Need NC1200 owners input

[NTK]

These are interesting details.
There are a few things not aligning in my mind:
- if March case cools better by conduction, I would expect it to be hotter in the same conditions, seems not to be the case (pun not intended but self appreciated)
- as already written, the SMPS1200 and the NC1200 dissipate a good proportion of heat in the air, and in this situation a higher thermal mass and lower surface of vents may not be an advantage.
- as you write, both cases are not optimized for cooling, they even have an insulation layer between themselves and what they are supposed to cool. How can one be much better than the other?

I am at a lost to understand/explain how the averaged surface temperatures of the March case and the Ghent case can be very different, when given the same external environment and same heat generation. The external geometries of the 2 cases aren't very different. Their total surface areas are almost the same. What I am saying is the March design is better at letting the heat from the Hypex modules transfer and distribute to the case. From the case to the room air, the March and Ghent designs are similar IMO.

When we use the electrical analogy, the average case temperature is V_in, room air temperature is V_out, thermal resistance for convective heat transfer is R, and heat flow is I. Ohm's law also applies here (V_in - V_out = I * R). When R, I and V_out are fixed, V_in is fixed too and cannot be different between the 2 cases.

If the "system" has reached steady state, which it should be after a few hours of idling, the different thermal masses shouldn't matter. Hard to imagine the tiny holes in the top cover of the VTV (Ghent) case will make much differences for natural convection.

 

[echopraxia]

I am at a lost to understand/explain how the averaged surface temperatures of the March case and the Ghent case can be very different, when given the same external environment and same heat generation.

I have my P451's connected to my Salon2's currently, so I left them running for 4 hours or so playing background music at moderate volume (I can't tell you exact how much power draw that is though, and yeah I know the Purifi 1ET400A is much more efficient than the Hypex NC1200). Both cases are still cool to the touch on all sides. The bottom would barely qualify as "warm", if that. I can also test the P701s later when they arrive, but I don't expect it will be fundamentally different from what March Audio has shown from it idling over night. I have no idea how hot the underside of the P701 will be though, but can attest to the P451's underside remining quite cool even after hours of operation.

My guess would be that March Audio's case channels airflow more efficiently over the hottest components. Looking at the case design, it looks quite clear the airflow outlet was intentionally positioned. I know that March Audio mentioned he uses thermal compound as well, but I don't know how much of a difference that makes. Beyond that, I don't know.

Lastly, I'm surprised March Audio is as generous as he is with build & design information here... IMO, I wouldn't expect March Audio to reveal all their secrets, and certainly wouldn't blame them if they don't want to publish the full blueprints and manufacturing instructions for a cooling design he's clearly worked carefully to perfect

 

[boXem]

I am at a lost to understand/explain how the averaged surface temperatures of the March case and the Ghent case can be very different, when given the same external environment and same heat generation. The external geometries of the 2 cases aren't very different. Their total surface areas are almost the same. What I am saying is the March design is better at letting the heat from the Hypex modules transfer and distribute to the case. From the case to the room air, the March and Ghent designs are similar IMO.

When we use the electrical analogy, the average case temperature is V_in, room air temperature is V_out, thermal resistance for convective heat transfer is R, and heat flow is I. Ohm's law also applies here (V_in - V_out = I * R). When R, I and V_out are fixed, V_in is fixed too and cannot be different between the 2 cases.

If the "system" has reached steady state, which it should be after a few hours of idling, the different thermal masses shouldn't matter. Hard to imagine the tiny holes in the top cover of the VTV (Ghent) case will make much differences for natural convection.

Since I agree on what you wrote, I also guess that we agree that the two cases are not fundamentally different in terms of cooling

... cut...
for a cooling design he's clearly worked carefully to perfect

I know internet doesn't encourage to nuances, but some would be needed here.

 

[echopraxia]

I know internet doesn't encourage to nuances, but some would be needed here.

I don’t understand what you’re trying to say in relation to what you quoted from me.

 

[boXem]

I don’t understand what you’re trying to say in relation to what you quoted from me.

Isn't "perfect" a bit on the exaggeration side?

 

[echopraxia]

Isn't "perfect" a bit on the exaggeration side?

(Sorry, just saw this reply now.)

In a phrase like "worked to perfect a design", the word "perfect" here does not mean "perfect" as an adjective, but as a verb. In this sense, "perfect" is mostly synonymous with "optimize" as a process/verb: It's not making any claims that perfection has been reached (which is impossible anyway, in most cases), but rather describes actions performed towards moving in that general direction.

Sometimes, "perfect" is used as a verb instead of "optimize" when the latter might be too clinical (agnostic to the 'why' aspect). In contrast, "perfect" as a verb may convey more of a sense of the underlying motivations / meta-objective of the optimization being rooted in perfectionism. I chose this word simply to convey my perception that March Audio's products are motivated in this way. I could be wrong about that, but that is my takeaway as an owner of several of their products.

Sorry for any confusion this caused... I hope that makes sense. In retrospect maybe if I used the word "optimize" instead of "perfect", it would have been less ambiguous.

 

[muslhead]

(Sorry, just saw this reply now.)

In a phrase like "worked to perfect a design", the word "perfect" here does not mean "perfect" as an adjective, but as a verb. In this sense, "perfect" is mostly synonymous with "optimize" as a process/verb: It's not making any claims that perfection has been reached (which is impossible anyway, in most cases), but rather describes actions performed towards moving in that general direction.

Sometimes, "perfect" is used as a verb instead of "optimize" when the latter might be too clinical (agnostic to the 'why' aspect). In contrast, "perfect" as a verb may convey more of a sense of the underlying motivations / meta-objective of the optimization being rooted in perfectionism. I chose this word simply to convey my perception that March Audio's products are motivated in this way. I could be wrong about that, but that is my takeaway as an owner of several of their products.

Sorry for any confusion this caused... I hope that makes sense. In retrospect maybe if I used the word "optimize" instead of "perfect", it would have been less ambiguous.

Doing so would have "optimized" your post

 

[Rick Sykora]

Hypex confirms that the NC1200 runs warm in the Ghent case btw. If you are handy (and careful), might want to try some thermal paste.

 

[muslhead]

Hypex confirms that the NC1200 runs warm in the Ghent case btw. If you are handy (and careful), might want to try some thermal paste.

Where is it applied, under the arms and legs?

 

[Rick Sykora]

Where is it applied, under the arms and legs?

only if they are aluminum!

 

[HomerJay]

I recently took delivery of two NC1200 amps built by a local guy here. One is stereo with two SMPS 1200 power supplies with input buffers/Sparkos OpAmps and the other a Tri-channel with two SMPS 1200 power supplies (one Stereo pair share one supply).

Both are in similar aluminium chassis with huge heat sinks on sides and smaller heat sink on fronts.

Tri-channel during final wiring: (you can see the existing thermal pad on right SMPS)

After 4.5 hours of playing full range 5.1 music at close to reference levels both measured 46 deg C on top, a little less on front and a little less on sides heat sinks. Ambient temperature was around 23 deg C.

I‘ve ordered new 15W/m-k thermal pads (Gelid) to cover the entire SMPS heatsink to chassis sides (existing are 3.2W/m-k 1mm thick and does not cover entire heat sink) so hoping I can get the temperature down a few degrees.

Since I‘ll need to remove all four SMPS (two amps) I thought I’d also add thermal pads to the bottom heat sinks for all SMPS and the single NC1200 mounted on the bottom plate in the tri-channel. I found some 20W/m-k Graphite pads that can be cut down to exact heat sink size from 300x300mm sheets.

I didn’t go for thermal paste as it breaks down in relatively short time (less than 3 years) and thermal transfer is higher for high performance pads vs paste and way better longevity for pads.

I’ll post my results in about a week or so once Ive received the pads and work is complete. I’ll do just the heat sink sides first, test; then do bottom heat sinks (alone) and test; then all heat sinks (sides and bottoms) and test.