Think about the word: heat-sink. To "sink" or wick away heat. Its purpose (ironically) isn't as a reservoir of heat, no. It is just the opposite: the least amount of "reservoirishness" and the greatest amount of "heat-moving-ness". Heat conduction.
This is why modern super-high end PC's have liquid cooling for the processors: not a huge block of aluminum. The water efficiently comports the produced heat AWAY from the source. The chip. This is also why aluminum is so often used as a heat-sinking building material. Its cheap, per unit mass, it is an excellent heat conductor, and it doesn't offer significant extrusion challenges (for cheap manufacture) into quality sinks.
Oh, you could use all-copper. But it'd cost a king's ransom. Or liquid cooling, but the pumping noise is rather distracting to a fine playback experience. Stick to aluminum. Its really, really good.
GoatGuy
This is why modern super-high end PC's have liquid cooling for the processors: not a huge block of aluminum. The water efficiently comports the produced heat AWAY from the source. The chip. This is also why aluminum is so often used as a heat-sinking building material. Its cheap, per unit mass, it is an excellent heat conductor, and it doesn't offer significant extrusion challenges (for cheap manufacture) into quality sinks.
Oh, you could use all-copper. But it'd cost a king's ransom. Or liquid cooling, but the pumping noise is rather distracting to a fine playback experience. Stick to aluminum. Its really, really good.
GoatGuy
Im old Maine, where rock was our best crop, a stone barn was considered "warm" (insulated), not a heat-sink. Of course in part this is because a foot of rock costs little more than an inch of wood board (rock on site and board hauled from sawmill). If it keeps the cows warm (good), won't it keep the chip warm (bad)?
Get some numbers.
Thermal Conductivity of common Materials and Gases
Specific Heat of common Substances
I thought specific heat might play a part in short-term rock fondling, but apparently no real effect.
Thermal Conductivity (more below)
Hand, rock, paper, scissors-
Beef 0.45 (hand substitute)
Rock 2 - 7
Paper 0.05
Steel 43
In the hand, stone sucks heat faster than paper, slower than steel. Cold paper is not real chilly. Cold steel is cold! Cold rock is cold but not like steel.
Air 0.024
Aluminum 205
Copper 401
Aluminum whups stone 50X. Aluminum whups steel 5X. Copper beats Aluminum 2X.
The heat path from a small hot chip to all the air in the world is a pyramid. You need your best stuff at the point. Hot chips use a copper plate to take heat from the chip and spread it about 50X bigger area. Here Aluminum is plenty good and cheaper than copper. Cheap enough to use out to 1,000X the area where we transfer to air. That's a serious mismatch, but nothing else has good conductivity, low cost, and producible in complex shapes. (Actually in larger work we transfer to metal to water and pump to a large water/air heat exchanger- your car radiator.)
The chips are surely designed to pass heat to Aluminum. With stone the heat rise in the first mm away from the chip would be about 50X what the designer expected. "Maybe" you could run a 100W chip at 2 Watts actual output (if it would work on the low voltage).
Also the contact between two surfaces must be INTIMATE. Air voids conduct 10,000X worse than Aluminum. Grease is only slightly better, not a cure. The stone must be finished as smooth as a metal heatsink. Yes, you can get Marble finished finger-smooth, but the surface is pitted.
The engineered path would be chip to Aluminum heat-spreader at least 50X the area of the chip heat pad. Roughly the center of a very good Al sink, fins cut off. Still it will take a LOT of rock to even approach the total surface area of a finned heatsink. It needs to be very much bigger than a finned sink.
Glue some pretty pebbles to 1% of the Al fins and rave about the rock-solid imaging you hear.
______________________
Thermal Conductivity
Air 0.024
Aluminum 205
Beef 0.45
Copper 401
Glass 1.05
Marble 2.08 - 2.94
Paper 0.05
Plywood 0.13
Rock, solid 2 - 7
Softwoods (fir, pine ..) 0.12
Steel, Carbon 43
Stainless Steel 16
Water 0.58
Get some numbers.
Thermal Conductivity of common Materials and Gases
Specific Heat of common Substances
I thought specific heat might play a part in short-term rock fondling, but apparently no real effect.
Thermal Conductivity (more below)
Hand, rock, paper, scissors-
Beef 0.45 (hand substitute)
Rock 2 - 7
Paper 0.05
Steel 43
In the hand, stone sucks heat faster than paper, slower than steel. Cold paper is not real chilly. Cold steel is cold! Cold rock is cold but not like steel.
Air 0.024
Aluminum 205
Copper 401
Aluminum whups stone 50X. Aluminum whups steel 5X. Copper beats Aluminum 2X.
The heat path from a small hot chip to all the air in the world is a pyramid. You need your best stuff at the point. Hot chips use a copper plate to take heat from the chip and spread it about 50X bigger area. Here Aluminum is plenty good and cheaper than copper. Cheap enough to use out to 1,000X the area where we transfer to air. That's a serious mismatch, but nothing else has good conductivity, low cost, and producible in complex shapes. (Actually in larger work we transfer to metal to water and pump to a large water/air heat exchanger- your car radiator.)
The chips are surely designed to pass heat to Aluminum. With stone the heat rise in the first mm away from the chip would be about 50X what the designer expected. "Maybe" you could run a 100W chip at 2 Watts actual output (if it would work on the low voltage).
Also the contact between two surfaces must be INTIMATE. Air voids conduct 10,000X worse than Aluminum. Grease is only slightly better, not a cure. The stone must be finished as smooth as a metal heatsink. Yes, you can get Marble finished finger-smooth, but the surface is pitted.
The engineered path would be chip to Aluminum heat-spreader at least 50X the area of the chip heat pad. Roughly the center of a very good Al sink, fins cut off. Still it will take a LOT of rock to even approach the total surface area of a finned heatsink. It needs to be very much bigger than a finned sink.
Glue some pretty pebbles to 1% of the Al fins and rave about the rock-solid imaging you hear.
______________________
Thermal Conductivity
Air 0.024
Aluminum 205
Beef 0.45
Copper 401
Glass 1.05
Marble 2.08 - 2.94
Paper 0.05
Plywood 0.13
Rock, solid 2 - 7
Softwoods (fir, pine ..) 0.12
Steel, Carbon 43
Stainless Steel 16
Water 0.58
This is αιγινήτικο κανάτι (Aegina's jug). It keeps the water inside it cool by vaporizing some through the ceramic body. Simple physics! Once, I thought how to use this as a heatsink but the drawbacks are humidity and necessity to fill water all the time...
Attachments
Very thin. NOT all that thermally conducty; not like metals. Advantage is real low electrical conductivity and high breakdown. A thin electric-gap, not a heat spreader.
Yes, evaporation is effective *after* you spread the heat over a large area to get a lot of evaporation. A jug of water is naturally large. Power chips, with present production economics, are invariably "small".
Small engines used to be evaporative cooling. Just an open water jacket. For a short run, the specific heat of water would absorb some HP-Hrs. Longer run, water gets hot and there was some evaporative cooling. Keep running, water boils away, you get a LOT of cooling but the water consumption may be high.
Yes.
Don't use stone as a heat sink.
Aluminum, or copper, or both esp. with water cooling of chips.
GoatGuy
Let's review why a stone heatsink would be a bad idea. Your transistor is a concentrated heat source, that is to say it is making a lot of heat in a small space. You need to transfer this heat away as fast as possible to keep the die cool. The total rate of heat transfer is dependent on the contact area and delta-T at the interface, and the thermal conductivity of the "sink" material. Since the area is small, you must have high thermal conductivity to keep delta-T down. Since stone has relatively low thermal conductivity, if it was used as a heat sink the delta-T would be high, meaning your transistor would be able to generate heat faster than it could be removed thru the contact are with the "sink". It would soon heat up and fail.
If you really want to use stone, you could first transfer the heat into a metal plate, and then from the plate via a large contact area to stone, which could then transfer the heat to air, also via a large contact area. But why do that when you can just use a finned aluminum heat sink to begin with? When sized appropriately it works very well, is robust, easy to obtain and fabricate, and is relatively inexpensive.
Was the OP serious? He hasn't been back. Is anyone else here serious? As well as getting a very smooth and fine contact between surfaces so as not to use too much thermal grease whose only function is to eliminate air gaps, pressure is the other important factor in thermal transfer, how are you going to attach the aluminium to the stone tightly without it cracking due to expansion and contraction?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.