I've been feeling brave and though about testing to turn the NCore into a current amp as described here https://docs.google.com/viewer?url=http%3A%2F%2Fwww.essex.ac.uk%2Fcsee%2Fresearch%2Faudio_lab%2Fmalcolmspubdocs%2FJ12%2520Distortion%2520reduction%2520MC%2520current%2520drive.pdf&docid=2075c0eb4fcc61972947f0e57b6c0e84&a=bi&pagenumber=16&w=766
in fig 23. And to measure success or failure I'd take a pair of cheap speakers and then compare distortion sweep between the current one and the normal voltage NCore.
The question I have first though is if I could damage the NCore by connecting it as in fig 23, or if the amp itself is safe and I only have to worry about my sacrificial speaker?
Secondly I've wondered about how to connect all the stuff in practice, say where am I to connect the unbalanced input ground?
in fig 23. And to measure success or failure I'd take a pair of cheap speakers and then compare distortion sweep between the current one and the normal voltage NCore.
The question I have first though is if I could damage the NCore by connecting it as in fig 23, or if the amp itself is safe and I only have to worry about my sacrificial speaker?
Secondly I've wondered about how to connect all the stuff in practice, say where am I to connect the unbalanced input ground?
Last edited:
Come on, just because Sweden won the Eurovision Song Contest...
I think you'll be OK.
To the "ground" end of Rf, but not sure if it needs to be connected anywhere else.
Ah, nice.
Read some more here: Amplifier topologies for current-drive - Current-Drive - The Natural Way of Loudspeaker Operation
The author there recommends a parallell RC network to ensure stability at high frequencies, do you think it's a good idea to include one in the case with the NCore or is it unnecessary for some reason?
Read some more here: Amplifier topologies for current-drive - Current-Drive - The Natural Way of Loudspeaker Operation
The author there recommends a parallell RC network to ensure stability at high frequencies, do you think it's a good idea to include one in the case with the NCore or is it unnecessary for some reason?
OllBoll
Those circuits only work as intended/according to the maths given, if you have a high (open) loop gain
The 26dB (voltage amplifier closed loop) gain of the ncore 'out of the box' isn't really high enough for the current feedback loop to work properly, and really the ncore's loop needs to be opened before you apply current feedback - at which point the compensation components may well be required
By all means try it and see but don't use it as a conclusion about what can be achieved with current feedback amps, or what an ncore transconductance amp is capable of achieving
Those circuits only work as intended/according to the maths given, if you have a high (open) loop gain
The 26dB (voltage amplifier closed loop) gain of the ncore 'out of the box' isn't really high enough for the current feedback loop to work properly, and really the ncore's loop needs to be opened before you apply current feedback - at which point the compensation components may well be required
By all means try it and see but don't use it as a conclusion about what can be achieved with current feedback amps, or what an ncore transconductance amp is capable of achieving
What if you say put a preamplifier circuit on the negative input to increase gain? Would it be a good idea or would there just be lots of new problems created?
Or in other words, what would be the easiest way to make the ncore enough of a transconductance amp to reap most of the benefits
Recently people seem to be talking about the heat produced by nc400s and smps600s. I have read this entire thread over the past couple of months (must have too much spare time), and a recurring point seems to be how cool these class D devices operate.
I took the plunge for the wk 18 batch and have just completed a pair of monoblocks, installed in the ubiquitous Italian Galaxy Maggiorato aluminium cases (all ally, 2mm top, bottom and rear, 10mm fronts and sides), and these things get hot. Compared to the 50w class AB Arcam amp that these are replacing, they get much hotter even when left just on standby.
Please excuse my ignorance (I am a skilled DIYer but a relative electronics novice), but is this normal and more importantly acceptable? I find it makes my life easier to keep all my gear hidden away within a built in cabinet, so ventilation is obviously not so good. That said it has never been a problem with the Arcam.
Are these units safe to leave switched on permanently? Sorry to ask but there are plenty of reports of wallwarts and other such consumer electronics goods going up in flames. I assume these modules have built in thermal trips of some kind but I need to know that the heat produced is within normal parameters rather than being down to a pair of "Friday" amps or dodgy design on my part. I used heatsink compound beneath the ncores during build up but I am now wondering whether to add heatsinks to the enclosures.
Sound quality through my PMC speakers is superb by the way, and a clear upgrade from the Arcam - not suprising really! When I have done some more listening, I will post a more in depth review.
On a separate note, can anyone suggest a good DIY dac/preamp to go with the ncore? I understand there may be a Hypex unit in the pipeline. If so does anyone have anymore info?
Thanks in advance
I took the plunge for the wk 18 batch and have just completed a pair of monoblocks, installed in the ubiquitous Italian Galaxy Maggiorato aluminium cases (all ally, 2mm top, bottom and rear, 10mm fronts and sides), and these things get hot. Compared to the 50w class AB Arcam amp that these are replacing, they get much hotter even when left just on standby.
Please excuse my ignorance (I am a skilled DIYer but a relative electronics novice), but is this normal and more importantly acceptable? I find it makes my life easier to keep all my gear hidden away within a built in cabinet, so ventilation is obviously not so good. That said it has never been a problem with the Arcam.
Are these units safe to leave switched on permanently? Sorry to ask but there are plenty of reports of wallwarts and other such consumer electronics goods going up in flames. I assume these modules have built in thermal trips of some kind but I need to know that the heat produced is within normal parameters rather than being down to a pair of "Friday" amps or dodgy design on my part.
I used heatsink compound beneath the ncores during build up but I am now wondering whether to add heatsinks to the enclosures. Sound quality through my PMC speakers is superb by the way, and a clear upgrade from the Arcam - not suprising really! When I have done some more listening, I will post a more in depth review.
On a separate note, can anyone suggest a good DIY dac/preamp to go with the ncore? I understand there may be a Hypex unit in the pipeline. If so does anyone have anymore info?
Thanks in advance
I built mono blocks using two W x H x L = 215mm x 70mm x 228mm aluminum cases. There are no vents in the cases, they are completely closed. Front is 10mm, sides and back 4mm, top and bottom 3mm. I've had them running since about 3pm yesterday at low volume. Not soft enough to carry on a conversation, but not at my usual listeng levels. I have checked the case temp at various times and the room temp at the same time. The cases seem to run about 21-22F higher than the room temp. Last night before I went to bed the room had gotten to about 80F. The cases measured 102F. This morning the room had cooled to about 67F. Cases were 88F. This afternoon, room temp 73F, cases 95F.
So far, very consistent. The mono blocks have measured within 1 degree of each other every time.
So far, very consistent. The mono blocks have measured within 1 degree of each other every time.
What if you say put a preamplifier circuit on the negative input to increase gain? Would it be a good idea or would there just be lots of new problems created?
Or in other words, what would be the easiest way to make the ncore enough of a transconductance amp to reap most of the benefits
No, it's the open loop gain of the amplifier that matters - how much gain you have in the feedback loop before you apply feedback. You won't be applying feedback from the output of the amplifier to the input of the preamp, so preamp gain is irrelevant
Have a look at some OP amp theory - the concept is that the open loop gain is infinite, so that when you apply negative feedback your errors are infinitesimally small. Reality is slightly different
but, within some practical constraints, the higher your open loop gain, the lower your overall distortionHave a look at the maths in the article I think I pointed pos at. Have a look at the maths next to fig 23 in the article you quoted - look at the importance of the value of A, the open loop gain of the amp
You need to open the loop of the ncore, and no one has yet analysed the circuit (as far as I am aware) to know where this would be within the circuit
No, it's the open loop gain of the amplifier that matters - how much gain you have in the feedback loop before you apply feedback. You won't be applying feedback from the output of the amplifier to the input of the preamp, so preamp gain is irrelevant
Have a look at some OP amp theory - the concept is that the open loop gain is infinite, so that when you apply negative feedback your errors are infinitesimally small. Reality is slightly different
Have a look at the maths in the article I think I pointed pos at. Have a look at the maths next to fig 23 in the article you quoted - look at the importance of the value of A, the open loop gain of the amp
You need to open the loop of the ncore, and no one has yet analysed the circuit (as far as I am aware) to know where this would be within the circuit
Ah, then it seems a bit more complicated than I would have liked =) I'm not sure I want to fiddle with the actual NCore circuit...
Though if I understood correctly the formula for output impedance was: (1 + amp gain)*Rf )
If amp gain is low then there isn't much to be done, if I understood correctly amp gain in the formula was 26 db, or 20x ( or is it open loop gain of 56 db here?). So with Rf = 1 Ω output impedance is:
(1+20)*1 = 21.
But if I've understood correctly it is also possible to raise output impedance then by increasing Rf, though you waste output power doing it.
Say I have a total load of 16 Ω and then set Rf to 16 Ω. That would net 336 Ω output impedance and still only waste 50% output power. This is for dedicated midrange and tweeters so power wastage shouldn't be that much of an issue, as long as it doesn't waste it when idle.
The real question though is if this would approximate a transconductance amp enough to yield significantly less distortion in the speaker or if it's just a handicapped voltage source in disguise.
Last edited:
I have had very good reults in distortion reduction with a 44 ohms output impedance on a 16 ohms driver:
http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-403.html#post3026510
even 22 ohms already gives good results
http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-403.html#post3026510
even 22 ohms already gives good results
I look forward to that, I too have PMC's but have yet to take the ncore plunge...
I would recommend the AMB audio Gamma 2 DAC. If you're good at soldering you can build one yourself if you get/have got all the right tools. It's extremely sensitive to the PSU used so a bit of trial and error might be in order. You'll need to pair it with some kind of pre-amp for the best results. Sound quality - it's almost as good as if not on a par with my Primare SP32 (£3-4K of DAC/Processor). Once I've got round to making a proper PSU, I think it may well sound better. All subjective of course but my opinions are shared with 2 other unbiased individuals
Brought them to life this morning - nice
is this case from ebay? i think i have the same ones, nice and tight fit with these cases
Yep it is - its a very tight fit but seems to operate well and stay ok heat wise. The speaker terminals are the correct spacing for Eichmann cable pods if anyone was interested, I will be changing mine when I get the motivation.
Thanks for your input Kjartan
Some years ago I build some LC Audio 2.3 dual mono Class D amps, which had an unbelievable control in the bottom, but sounded a little bit harsh in the upper mids. But still I liked them.
Later I build a NewClassD amp, with the high end discrete opamps in the input stage. That one sounded much better, and I think I would have kept it, if it wasnt for a not too well designed crossover in my ScanSpeak based speakers.
After that I had a Bewitch 6550 tube amp, which sounded good but had too little power to drive my speakers.
Right now I use a Musical Fidelity A3 for the side systems (1", 6 1/2" and 10" SS), and a Hypex 4.0 for the bottom (12" SS).
It sounds quite good, but I suppose youre allways on the hunt for something better ;-)
The Hypex substituted another plateamp, and gave a substantial better control.
Thanks
Arthur.
Arthur, I have a MF A3 right now, which I like, but I am thinking of either an ncore, ucd400, or LC audio 2.3SE. Any advice?
Brought them to life this morning - nice
What is this red thingy on your power cord (I presume this is the power cord) ?
I guess it is an IEC plug like this one on ebay: Oyaide C-079 IEC-Buchse C 079 Oyaide C-079 IEC-Buchse Netz Kaltgetäte Buchse C 079 / Original Oyaide | eBay
Brought them to life this morning - nice
On the power supply, why is the yellow bit wrapped in yellow paper? Did it come this way? Mine isn't wrapped. Is this a 2.0 upgrade?
That is right. It's an upgrade. Mine has it too so I asked Hypex. It is extra isolation.
- Status
- Not open for further replies.