Has anybody ever stacked two CPU coolers?...

[Commander Kinsey]

Has anybody ever stacked two CPU coolers? First one fitted normally, then the second welded on top of it. Theoretically, you have twice the cooling, if it conducts through the first fast enough. I\'m referring to the type with the fan on the side.

 

[Phil Hobbs]

On 2020-07-25 14:01, Commander Kinsey wrote:

Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[George Herold]

On Saturday, July 25, 2020 at 2:02:02 PM UTC-4, Commander Kinsey wrote:
> Has anybody ever stacked two CPU coolers? First one fitted normally, then the second welded on top of it. Theoretically, you have twice the cooling, if it conducts through the first fast enough. I\'m referring to the type with the fan on the side.

If you\'ve got two of the same size TECs, then you might do better to
get a bigger heat sink and run them in parallel. (not series stacked)
More current rather than more voltage.

George H.

 

[Commander Kinsey]

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

 

[legg]

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
<CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

You\'d get a better result with a bigger mounting interface to the heat
source, coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

RL

RL

 

[Commander Kinsey]

On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on a bookshelf. It\'s a research machine sat in my garage made out of spare parts, and no case was the right shape to fit a dual server board into.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the new one would also be.

 

[Commander Kinsey]

On Sat, 25 Jul 2020 20:04:35 +0100, George Herold <ggherold@gmail.com> wrote:

On Saturday, July 25, 2020 at 2:02:02 PM UTC-4, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally, then the second welded on top of it. Theoretically, you have twice the cooling, if it conducts through the first fast enough. I\'m referring to the type with the fan on the side.

If you\'ve got two of the same size TECs, then you might do better to
get a bigger heat sink and run them in parallel. (not series stacked)
More current rather than more voltage.

I\'m not using TECs. Just a heatsink with heatpipes, and a fan.

 

[Ricketty C]

On Saturday, July 25, 2020 at 5:39:43 PM UTC-4, Commander Kinsey wrote:

On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on a bookshelf. It\'s a research machine sat in my garage made out of spare parts, and no case was the right shape to fit a dual server board into.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the new one would also be.

Why would you not just get a larger heat sink device, AKA a *better* heat sink?

The first heat sink is not designed to carry heat through to the second heat sink, so it is unlikely to work well in a stacked mode. If you were doing that you would make the bottom heat sink fins thicker to conduct more heat to the upper heat sink with less drop in temperature.

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

 

[Carl]

\"Commander Kinsey\" wrote in message newsp.0obrcuevwdg98l@glass...

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally
disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on
the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough
heat is transferred to the top of the first one, and the one I have is hot
to touch on the top, then that heat could be conducted away by the second
one.

Never done it but I can\'t see any reason it won\'t work, the question is how
well. Biggest limitation I see is that you are using just that top fin as
your \"heat spreader\" to conduct the heat laterally away from the lower heat
pipes, and that is a pretty thin piece of metal. I\'d make a flat aluminum
plate that is a tight fit between the top and second fins, and maybe another
one between the next pair down. Make the lower one stick out each side and
drill the appropriate holes in it so the upper heat sink assembly bolts to
it like it was the motherboard. That clamps the plates and fins and new
heat sink together for good conduction. Put heat sink compound on both
sides of each plate, and see how it works.

The other thing you can do if you have the room is to add a second fan to
your existing heat sink to pull air out the exit side. Go to
https://noctua.at/en/support/compatibility-lists/mainboard and look up your
motherboard to see what their biggest cooler is, and look at their models
for more ideas.

If you do stack two coolers, please report back on how well it worked -
enquiring minds want to know .

--
Regards,
Carl Ijames

 

[Commander Kinsey]

On Sat, 25 Jul 2020 23:26:31 +0100, Ricketty C <gnuarm.deletethisbit@gmail.com> wrote:

On Saturday, July 25, 2020 at 5:39:43 PM UTC-4, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on a bookshelf. It\'s a research machine sat in my garage made out of spare parts, and no case was the right shape to fit a dual server board into.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the new one would also be.

Why would you not just get a larger heat sink device, AKA a *better* heat sink?

Because they don\'t really exist or are phenominally expensive.

> The first heat sink is not designed to carry heat through to the second heat sink, so it is unlikely to work well in a stacked mode. If you were doing that you would make the bottom heat sink fins thicker to conduct more heat to the upper heat sink with less drop in temperature.

But the lower heatsink has heatpipes going all the way to the top. Those are what will deliver heat to the top heatsink.

 

[Jasen Betts]

On 2020-07-25, Commander Kinsey <CFKinsey@military.org.jp> wrote:
> Has anybody ever stacked two CPU coolers? First one fitted normally, then the second welded on top of it.

seems odd.

> Theoretically, you have twice the cooling, if it conducts through the first fast enough.

Yeah, if. I doubt that it would, often the fins are tapered.

> I\'m referring to the type with the fan on the side.

Just get a taller one that has heat pipes, then conduction is less of
an issue.

--
Jasen.

 

[Ricketty C]

On Saturday, July 25, 2020 at 6:49:02 PM UTC-4, Commander Kinsey wrote:

On Sat, 25 Jul 2020 23:26:31 +0100, Ricketty C <gnuarm.deletethisbit@gmail.com> wrote:

On Saturday, July 25, 2020 at 5:39:43 PM UTC-4, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on a bookshelf. It\'s a research machine sat in my garage made out of spare parts, and no case was the right shape to fit a dual server board into.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the new one would also be.

Why would you not just get a larger heat sink device, AKA a *better* heat sink?

Because they don\'t really exist or are phenominally expensive.

Some years ago I wanted to make a water block for the CPU. I was going to modify a cheap air heatsink which would have worked well, but the problem I had was with finding a good glue for plexi to aluminum.

Not sure what you consider expensive. A $20 aquarium pump and a 5 gal container of water will do the job for quite a while. One guy piped his computer to a 55 gal drum in his garage and never measured a significant temperature rise. I got a cracked radiator from a friend that I could have repaired for this task, but just never followed through. I guess the fear of leaking on the mobo was a bit daunting. I could have turned the mobo over so the water block was on the bottom! Why didn\'t I think of that then?

The first heat sink is not designed to carry heat through to the second heat sink, so it is unlikely to work well in a stacked mode. If you were doing that you would make the bottom heat sink fins thicker to conduct more heat to the upper heat sink with less drop in temperature.

But the lower heatsink has heatpipes going all the way to the top. Those are what will deliver heat to the top heatsink.

Sorry, I\'m not picturing this at all. How do you plan to make a good thermal and mechanical attachment?

--

Rick C.

+ Get 1,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209

 

[Commander Kinsey]

On Sun, 26 Jul 2020 00:01:37 +0100, Jasen Betts <jasen@xnet.co.nz> wrote:

On 2020-07-25, Commander Kinsey <CFKinsey@military.org.jp> wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally, then the second welded on top of it.

seems odd.

Seems it might work.

Theoretically, you have twice the cooling, if it conducts through the first fast enough.

Yeah, if.

The top of the first is hot to touch.

> I doubt that it would, often the fins are tapered.

They\'re not.

I\'m referring to the type with the fan on the side.

Just get a taller one that has heat pipes, then conduction is less of
an issue.

Not available and/or expensive.

 

[Commander Kinsey]

On Sat, 25 Jul 2020 23:45:51 +0100, Carl <carl.ijamesXYZ@zyxverizon.net> wrote:

\"Commander Kinsey\" wrote in message newsp.0obrcuevwdg98l@glass...

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally
disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on
the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough
heat is transferred to the top of the first one, and the one I have is hot
to touch on the top, then that heat could be conducted away by the second
one.

Never done it but I can\'t see any reason it won\'t work, the question is how
well. Biggest limitation I see is that you are using just that top fin as
your \"heat spreader\" to conduct the heat laterally away from the lower heat
pipes, and that is a pretty thin piece of metal. I\'d make a flat aluminum
plate that is a tight fit between the top and second fins, and maybe another
one between the next pair down. Make the lower one stick out each side and
drill the appropriate holes in it so the upper heat sink assembly bolts to
it like it was the motherboard. That clamps the plates and fins and new
heat sink together for good conduction. Put heat sink compound on both
sides of each plate, and see how it works.

I\'d not go to the extra bother. I\'d prefer just to buy an expensive one if the top fin isn\'t enough to conduct. And since ther plate on the bottom of the top one will be close to the heatpipes, it\'s not a long distance to go through that top fin.

The other thing you can do if you have the room is to add a second fan to
your existing heat sink to pull air out the exit side. Go to
https://noctua.at/en/support/compatibility-lists/mainboard and look up your
motherboard to see what their biggest cooler is, and look at their models
for more ideas.

No, tried that already. If you blow both inwards, they fight as they\'re too close, it makes it 3C warmer! If you blow one in and one out, it only makes it 3C cooler. I\'m just making it easier on the fan and not really increasing the airflow - maybe due to turbulence - if the out of the first fan is producing turbulent air into the in of the second fan, the second fan can\'t do much work.

If you do stack two coolers, please report back on how well it worked -
enquiring minds want to know .

I will try it if the CPU overheats again, out of the 4 Xeons, two have never done it, one did it once, and one did it twice. I think it depends on what program is running (different instruction set?). So you\'ll have to wait until it happens, then the wait for the thing to arrive in the post (which could be form China if they\'re cheaper), then hope I\'ve remembered to report in here.

 

[Bill Sloman]

On Sunday, July 26, 2020 at 6:04:10 AM UTC+10, legg wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

You\'d get a better result with a bigger mounting interface to the heat
source, coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

Wrong, if the heat-pipe has been properly evacuated before it was sealed off.

We ran into this, and had to set up a test rig to make sure that the heat pipes we had got had been properly evacuated. We only had to send back a few before we stopped get parts that failed the test.

Sloman A.W., Buggs P., Molloy J., and Stewart D. “A microcontroller-based driver to stabilise the temperature of an optical stage to 1mK in the range 4C to 38C, using a Peltier heat pump and a thermistor sensor” Measurement Science and Technology, 7 1653-64 (1996)

The problem comes if there\'s enough pressure of non-condensible gas in the heatpipe to slow the evaporation at hot end, and - more important - to slow down the evaporating vapour on its way to heat-dissipating end.

If there\'s no non-condensible gas to get in the way, the vapour can flow very fast.

--
Bill Sloman, Sydney

 

[Martin Brown]

On 25/07/2020 22:39, Commander Kinsey wrote:

On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

Stacking CPU coolers will have a huge impedance mismatch on the second
one - essentially wasting most of its capability.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on
a bookshelf.  It\'s a research machine sat in my garage made out of spare
parts, and no case was the right shape to fit a dual server board into.

It is antisocial to operate a fast PC outside an enclosure other than
for testing because it is a broadband emitter of RF interference to GHz.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

You can get oversize heatsinks using heatpipes to shift more heat from
the CPU to the finned metal. Larger fans are quieter too. If all else
fails water cool the thing which is what the serious overclockers do.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the
new one would also be.

The way to cool it is use a larger heatsink designed to match the CPU
surface and use the right compound to make the best thermal joint.

--
Regards,
Martin Brown

 

[Commander Kinsey]

On Sun, 26 Jul 2020 00:03:28 +0100, Ricketty C <gnuarm.deletethisbit@gmail.com> wrote:

On Saturday, July 25, 2020 at 6:49:02 PM UTC-4, Commander Kinsey wrote:
On Sat, 25 Jul 2020 23:26:31 +0100, Ricketty C <gnuarm.deletethisbit@gmail.com> wrote:

On Saturday, July 25, 2020 at 5:39:43 PM UTC-4, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

On Sat, 25 Jul 2020 20:06:20 +0100, \"Commander Kinsey\"
CFKinsey@military.org.jp> wrote:

On Sat, 25 Jul 2020 19:21:25 +0100, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-25 14:01, Commander Kinsey wrote:
Has anybody ever stacked two CPU coolers? First one fitted normally,
then the second welded on top of it. Theoretically, you have twice the
cooling, if it conducts through the first fast enough. I\'m referring to
the type with the fan on the side.

Stacked TECs are an old technology. They work at some level if you\'re
careful, but the inefficiency of the earlier stages produces a huge heat
load on the later ones, so the available delta-T is generally disappointing.

I wasn\'t considering peltier. Just a finned lump of metal with a fan on the side. Like this:
http://www.shoppingexpress.com.au/assets/full/CL-P039-AL12BL-A.jpg
Now consider another of those clamped to the top of that one. If enough heat is transferred to the top of the first one, and the one I have is hot to touch on the top, then that heat could be conducted away by the second one.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on a bookshelf. It\'s a research machine sat in my garage made out of spare parts, and no case was the right shape to fit a dual server board into.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the new one would also be.

Why would you not just get a larger heat sink device, AKA a *better* heat sink?

Because they don\'t really exist or are phenominally expensive.

Some years ago I wanted to make a water block for the CPU. I was going to modify a cheap air heatsink which would have worked well, but the problem I had was with finding a good glue for plexi to aluminum.

Been there done that, flooded a motherboard (it recovered), and caused a power supply to go bang (I\'d cooled the main heatsink, but not some of the smaller components that required a bit of airflow).

> Not sure what you consider expensive. A $20 aquarium pump and a 5 gal container of water will do the job for quite a while. One guy piped his computer to a 55 gal drum in his garage and never measured a significant temperature rise. I got a cracked radiator from a friend that I could have repaired for this task, but just never followed through. I guess the fear of leaking on the mobo was a bit daunting. I could have turned the mobo over so the water block was on the bottom! Why didn\'t I think of that then?

A central heating radiator is good, especially for multiple computers, or someone I know with SEVEN graphics cards on one motherboard - they\'re right up against each other, the water blocks make them only one unit wide instead of two.

The first heat sink is not designed to carry heat through to the second heat sink, so it is unlikely to work well in a stacked mode. If you were doing that you would make the bottom heat sink fins thicker to conduct more heat to the upper heat sink with less drop in temperature.

But the lower heatsink has heatpipes going all the way to the top. Those are what will deliver heat to the top heatsink.

Sorry, I\'m not picturing this at all. How do you plan to make a good thermal and mechanical attachment?

Thermal grease and cable ties (zip ties if you\'re American).

Anyway I fixed it, by changing the fan, I\'d forgotten that some fans are really crap - they can be anything from 1100rpm to 4000rpm. I happened to have a few 3000rpm fans sitting about, so I replaced what looks like about 1100rpm with 3000rpm, and lost 20C of temperature (I don\'t care about the noise, it\'s in the garage and operated remotely). What\'s odd is I used to have one CPU at 65C and one at 75C (with identical coolers), and now the 75C has dropped to 55C. I guess the physical position meant cooler air for one of them.

 

[Commander Kinsey]

On Sun, 26 Jul 2020 10:43:46 +0100, Martin Brown <\'\'\'newspam\'\'\'@nezumi.demon.co.uk> wrote:

On 25/07/2020 22:39, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

Stacking CPU coolers will have a huge impedance mismatch on the second
one - essentially wasting most of its capability.

Since the heatpipes of each would be very close, the length of normal steel would be minimal.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on
a bookshelf. It\'s a research machine sat in my garage made out of spare
parts, and no case was the right shape to fit a dual server board into.

It is antisocial to operate a fast PC outside an enclosure other than
for testing because it is a broadband emitter of RF interference to GHz.

I don\'t give a fuck. And that does explain why the monitor jumps a bit. There are 5 such open cased machines there. I wonder if that\'s why my neighbour has installed some kind of WiFi repeater in his garden Pissed him off without even trying!

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

You can get oversize heatsinks using heatpipes to shift more heat from
the CPU to the finned metal.

But the interface between the CPU and the block would be the same size - the area of the CPU.

Larger fans are quieter too. If all else
fails water cool the thing which is what the serious overclockers do.

Expensive, hassle, leakages....

And serious overclockers use liquid nitrogen, or submerse the whole thing in oil!

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the
new one would also be.

The way to cool it is use a larger heatsink designed to match the CPU
surface and use the right compound to make the best thermal joint.

I just replaced the fan with a faster one. Lost 20C. I\'d forgotten how shit some fans are. 1100rpm to 4000rpm!

 

[Ricketty C]

On Sunday, July 26, 2020 at 5:43:51 AM UTC-4, Martin Brown wrote:

On 25/07/2020 22:39, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:


Stacking CPU coolers will have a huge impedance mismatch on the second
one - essentially wasting most of its capability.

A highly faulty analysis.

People often think the best match for power is equal impedance, but that is not the case when the impedance of the load is fixed. In that case the lowest possible impedance in the supply provides for the optimum power transfer. That\'s why power supplies are designed with very low output impedances.. Same thing here, only the guy is trying to optimize his costs, rather than ultimate effectiveness of the cooling.

I have no idea what his heat sinks look like or how he plans to attach them, but adding a second heat sink will most likely improve his cooling. The only question is by how much. That can\'t be answered without more information. He talks about \"a\" heat pipe being near the top of the heat sink, but any one heat pipe has limited heat capacity, so still insufficient info.

--

Rick C.

-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209

 

[Ricketty C]

On Sunday, July 26, 2020 at 10:42:49 AM UTC-4, Commander Kinsey wrote:

On Sun, 26 Jul 2020 10:43:46 +0100, Martin Brown <\'\'\'newspam\'\'\'@nezumi.demon.co.uk> wrote:

On 25/07/2020 22:39, Commander Kinsey wrote:
On Sat, 25 Jul 2020 21:09:52 +0100, legg <legg@nospam.magma.ca> wrote:

Stacking CPU coolers will have a huge impedance mismatch on the second
one - essentially wasting most of its capability.

Since the heatpipes of each would be very close, the length of normal steel would be minimal.

WTF are you talking about \"normal steel\"??? Each heat pipe can only carry a small amount of heat, that\'s why they use many. Otherwise they would just use one pipe snaked along the fins.

It would be an odd construction - that had so much free real estate in
that particular direction, away from the heat source.

If you mean the computer case, it\'s not in one, the motherboard sits on
a bookshelf. It\'s a research machine sat in my garage made out of spare
parts, and no case was the right shape to fit a dual server board into.

It is antisocial to operate a fast PC outside an enclosure other than
for testing because it is a broadband emitter of RF interference to GHz.

I don\'t give a fuck. And that does explain why the monitor jumps a bit. There are 5 such open cased machines there. I wonder if that\'s why my neighbour has installed some kind of WiFi repeater in his garden Pissed him off without even trying!

You are starting to sound like a dick who used to post here under another name that I can\'t recall. What was it? You used to blow up stuff and get mad at the people who made it.

You\'d get a better result with a bigger mounting interface to the heat
source,

Can\'t be done, the CPU is a specific size that cannot be changed.

You can get oversize heatsinks using heatpipes to shift more heat from
the CPU to the finned metal.

But the interface between the CPU and the block would be the same size - the area of the CPU.

Yes, you are very definitely not understanding what is going on. If the interface to the CPU was the limiting factor in heat conduction, then you would not get better performance by stacking your heat sinks either.

Larger fans are quieter too. If all else
fails water cool the thing which is what the serious overclockers do.

Expensive, hassle, leakages....

And serious overclockers use liquid nitrogen, or submerse the whole thing in oil!

So you just want to do something lame and pathetic?

coupling the heatpipes to it in parallel, rather in series,
or just by pushing more air.

Heatpipes only work if the temperature of the source exceeds a certain
level.

The top of the current single one feels pretty hot, so the bottom of the
new one would also be.

The way to cool it is use a larger heatsink designed to match the CPU
surface and use the right compound to make the best thermal joint.

I just replaced the fan with a faster one. Lost 20C. I\'d forgotten how shit some fans are. 1100rpm to 4000rpm!

Yeah, it\'s not uncommon for people to rush to a solution before they have all the facts. Still wanting to stack heat sinks?

--

Rick C.

-+ Get 1,000 miles of free Supercharging
-+ Tesla referral code - https://ts.la/richard11209