Finally finished build, tidied up wiring and sat down to do more listening;
Source is an Arcam Solo Neo (bought before upgrade itis set in) and speakers are PMC fb1s. Home made RCAs to XLRs using Gotham GAC3 cable. I originally decided to build the ncores because I needed a more powerful amp to drive the PMCs. I had around £1k to spend and decided that these would be way better than any ready built consumer power amp at that price point. And they are!!
Theres probably not much point in my going into lots of detail about the improvements to SQ given the limiting factor of my front end. Everybody else has already been through the thesaurus and used up all the adjectives but these amps really are good. I always thought that my speakers had more to give and my god they do. Transmission line designs are said to have a transparent midrange given the control afforded the driver by the mass of air inside the tube.
Suffice to say, bass is now tight and controlled, mids are sooo detailed and clear, highs are crisp and with no harshness. Soundstage is wide and deep (listening to Sonny Rollins - Saxophone Collosus as I type, sounds absolutely amazing)
Next upgrade will be separate preamp/DAC (thinking Benchmark HDR???), using 24/48khz Toslink output of CD/Fm/DAB/IPOD from the Arcam. Any news of Hypex modules? Any other suggestions would appreciated.
Only concern with Ncores is about heat. Asked about this a couple of weeks back - only one reply. Even on standby, these units get hot - far more so than the class AB Arcam. Are they safe to leave switched on?? All my gear goes inside a cabinet to keep out of sight of the nippers (and wife) and I havent dared to close the door in case they overheat. Am I being overly cautious or is there nothing to worry about??
Thanks
Source is an Arcam Solo Neo (bought before upgrade itis set in) and speakers are PMC fb1s. Home made RCAs to XLRs using Gotham GAC3 cable. I originally decided to build the ncores because I needed a more powerful amp to drive the PMCs. I had around £1k to spend and decided that these would be way better than any ready built consumer power amp at that price point. And they are!!
Theres probably not much point in my going into lots of detail about the improvements to SQ given the limiting factor of my front end. Everybody else has already been through the thesaurus and used up all the adjectives but these amps really are good. I always thought that my speakers had more to give and my god they do. Transmission line designs are said to have a transparent midrange given the control afforded the driver by the mass of air inside the tube.
Suffice to say, bass is now tight and controlled, mids are sooo detailed and clear, highs are crisp and with no harshness. Soundstage is wide and deep (listening to Sonny Rollins - Saxophone Collosus as I type, sounds absolutely amazing)
Next upgrade will be separate preamp/DAC (thinking Benchmark HDR???), using 24/48khz Toslink output of CD/Fm/DAB/IPOD from the Arcam. Any news of Hypex modules? Any other suggestions would appreciated.
Only concern with Ncores is about heat. Asked about this a couple of weeks back - only one reply. Even on standby, these units get hot - far more so than the class AB Arcam. Are they safe to leave switched on?? All my gear goes inside a cabinet to keep out of sight of the nippers (and wife) and I havent dared to close the door in case they overheat. Am I being overly cautious or is there nothing to worry about??
Thanks
Attachments
![IMG_0677[1].jpg](/community/data/attachments/259/259499-64c5769d477555123aee105d5c860eb0.jpg)
![IMG_0675[1].jpg](/community/data/attachments/259/259534-aa412a98e4d28460eb3abf92979034bd.jpg)
![IMG_0676[1].jpg](/community/data/attachments/259/259564-86b4b0e3e500aa364fbe3c7d6bb24dc7.jpg)
IMO this is something to consider.
Heat is the enemy of some parts, electrolytic capacitors especially.
When checking the NC400 and SMPS600 datasheets total dissipation at idle should be some 10 watts (one amp), and it seems to get hotter.
More passive cooling than normally required for class D amps might be required?
Maybe Bruno can comment.
He mentioned it here: http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-149.html#post2856720
and here http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-87.html#post2822328 and more places if you look hard enough. Bottome line, Bruno doesn't seemed concerned otherwise he would make it more of a point in the data sheet on online. I think so at least.
and here http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-87.html#post2822328 and more places if you look hard enough. Bottome line, Bruno doesn't seemed concerned otherwise he would make it more of a point in the data sheet on online. I think so at least.
Last edited:
He mentioned it here: http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-149.html#post2856720
and here http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore-87.html#post2822328 and more places if you look hard enough. Bottome line, Bruno doesn't seemed concerned otherwise he would make it more of a point in the data sheet on online. I think so at least.
I know, but user reports indicate that there is more dissipation than the datasheets suggest.
Maybe some parts (output coils) get hot?
Strange. I thought my units are actually cooler then I thought they would be. I could certainly do a temp rise test as I've done on my Class A amps but really don't think it's needed here.
How long does it take for your amp to heat up?
Maybe Bruno will chime in.
The output inductor definitely gets pretty warm, I think mine were around 150F but I'd need to measure again. The heat sinks on the SMPS600 are the only things on mine that get very warm but I can't measure them accurately with an IR temp gauge since they are not black and that is what the IR device is calibrated to. Now were did I put that themocouple probe???
There are electrolytics on the SMPS600 "covered" by the heatsinks; does not seem good for life expectancy, even if the caps are 105 degrees types.
I tend to agree and have decided my chassis are getting some ventilation holes.
On the upside, the caps are below the heat sink and hot air rises, provided of course there is somewhere for it to go...
For initial testing, I put my pairs of NC400s and SMPS600s into one cardboard box large enough to leave quite some space between all of them. I got hooked immediately to what those tiny beasts do to my makeshift OBs - they carry 4x MarkAudio Alpair 10.2 each, supposed to go into a Silbury one by one later this year. Wired in parallel the impedance dips down to 1.5ohm, hence my vintage Denon PMA 1560 was not so happy about my break-in setup...
Anyway, after one night's sleep spontaneously replaced with extended listening, I went on to check their local weather report. Deserts are supposed to be freezing at night, so that would not be the ideal analogy. The VDE would brandish me enemy number one if they could see the wiring job right there, so I didn't have the nerve to do the usual maintainance-without-interrupting-the-service thing I do to computer stuff as much as possible. After ten minutes of silence, they came to rest on the currently unused side plate of a steel PC case, with some narrow cardboard boxes as spacers to give the bottom side some air.
Much better now, but still, IMHO the SMPS should have its heat sinks mounted such that the upper surfaces are located in one plain - no point in turning them upside down now if they won't go full contact with the steel plate beneath anyhow, and I'd rather not loosen the screws and start bending stuff. The large chunk of aluminium underneath the amp is a rather nice heat spreader though, and works quite well through gravitational contact alone. No thermal grease messing up the place so far, but I am seriously considering compressible adhesive thermal pads for the SMPS to bind the heat sink surface to the top of the future case.
Anyway, after one night's sleep spontaneously replaced with extended listening, I went on to check their local weather report. Deserts are supposed to be freezing at night, so that would not be the ideal analogy. The VDE would brandish me enemy number one if they could see the wiring job right there, so I didn't have the nerve to do the usual maintainance-without-interrupting-the-service thing I do to computer stuff as much as possible. After ten minutes of silence, they came to rest on the currently unused side plate of a steel PC case, with some narrow cardboard boxes as spacers to give the bottom side some air.
Much better now, but still, IMHO the SMPS should have its heat sinks mounted such that the upper surfaces are located in one plain - no point in turning them upside down now if they won't go full contact with the steel plate beneath anyhow, and I'd rather not loosen the screws and start bending stuff. The large chunk of aluminium underneath the amp is a rather nice heat spreader though, and works quite well through gravitational contact alone. No thermal grease messing up the place so far, but I am seriously considering compressible adhesive thermal pads for the SMPS to bind the heat sink surface to the top of the future case.
A couple of questions and I apologize for asking. I know they have been answered somewhere here, but I have searched without finding the exact answer I am looking for and the thread is a beast already.
!. Does twisting all of the wiring (I know the speaker output wire should be twisted) really offer any benefit?
2. Pros and cons to using thermal grease under the NCore heatsink?
!. Does twisting all of the wiring (I know the speaker output wire should be twisted) really offer any benefit?
2. Pros and cons to using thermal grease under the NCore heatsink?
Any bunch of wires that is supposed to pick up interference as equally as possible (in order to be rejected as a common mode : ) should be tightly twisted IMHO. NCore has an excellent CMRR, so builders are invited to make use of it. I did not compare, and my current setup is temporary at best, but I twisted the power cords for improved handling / mechanical properties alone. Lacking any means of fixation, the opposite holds true for the smps/ncore interconnect harness. Ooh, let's twist again // once the case is delivered...
Applying just a little bit should improve heat transfer by literally filling the gap. The only possible negative aspect might be the mess it leaves, but nothing a cotton cloth and a few drops of alcohol won't cure. Once you're sure about the spot the amps will end up, with holes drilled, ready to kindly tighten each screw a half-turn at a time, feel free to dispense a few tiny drops of thermal grease. Depending on its viscosity, you might need to spread the stuff with a razor blade.
For the SMPS I suspect some 0.5mm pad will be more suitable, depending on the heat sink's angle maybe even thicker material. Will keep my eyes peeled for some stuff with the highest possible heat transfer rate - at a glance I could find 5 and 7 W/mK, 0.5 or 1mm, coming in stripes of 120x20mm - one of those can be cut to suit the two heat sinks on a single SMPS600 pretty neatly.
Sure it gets hot, it's an amplifier. 30 Watts inside a closed case, even an aluminum one, will get plenty hot. Just because ncores are superbly efficient there is no good reason to put them in unventilated cases. People here have been using cases originally designed for line-level components; the cases should be drilled.
I'm even thinking about using the top and bottom plate as planar heat sink for SMPS and NCore resp., and perhaps even place the cases vertically. Some feet for spacing below, and a few inches of air above, combined with drilled "side" panels, and the improved chimney effect should do the job.
Granted, that will not quite fit into any ordinary rack, and even my Double Flexy might have issues, so I'm still at odds with my current furniture. Anyone who is going to put them on top of some sideboard or whatever surface with enough vertical clearance should not have any trouble though. Might even look cute with two vertical mono blocks lining a horizontal preamp.
To each his own - using these splendid amps and smps, putting them in a shiny case and adding some nifty featuers is most of what remains of the DIY aspect after all. Folks keep saying I worry too much, so anyway...
Even a closed case works a bit as a heat sink, though that effect can be improved one way or another.
Thinking out loud when it comes to possibly constructive ideas, just leaving my 0.02€ - take it or leave it, no hard feelings either way
Depends. IIRC for safety class 2 (chassis not connected to mains safety earth) there needs to be a safety distance of at least 6mm between parts conducting mains voltage and the case, or some insulation material at least 0.5mm in thickness. If the chassis is safety-grounded anyway, that would be safety class 1 and there should not be a problem. It's not so much about regulations than the lives of yourself and your loved ones...
I touched the heat sink surface with my bare hands already to roughly check on temperature, so if that would carry mains voltage under normal operating conditions, I would have noticed
If you stick your SMPS heat sinks to the chassis with some conductive thermal grease (think Arctic Silver et.al.) and perhaps lock them in place with a screw or two, any mains-carrying conductors on the loose should trigger the breaker as soon as they touch the heat sink, read: thoroughly wire the chassis to safety earth with a mechanically stable, low-resistance connection.
Depends. There are various kinds of thermal pads around, and the ones I have in mind are 1mm thick and non-conductive.
There are thermal adhesives available in a two-component form (not so sure about the terminology - two substances, only the mix will cure...) with a high silver content, but still, not electrically conductive. Safe distance is still an issue though.
In order to remain within the constraints of safety class 2 though, the pad would need to extend a tad bit over the heat sinks' surface areas, overlapping some other insulation material where the heat transfer properties don't matter that much.
You would rather not want the mere insulation to introduce gaps between pad, heat sink and case, that's counterproductive; if the pad is larger than the area to tightly bind, gaps a few mm away should be less tragic.
That safety insulation should probably cover an area slightly larger than the mains-carrying part of the smps. 6mm distance from mains voltage to case, or 0.5mm of insulation inbetween. This posting sucks because it always repeats itself, repeats itself, repeats itself... Of course, one can always go for a thermal pad the size of the smps plus 12mm in each dimension... not very cost efficient in case of material with a high heat transfer rate. The stuff I have in mind is meant for binding watercooling heat sinks to overclocked RAM sticks and comes at a whopping 7W/mK.
I will probably go for safety class 1, doing the cleanly balanced wiring setup and all - case and cable shields go straight to mains safety earth, and in case an unbalanced source is connected, "signal ground" / "signal return" is equal to smps / amp ground. The temporary setup works quite well that way, except there is no case, no shield, and no balanced input so far, but the signal gnd to smps gnd works just fine.
- Status
- Not open for further replies.