Syn Jets!!!

Wang · 03-23-2005, 06:59 PM

I hope this belongs here........ A link to "Synthetic Jets"
http://gtresearchnews.gatech.edu/new...se/synjets.htm

Mount on a heatercore is my first thought.... Intercooler type tech.?
Maybe I should just get a spray bottle for now and mist my heatercore!

What do you guys think of this tech?

Bernie.

Hoot · 03-24-2005, 09:24 AM

Boy, if that isn't food for thought!

Imagine getting your hands on a piezo-electric transducer from say, an old ultrasonic parts cleaner and integrating it into a water block. With a fairly slow lateral flow rate through the block, the water as it passes between the transducer and the cups/pins/whatever shape lends well, would be excited both towards and away from them. I think you get the drift. Lots of parameters to work out and hoping the transducer wouldn't dump too much heat into the water, but just imagine...

Hoot

Jimbo Mahoney · 03-24-2005, 10:05 AM

Interesting Hoot..........

BillA · 03-24-2005, 10:14 AM

I have discussed this,
can't get any enthausam from the pros for the ultrasonics coupled to the CPU
freq, power, dedicated hi freq psu (25k to 70k ?)
and the transducers are NOT cheap

killernoodle · 03-24-2005, 02:20 PM

Very interesting.

MadHacker · 03-24-2005, 02:27 PM

Would it affect the CPU any?

Althornin · 03-24-2005, 04:21 PM

“We have so far been able to cool about 420 watts per square centimeter, and ultimately expect to increase that to 1,000 watts per square centimeter.”

Is it just me, or is this meaningless?
Without knowing at what delta T to ambient they do this, these numbers mean nothing...

MadHacker · 03-24-2005, 04:57 PM

Quote:

Originally Posted by Althornin

“We have so far been able to cool about 420 watts per square centimeter, and ultimately expect to increase that to 1,000 watts per square centimeter.”

Is it just me, or is this meaningless?
Without knowing at what delta T to ambient they do this, these numbers mean nothing...

Marketing at it's greatest

Ls7corvete · 03-24-2005, 04:59 PM

Quote:

Originally Posted by unregistered

I have discussed this,
can't get any enthausam from the pros for the ultrasonics coupled to the CPU
freq, power, dedicated hi freq psu (25k to 70k ?)
and the transducers are NOT cheap

HHmmm not encouraging. If you cant get it done I dont see anyone putting it together. Maybe we can get some info from you, transducer price, heat output, expected results etc?

I would guess you have limited results from such a system as it is still limited to the water temps. If your using a bunch of power to remove heat most people would expect sub-ambient similar to peltier. Even with peltiers swiftech has little market for them.

jaydee · 03-24-2005, 06:56 PM

I tried to figure out a way to shake the water inside the block before. Never thought about using ultrasonics. Interesting.

BillA · 03-24-2005, 07:05 PM

think on the TIM joint a while

Hoot · 03-24-2005, 09:52 PM

Given the mass of a copper base (body at rest), would the motion transfer through to the TIM making it lap the core/face junction, or walk out from in between?

Hoot

samualt · 03-25-2005, 02:36 AM

unregistered:
Why would transducers be expensive? You can buy all kinds of cheap transducers. Drive them with the signal generator of your choice. It's not brain science to make a plate vibrate using the peizoelectric effect. It's easy to drive stuff like that with a simple 555 circuit. You could build that whole device for under $20.

A good idea might be to put the transducer inside the top of the water block instead of "on" the water block. You want to shake the water, not the metal water block itself! Of cource, routing your wires out of the water block might be fun.
Not sure I buy the idea that it makes heat transfer better, but it could be a cheap and easy experiment for someone who plays with custom water blocks.
Also, it would be fairly easy to use transducers to push air through the radiator. But again, I don't think it really helps transfer heat. But I'm not a thermo-dynamasist either.

Also, you might check out Electrostatic Cooling. There is also an abstract at ECD. And here is Cool Chips PLC proclaiming to use similar technology (Thermotunneling) in a device that looks like a peltier but isn't.

Anyone else got some weird science for us?

Nugit · 03-25-2005, 04:54 AM

Interesting indeed...

And hey, if you got a transducer with the right frequency you wouldn't have to clean your block again

Hoot · 03-25-2005, 07:00 AM

I should have been more clear in my diagram that I was pondering a block with the transducer inside of it, not on the outside. As for the electrostatic idea, that is interesting also, but it would be hard, if not impossible to implement it in a water cooled environment since it would be near impossible to find a cooling solution with a high enough dielectric constant while maintaining the positive cooling attributes associated with water. In both situations, the question of impact upon the electronic functionality of the CPU while either being bombarded with piezoelectric induced energy, or high voltage fields in the case of the electrostatic approach, remains to be assessed.

Hoot

samualt · 03-25-2005, 03:48 PM

Hoot:
Obviously you couldn't use the electrostatic cooling inside the system in any way. I was thinking someone could use it on the radiator, mounted outside the case. It would be simple to make a fine wire mesh that is negativly charged and place it an inch away from the grounded radiator. This would cause a fine spray of negitive electrons that would hit the grounded radiator. If it really helps to exchange heat it might be worth a try.
Since the radiator is grounded it shouldn't introduce any ions into the system. Although, if you start smelling a sweet musky odor you might want to worry about the ozone generator you just created.

jaydee · 03-27-2005, 06:07 PM

Quote:

Originally Posted by Hoot

Given the mass of a copper base (body at rest), would the motion transfer through to the TIM making it lap the core/face junction, or walk out from in between?

Hoot

The purpose of the TIM is to fill in the micro crevices in the materials so the TIM would be trapped in those crevices and the rest might push out. This would be a good thing I believe. The TIM that remains in the crevices would possibly be stirred by the sound waves. Turbulence is good in the water so why not in the TIM?

Probably a lot I am missing here though.

BillA · 03-28-2005, 09:21 AM

the rheological properties of the TIM are of importance throughout the service life of the joint,
what is going to happen re the consolidation due to the ultrasonics ?
and understand that the transducer must be coupled to the fluid - not the wb

Wang · 03-28-2005, 06:12 PM

Quote:

Originally Posted by samualt

unregistered:
Anyone else got some weird science for us?

Heh, admittedly most of the posts are beyond my ken.
If I were to post only what I know.... I'd forget how to type!!

good reading tho!

maxSaleen · 03-28-2005, 07:25 PM

Quote:

I have discussed this,
can't get any enthausam from the pros for the ultrasonics coupled to the CPU
freq, power, dedicated hi freq psu (25k to 70k ?)
and the transducers are NOT cheap

Transducers simply create vibrations, correct (practically speaking, from what I've ascertained they actually move a fluid by contractions and expansions)? If so, for the purpose of creating turbulance, I'm sure lower cost options exist. This could be an interesting way to increase the performance of a waterblock and could make passive radiators a viable option.

Turbulance, to a point, equals cooling performance. This is part of the reason jet impignation blocks work so well. Imagine being able to achieve large amounts of turbulance while having a relatively free flowing block. There is potential here.

03-23-2005, 06:59 PM	#1
Wang Cooling Neophyte Join Date: Apr 2003 Location: Ottawa,ON Posts: 46	Syn Jets!!! I hope this belongs here........ A link to "Synthetic Jets" http://gtresearchnews.gatech.edu/new...se/synjets.htm Mount on a heatercore is my first thought.... Intercooler type tech.? Maybe I should just get a spray bottle for now and mist my heatercore! What do you guys think of this tech? Bernie.

03-24-2005, 09:24 AM	#2
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	Boy, if that isn't food for thought! Imagine getting your hands on a piezo-electric transducer from say, an old ultrasonic parts cleaner and integrating it into a water block. With a fairly slow lateral flow rate through the block, the water as it passes between the transducer and the cups/pins/whatever shape lends well, would be excited both towards and away from them. I think you get the drift. Lots of parameters to work out and hoping the transducer wouldn't dump too much heat into the water, but just imagine... Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-24-2005, 02:20 PM	#5
killernoodle Thermophile Join Date: Sep 2002 Location: Charlotte, NC Posts: 1,014	Very interesting. __________________ I have a nice computer.

03-24-2005, 02:27 PM	#6
MadHacker Cooling Savant Join Date: May 2003 Location: Okotoks, A.B. Canada Posts: 726	Would it affect the CPU any? __________________ "Great spirits have always encountered violent opposition from mediocre minds" - (Einstein)

03-24-2005, 09:52 PM	#12
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	Given the mass of a copper base (body at rest), would the motion transfer through to the TIM making it lap the core/face junction, or walk out from in between? Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-24-2005, 10:05 AM	#3
Jimbo Mahoney Cooling Neophyte Join Date: Jan 2005 Location: UK Posts: 96	Interesting Hoot..........

03-24-2005, 10:14 AM	#4
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	I have discussed this, can't get any enthausam from the pros for the ultrasonics coupled to the CPU freq, power, dedicated hi freq psu (25k to 70k ?) and the transducers are NOT cheap

03-24-2005, 04:21 PM	#7
Althornin Cooling Savant Join Date: Dec 2002 Location: Houston, TX, USA Posts: 221	“We have so far been able to cool about 420 watts per square centimeter, and ultimately expect to increase that to 1,000 watts per square centimeter.” Is it just me, or is this meaningless? Without knowing at what delta T to ambient they do this, these numbers mean nothing...

03-24-2005, 06:56 PM	#10
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	I tried to figure out a way to shake the water inside the block before. Never thought about using ultrasonics. Interesting.

03-24-2005, 07:05 PM	#11
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	think on the TIM joint a while

03-25-2005, 02:36 AM	#13
samualt Cooling Neophyte Join Date: Jan 2005 Location: Texas, USA Posts: 30	unregistered: Why would transducers be expensive? You can buy all kinds of cheap transducers. Drive them with the signal generator of your choice. It's not brain science to make a plate vibrate using the peizoelectric effect. It's easy to drive stuff like that with a simple 555 circuit. You could build that whole device for under $20. A good idea might be to put the transducer inside the top of the water block instead of "on" the water block. You want to shake the water, not the metal water block itself! Of cource, routing your wires out of the water block might be fun. Not sure I buy the idea that it makes heat transfer better, but it could be a cheap and easy experiment for someone who plays with custom water blocks. Also, it would be fairly easy to use transducers to push air through the radiator. But again, I don't think it really helps transfer heat. But I'm not a thermo-dynamasist either. Also, you might check out Electrostatic Cooling. There is also an abstract at ECD. And here is Cool Chips PLC proclaiming to use similar technology (Thermotunneling) in a device that looks like a peltier but isn't. Anyone else got some weird science for us?

03-25-2005, 04:54 AM	#14
Nugit Cooling Neophyte Join Date: Jun 2003 Location: Aalborg, Denmark Posts: 85	Interesting indeed... And hey, if you got a transducer with the right frequency you wouldn't have to clean your block again

03-25-2005, 07:00 AM	#15
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	I should have been more clear in my diagram that I was pondering a block with the transducer inside of it, not on the outside. As for the electrostatic idea, that is interesting also, but it would be hard, if not impossible to implement it in a water cooled environment since it would be near impossible to find a cooling solution with a high enough dielectric constant while maintaining the positive cooling attributes associated with water. In both situations, the question of impact upon the electronic functionality of the CPU while either being bombarded with piezoelectric induced energy, or high voltage fields in the case of the electrostatic approach, remains to be assessed. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-25-2005, 03:48 PM	#16
samualt Cooling Neophyte Join Date: Jan 2005 Location: Texas, USA Posts: 30	Hoot: Obviously you couldn't use the electrostatic cooling inside the system in any way. I was thinking someone could use it on the radiator, mounted outside the case. It would be simple to make a fine wire mesh that is negativly charged and place it an inch away from the grounded radiator. This would cause a fine spray of negitive electrons that would hit the grounded radiator. If it really helps to exchange heat it might be worth a try. Since the radiator is grounded it shouldn't introduce any ions into the system. Although, if you start smelling a sweet musky odor you might want to worry about the ozone generator you just created.

03-28-2005, 09:21 AM	#18
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	the rheological properties of the TIM are of importance throughout the service life of the joint, what is going to happen re the consolidation due to the ultrasonics ? and understand that the transducer must be coupled to the fluid - not the wb Last edited by BillA; 03-28-2005 at 09:28 AM.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)