each module. Would it be sonically equivalent to use a single Hypex switching supply for three modules, considering that the total power per speaker does not need to exceed 400w?
Considering that I feel the future of Loudspeakers is to get rid of passive crossovers, and the future of Amplification is NCore, there is a mismatch in power/cost. Unless of course the NCore technology is not required when there is an amplifier for each driver.
Guy
Last edited:
Wouldn't it make sense for you to realize that what's driving your amplifiers will have a far greater sonic impact? I.e. the large collection of opamps used in the Orion's filter. A chain is really only as strong as its weakest link, which is perfectly applicable to audio as well. I think it would be more worthwhile to suggest to Mr. Linkwitz to design a properly implemented DSP solution for the Orion, one that processes everything in the digital domain. If you've heard the Grimm LS1 you'll realize analog filtering/processing really is a relic of the Mesozoic Era.
"A chain is really only as strong as its weakest link"
Learn to evaluate the weakness of each link properly.
Loudspeakers and air are an order of magnitude weaker than a dozen op-amps.
For example, direct radiating drivers and air absolutely can't go below 0.1% THD, rising to >1% when playing loud, and reflected sound in rooms and within speaker drivers and enclosures give SNR far worse than 40dB.
The advantage of DSP is to be able to stack and adjust complex filters on the fly, without changing component values or making new PCBs.
Learn to evaluate the weakness of each link properly.
Loudspeakers and air are an order of magnitude weaker than a dozen op-amps.
For example, direct radiating drivers and air absolutely can't go below 0.1% THD, rising to >1% when playing loud, and reflected sound in rooms and within speaker drivers and enclosures give SNR far worse than 40dB.
The advantage of DSP is to be able to stack and adjust complex filters on the fly, without changing component values or making new PCBs.
Indeed wise words.
It depends what measurements are chosen, how the measurements are made and what the opamps are built into. As well as which opamps
Why would reflected sound be considered 'N' rather than 'S' ? Also why choose THD when evaluating the sound quality of an audio signal chain?
Yep, big advantage. But to do this on the fly, DSP filters need de-zippering and that's a considerable overhead which many don't provide for. Without de-zippering, the real-time adjustment of DSP filters gives positively nasty zipper noise. Analog filters don't suffer from such unpleasant side effects when adjusted.
I was of the assumption that the moving parts (e.g. the drivers) and the room acoustics were the weakest link, but I'm never to old to learn.
I'm saving to build the Orions after I've auditioned a lot (>40) so called "high end systems"; wilson, focal, quad, to name a few. To my ears the Orions are among the most neutral sounding loudspeakers I've heard at any price point. However they only cost about 4500 euro to build, including all (hypex UcD180HG) amps.
You are one of the first to comment that the Orion speaker system is not well implemented.
If someone thinks that the opamps in the Orion ASP degrade the speakers sonic performance, then there are a number of approaches for trying to prove that. All involve that the terminal voltage at each driver remains unchanged in magnitude and phase over the audible frequency range. If they change by more than 0.1 dB, then the speaker can sound different.
1 Change the opamps. Note that the circuit impedance levels affect noise and stability.
2 Replace opamps with discrete devices. Note that the filter frequency response can change.
3 Use high level passive filters after the power amps. Good luck with the driver impedance interaction and non-linear distortion.
4 Use line level passive filter networks. Good luck with your inductors.
5 Use DSP without changing the phase response relative to the Orion's filter. Not for the uninitiated.
I hope you have had the opportunity to hear the Orions; if not, you are more than welcome to hear them when they are finished.
Furthermore, I am really interested in you implementing a DSP filter in it.
If you manage to maintain the terminal voltages and to find something worthwhile in this exercise, then I would surely like to know about it.
Maybe this would become even better than the current version and you get the credit for it.
I don't say it cannot be done, but not for the inexperienced.
I didn't have the opportunity to hear the Grimm Ls1, but I'd very much like to.
Even though I cannot afford them at 10.000 euro, without subs in order to make them full-range.
what diffirence between ucd and abletec
hello everybody i am brand new at this forum but
i have read this thread from the begining and i am very interested in the new
ncore modul.
just like tymen (that i totally agree with about dsp vs. asp) i am about to build
a linkwitz speaker, the speaker i planing to build is the pheonix.
this is were ncore comes in, i need 6 or maybe 8 channels to drive the loudspeakers.
the only thing that make me a sceptic to the class d module in general
is that i and some friends of mine have tested the abeltec modul.
the amp has it`s points with wery good focus and really good dynamics
both micro and macro.
but image is kind of flat and it sounds rather thin compared to
mark levinson 335 and ayre v-1
the thing i am wondering about is, are the ucd/ncore as bright
as the abletec
have anyone compared the ucd to abletec? i really like to know the diffirence
between those two
thanks for a really good site
hello everybody i am brand new at this forum but
i have read this thread from the begining and i am very interested in the new
ncore modul.
just like tymen (that i totally agree with about dsp vs. asp) i am about to build
a linkwitz speaker, the speaker i planing to build is the pheonix.
this is were ncore comes in, i need 6 or maybe 8 channels to drive the loudspeakers.
the only thing that make me a sceptic to the class d module in general
is that i and some friends of mine have tested the abeltec modul.
the amp has it`s points with wery good focus and really good dynamics
both micro and macro.
but image is kind of flat and it sounds rather thin compared to
mark levinson 335 and ayre v-1
the thing i am wondering about is, are the ucd/ncore as bright
as the abletec
have anyone compared the ucd to abletec? i really like to know the diffirence
between those two
thanks for a really good site
DSP for Linkwitz Orion
As others have pointed, the difficult part is to do at least as good a job as the current analog design. If successful it would be wonderful to be able to change values without looking for discontinued non-surface mount capacitors, how does the DIY community deal with that?
From what I gather Siegfried Linkwitz is not against a DSP implementation but unless he had validated it and makes it official he would not agree that an unofficial implementation be called an Orion. Also an Orion can be DIY as long as you buy the construction plans, it is not in the public domain.
It seems that attempts at implementing it have been challenging. I don't think SL will do it himself as he is happy with the analog solution. It would take somebody like Bruno to do it right, but why would he do that when he has done a great job with the Grimm LS1? The LS1 is $15000 in the U.S. and yes I would love to hear it. Will it be at the RMAF in Denver this year?
If someone comes up with a DSP implementation, I sure would like to try that. As a computer scientist major and physics minor, my electronic skills are sufficient to understand the design of a linear power supply, but I don't have enough signal processing background to fully understand the mathematics and physics to undertake this myself.
Guy
As others have pointed, the difficult part is to do at least as good a job as the current analog design. If successful it would be wonderful to be able to change values without looking for discontinued non-surface mount capacitors, how does the DIY community deal with that?
From what I gather Siegfried Linkwitz is not against a DSP implementation but unless he had validated it and makes it official he would not agree that an unofficial implementation be called an Orion. Also an Orion can be DIY as long as you buy the construction plans, it is not in the public domain.
It seems that attempts at implementing it have been challenging. I don't think SL will do it himself as he is happy with the analog solution. It would take somebody like Bruno to do it right, but why would he do that when he has done a great job with the Grimm LS1? The LS1 is $15000 in the U.S. and yes I would love to hear it. Will it be at the RMAF in Denver this year?
If someone comes up with a DSP implementation, I sure would like to try that. As a computer scientist major and physics minor, my electronic skills are sufficient to understand the design of a linear power supply, but I don't have enough signal processing background to fully understand the mathematics and physics to undertake this myself.
Guy
It doesn't need to be a single individual coming up with a DSP implementation - how about a team effort? Seeing there are so many blocks in the overall design if interested people divided them up amongst themselves, then a digital solution could gradually come together. Kind of 'open source' DSP.
Here's an article which could help set the ball rolling:
Which filters are noisier – analog or digital? part 2
Anyone who knows how to use LTSpice can now play with digital filters using Kendall's delay-line model. Who's up for it?
Let's not hijack the thread
Let's not hijack the thread which is about NCore.
The team work to implement various Linkwitz design in DSP is already on-going here: http://www.diyaudio.com/forums/minidsp/174145-minidsp-linkwitz-orion-asp.html
And general Hypex DSP discussion is here: http://www.diyaudio.com/forums/class-d/144412-hypex-ucd-as2-100-a.html?postid=1833533#post1833533
Let's not hijack the thread which is about NCore.
The team work to implement various Linkwitz design in DSP is already on-going here: http://www.diyaudio.com/forums/minidsp/174145-minidsp-linkwitz-orion-asp.html
And general Hypex DSP discussion is here: http://www.diyaudio.com/forums/class-d/144412-hypex-ucd-as2-100-a.html?postid=1833533#post1833533
It doesn't need to be a single individual coming up with a DSP implementation - how about a team effort? Seeing there are so many blocks in the overall design if interested people divided them up amongst themselves, then a digital solution could gradually come together. Kind of 'open source' DSP.
Here's an article which could help set the ball rolling:
Which filters are noisier – analog or digital? part 2
Anyone who knows how to use LTSpice can now play with digital filters using Kendall's delay-line model. Who's up for it?
Yeah I will start a new thread in the Digital Line section to see if there's any interest in open source DSP. The MiniDSP is entirely proprietary in my understanding so not really relevant.
<edit> New thread here: http://www.diyaudio.com/forums/digital-line-level/195791-open-source-dsp-xos.html#post2694708
<edit> New thread here: http://www.diyaudio.com/forums/digital-line-level/195791-open-source-dsp-xos.html#post2694708
Last edited:
Linkwitz Orion with NCore amps
Hi Plutten,
just like you I'm waiting for the new long throw Seas woofers and ASP circuit boards to be available at the end of September.
Hopefully the DIY NCore amps will be ready within a few months. Any dates yet?
The Orion-NCore combo has the potential to be a really nice neutral, reference system.
If I can afford 6 to 8 channels of NC400 that is
... still way too early to start an Orion/NCore group buy
Hi Plutten,
just like you I'm waiting for the new long throw Seas woofers and ASP circuit boards to be available at the end of September.
Hopefully the DIY NCore amps will be ready within a few months. Any dates yet?
The Orion-NCore combo has the potential to be a really nice neutral, reference system.
If I can afford 6 to 8 channels of NC400 that is
... still way too early to start an Orion/NCore group buy
yes it´s to early start a group buy, but my question still remain
is ncore as bright as the abletec.
maybe Bruno can give me an answer on this question
when a designer says that his product is neutral it doesnt say me anything
every designer that i can think of says that there product is neutral and true to the
source, but they dont sound the same!
so what it comes down to is comparison between different amp, loudspeaker and so on.
is ncore as bright as the abletec.
maybe Bruno can give me an answer on this question
when a designer says that his product is neutral it doesnt say me anything
every designer that i can think of says that there product is neutral and true to the
source, but they dont sound the same!
so what it comes down to is comparison between different amp, loudspeaker and so on.
Plutten: I will be able to try the Ayre (from Mats) against the abletec's here in due time. Here the abletec's are not flat at all, I have at least a 5m depthcontrast if the recording allows for it. I can on the other hand agree that they can be a bit lean in the bassregion even if they have a very good dynamic drive.
I'm quoting Stig-Erik as I think that he might as well had writen about the Abletec's, and I'm also wondering if the ncore would be described as "lean/clean" or "powerfull" in the bassregion.
I'm quoting Stig-Erik as I think that he might as well had writen about the Abletec's, and I'm also wondering if the ncore would be described as "lean/clean" or "powerfull" in the bassregion.
When an amplifier strongly under/overemphasizes certain frequency ranges, it suggests that distortion isn't sufficiently independent of frequency. It could also point toward EMC problems or subtle flaws in the small signal path. For instance, you could have the choice between either building the modulator differentially or single-ended with a diff amp in front. You'd have to do tests like drive a lot of current through the ground plane to spot the difference (which is what the power stage does anyway) but measured as a black box the added distortion would not be so obvious. It takes serious fanaticism to keep an eye on all such supposedly negligible effects which in sum can amount to quite some otherwise unexplicable colouration. The trouble is that they're hard to discover by trial and error. You have to see them coming before they happen.
Hello,
I believe, I know what you mean.
really in temporal order? That seems to be tricky, indeed.
I believe, I know what you mean.
@tiki I see your point: you see a problem coming, you solve it and then it fails to come, falsifying your predictive ability
This is the usual problem with prophesies of doom which go away when action is taken. One shining example is the Y2K bug. This is now held up just about everywhere as an example of a spurious prediction of doom. Excuse me? It is the prime example of a prediction duly heeded. Thousands of programmers have sifted through billions of lines of code to change 2-digit year fields to 4-digit and to look for time difference calculations with a wraparound problem. The whole effort started years before 2000 and has worked splendidly. Sorry for the OT, but it's an old hobby horse of mine.
@Tengizk
Electrostats work fine. The only time you get into trouble is when you clip. The culprit is the step-up transformer. Usually the waveform isn't symmetrical so one side clips before the other. Chopping off positive peaks causes negative net DC and vice versa. This happens with any amp. The only thing that makes the problem more severe with UcD and Ncore is the low output impedance which makes the amp insist on driving that DC voltage into an inductance (which has near zero DC resistance). The transformer goes into saturation straight away and the amp protects.
Other amps can have the same problem but most will simply give in gracefully. It's the downside to trying to make output impedance as low as possible.
This is the usual problem with prophesies of doom which go away when action is taken. One shining example is the Y2K bug. This is now held up just about everywhere as an example of a spurious prediction of doom. Excuse me? It is the prime example of a prediction duly heeded. Thousands of programmers have sifted through billions of lines of code to change 2-digit year fields to 4-digit and to look for time difference calculations with a wraparound problem. The whole effort started years before 2000 and has worked splendidly. Sorry for the OT, but it's an old hobby horse of mine.
@Tengizk
Electrostats work fine. The only time you get into trouble is when you clip. The culprit is the step-up transformer. Usually the waveform isn't symmetrical so one side clips before the other. Chopping off positive peaks causes negative net DC and vice versa. This happens with any amp. The only thing that makes the problem more severe with UcD and Ncore is the low output impedance which makes the amp insist on driving that DC voltage into an inductance (which has near zero DC resistance). The transformer goes into saturation straight away and the amp protects.
Other amps can have the same problem but most will simply give in gracefully. It's the downside to trying to make output impedance as low as possible.
- Status
- Not open for further replies.