I checked the temps on the SMPS large heat sink with the temp probe on my DMM and found it to be 20F cooler with the top off. Now, even though the temp was well within the limits of operation, I thought I would like some venting just to attempt to prolong life of electrolytic caps. I installed four aluminum standoffs to vent the top all around the edges.
Attachments
Do the same on the bottom, and you'll have a constant upwards air stream.
Chris,
Nice solution and easy to test the effect without drilling holes in my nice black case tops.
Tom
Yes, that's the whole point
The instructions in the data sheet are pretty clear on this.
The SMPS is class 2 (IIRC) and requires at least 6mm air space around the live areas - as indicated in the diagram in the data sheet. The 6mm spacing is inherently provided underneath by the standoffs
It's all designed to be class 2 so there's no need to fuss.
- bolt the amp to the case
- bolt the smps to the case using the standoffs - one of which is conductive for improved emi
- ensure 6mm clearance everywhere else
- connect as recommended in the data sheet (no mains earth)
- for a single module, don't worry about ventilation
- sit back and enjoy
Personally I think this is the one that matters now....and I'm doing as Bruno says. Compared to my ex Class A amps these things are very cool.
Hi,
I´m just in the process of building a pair of cases for my NCores and also encountered some problems with too much heat.
One amp plus smps produce around 14,5 watt of heat (measured the consumption at idle). I put them in 100 x 60 x 250mm cases (one for the amp, one for the power supply) with a little ventilation but they keep turning off after 2 or 3 hours of use with very efficient 12 ohm loudspeakers.
It´s the smps getting too hot.
Drilling some holes in the case (one 50mm on top, one at the bottom) helps to keep them from switching off but the little heatsinks still get quite hot.
Looking at the heatsinks it would be possible to extend them with a sheet of alumium (connecting both) to make them more efficient. The question is if this is possible?
I´ve connected my cases to safety earth.
William
I´m just in the process of building a pair of cases for my NCores and also encountered some problems with too much heat.
One amp plus smps produce around 14,5 watt of heat (measured the consumption at idle). I put them in 100 x 60 x 250mm cases (one for the amp, one for the power supply) with a little ventilation but they keep turning off after 2 or 3 hours of use with very efficient 12 ohm loudspeakers.
It´s the smps getting too hot.
Drilling some holes in the case (one 50mm on top, one at the bottom) helps to keep them from switching off but the little heatsinks still get quite hot.
Looking at the heatsinks it would be possible to extend them with a sheet of alumium (connecting both) to make them more efficient. The question is if this is possible?
I´ve connected my cases to safety earth.
William
I'm running a single NC400 + SMPS600 in a closed aluminum case with no ventilation holes. No problems at all.
For the mechanical part, you should ensure close contact for increased effectivity by means of thermal grease or pads. Depending on the case material, if feasible wrt. the layout, you could even use that as heat sink directly, since ...
... you're implementing class 1 anyway.
I'm running a single NC400 + SMPS600 in a closed aluminum case with no ventilation holes. No problems at all.
I'm doing the same with no issues. I don't get all the worry. I really don't.
Just so I get this right - correct me wherever I am wrong.
If I do not ensure 6mm clearance on the top of the SMPS by clamping its heat sinks directly to a case panel, I am in direct violation of class 2 rules unless I implement the alternative, which is 0.5mm of insulation material, IMHO.
I was not talking about anything else there, as the need for additional heat transfer seems subjective depending on what exact case is applied. Hence, "no need to fuss" -- as long as things remain reasonably cool, and the clearance is achieved by the steps you outlined. If I am completely paranoid about thermal issues, and my resulting implementation goes against the principles of class 2, I am supposed to either check how I can get back in line, or go for a class 1 setup altogether - or so I thought.
Hi,
yes it´s a class 1 setup. The problem is not about being completely paranoid but of a smps switching off because it get´s too hot. The space inside the case is rather cramped with only 52mm height and 92mm width.
I will try some ventilation holes directly above the heatsinks and see what happens before trying to increase the heatsinking.
thanks,
William
yes it´s a class 1 setup. The problem is not about being completely paranoid but of a smps switching off because it get´s too hot. The space inside the case is rather cramped with only 52mm height and 92mm width.
I will try some ventilation holes directly above the heatsinks and see what happens before trying to increase the heatsinking.
thanks,
William
From Bruno at AC:
At low power the FET temperature is very nearly equal to the temperature of the mounting base.
Capacitor lifetime works as follows: the manufacturer gives a baseline lifetime at a given temperature and under rated load conditions. From that baseline, lifetime doubles for every 10°C reduction in temperature. Rated ripple current is usually taken to contribute 10°C to temperature, so no or very low ripple current constitutes another doubling of lifetime. Running the capacitor at a lower than rated voltage once again multiplies lifetime by a factor that goes up fairly quickly with dropping voltage.
The baseline lifetime of the caps on the NC400 is given as 2000 hours at 105°C. At idle or during normal listening (i.e. low ripple current) and 60°C we can thus multiply lifetime with a factor 2^((105-60)/10+1)=45. Expected lifetime works out as 90000 hours, or just over 10 years of continuous operation. At that point the capacitor will go out of spec, but not fail outright.
There is some irony to temperature and class D. People expect class D to run cool. That would only be true if one substituted a class D power stage for a class A amp and kept the heat sink as it were. Of course it would run stone cold under those conditions. But the point of course is to cash in on the diminished heat output and reduce heatsink capacity until the final temperature is the same.
NC400 Cooling
At low power the FET temperature is very nearly equal to the temperature of the mounting base.
Capacitor lifetime works as follows: the manufacturer gives a baseline lifetime at a given temperature and under rated load conditions. From that baseline, lifetime doubles for every 10°C reduction in temperature. Rated ripple current is usually taken to contribute 10°C to temperature, so no or very low ripple current constitutes another doubling of lifetime. Running the capacitor at a lower than rated voltage once again multiplies lifetime by a factor that goes up fairly quickly with dropping voltage.
The baseline lifetime of the caps on the NC400 is given as 2000 hours at 105°C. At idle or during normal listening (i.e. low ripple current) and 60°C we can thus multiply lifetime with a factor 2^((105-60)/10+1)=45. Expected lifetime works out as 90000 hours, or just over 10 years of continuous operation. At that point the capacitor will go out of spec, but not fail outright.
There is some irony to temperature and class D. People expect class D to run cool. That would only be true if one substituted a class D power stage for a class A amp and kept the heat sink as it were. Of course it would run stone cold under those conditions. But the point of course is to cash in on the diminished heat output and reduce heatsink capacity until the final temperature is the same.
NC400 Cooling
I was refferring to myself, figuratively, and nobody else. I stopped projecting my own reasoning (or the lack thereof) onto others and jumping to conclusions a long time ago
Aluminium sheets are quite the heat sinks, so if you build a case out of those and have a reasonable amount of air inside along with your smps+ncore which can slowly circulate and gently carry away the heat from where it originates, you might soon end up with a thermal equilibrium. If the case is too small for that, some venting or heat sinking might be required - there is one report of smps shutdown.
Some just worry about the effect of heat on the life span of their caps - I touched those on my ncore and they were quite on the warm side, the output inductor being the hottest part. The transistors et.al. stuck to the circular heat sink are taken care of, but the rest is basically air cooled. My entire setup is still very open, all just put on a thin sheet of steel ~20cm above the floor, each amp heatsink-down and the smps heatsink-up, as they should. I am not worried about thermal conditions now.
"Considering" as I am (that's one way not to say paranoid), I merely wonder where all that heat will end up, if not carried away via the case directly, or by air as it is now. Granted, it's not much but it is there and unless taken away, will accumulate. I am mostly checking my options before going into any direction I cannot easily change later on. Adding a heat pipe from output inductor to case plate is not one of them, whereas binding smps heat sinks to case plate looks sensible to me, and a lot less invasive than spinning up that dremel.
Thanks a bunch for this, it will let me pursue thermal issues with a lot less motivation now. Efficiency lies in the details. Just wondering. If I can keep those things alive a bit longer by adding a measily strip of thermal pad and tinkering with the mechanical part of the setup, then so be it. Not much effort involved AFAICS. By no means I wanted to make the impression on anybody that such a thing would be necessary.
Perhaps you are sharing the raves of a lunatic (edit: whilst reading my mumblings), perhaps someone can fix an issue by looking at it from a different angle. As long as nobody gets hurt, I don't mind much
Last edited:
Same hereI'm running a single NC400 + SMPS600 in a closed aluminum case with no ventilation holes. No problems at all.
That's how I understand it
Again, that's how I understand it
However, I am not an electrical safety expert, you make your own choices and decisions, and my understanding should not be taken as advice
I look at it this way. Bruno says that the caps are good for 90000 hrs. I have no doubt that the amps will be replaced way before then.
I highly doubt that Bruno & Hypex will make me wait that long for a NC400 Rev A.
Interesting viewpoint. For me, I am tired of constantly changing my gear. I came to the DIY approach because in many cases I don't think you can buy gear as good as you can make it, unless you spend obscene amounts of money. I would rather spend the money on software. I plan on having my nc400s a long, long time -- my Musical Fidelity A3 has been continuously running since 1999.
I appreciate though that many of us here are constant tinkerers and enjoy that immensely. Hats off to that!
Hey there...
Anyone running their nCores with a Twisted Pear Buffalo/Legato combo with no output caps (perhaps no pin 1 connection at all)?
And:
"I plan on having my nc400s a long, long time -- my Musical Fidelity A3 has been continuously running since 1999."
Just because a piece of electronics is running does not mean it is running with optimal performance. If the A3 has been continuously powered for over ten years now it is very likely that some, if not all, of the electrolytic caps have lost some of their capacity due to drying out. It is likely time for some maintenance on this amp.
Anyone running their nCores with a Twisted Pear Buffalo/Legato combo with no output caps (perhaps no pin 1 connection at all)?
And:
"I plan on having my nc400s a long, long time -- my Musical Fidelity A3 has been continuously running since 1999."
Just because a piece of electronics is running does not mean it is running with optimal performance. If the A3 has been continuously powered for over ten years now it is very likely that some, if not all, of the electrolytic caps have lost some of their capacity due to drying out. It is likely time for some maintenance on this amp.
- Status
- Not open for further replies.