Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: Prev 1 [2] 3 4 5 6 7 Next   Go Down
  Print  
Author Topic: How to calculate water cooling systems?  (Read 13075 times)
W9IQ
Member

Posts: 1707




Ignore
« Reply #15 on: July 27, 2017, 07:39:07 AM »

We did water cooling of the tooling in diamond manufacturing presses (carbon under very high heat and high pressure). With a proper heat exchanger design, I don't see the power density as a limiting factor.

- Glenn W9IQ
Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
N3QE
Member

Posts: 4878




Ignore
« Reply #16 on: July 27, 2017, 07:42:00 AM »

Here's some rules of thumb for liquid cooling the YU-191 which is in the ballpark of required dissipation:



I'm trying to think of how conduction area on a tube (a few feet of tubing spiraled around the anode) vs conduction area on a semiconductors small thermal interface block (much smaller, presumably a shorter water path) will impact the required flow rates.
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #17 on: July 27, 2017, 08:41:56 AM »

That's an excellent write-up by W9IQ.

I would go with water and not oil. I tried to oil cool a 4CX25,000 for a HV linear regulator dissipating 10 kW. As Glenn pointed out the specific heat of oil is much less than water and so more must be pumped. And oil is more viscous and that means higher pressure.
Logged
KM1H
Member

Posts: 2465




Ignore
« Reply #18 on: July 27, 2017, 12:19:32 PM »

Just run a big water cooled tube way below its max, such as the commercially popular 3CW20000A7 which can be sometimes found for the cost of a new Eimac 8877. The 750W filament power may bother a few Roll Eyes
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #19 on: July 27, 2017, 04:04:21 PM »

I meant to type 4CW25000
Logged
N2WQ
Member

Posts: 99




Ignore
« Reply #20 on: July 28, 2017, 07:06:41 AM »

So I contacted one of the PC cooling manufacturers and got the following response:

EK-CoolStream XE 480 (Quad) https://www.ekwb.com/shop/ek-coolstream-xe-480-quad
with four EK-Furious Vardar EVO 120 (3000rpm) https://www.ekwb.com/shop/ek-furious-vardar-evo-120
can dissipate just a bit less than 1000W/10°C coolant rise.
With 2500W you can expect the coolant temperature 25°C above ambient temperature.

EK-CoolStream CE 560 (Quad) https://www.ekwb.com/shop/ek-coolstream-ce-560-quad
with four EK-Furious Vardar EVO 140 (2500rpm) https://www.ekwb.com/shop/ek-furious-vardar-evo-140
can dissipate ~1050W/10°C coolant rise so expected coolant temp is 24°C above ambient temperature.

With D5 based pump and ZMT tubing you can safely run the coolant temperature in the range of 50-60°C.
Logged
N3QE
Member

Posts: 4878




Ignore
« Reply #21 on: July 29, 2017, 06:54:04 AM »

So I contacted one of the PC cooling manufacturers and got the following response:

EK-CoolStream XE 480 (Quad) https://www.ekwb.com/shop/ek-coolstream-xe-480-quad
with four EK-Furious Vardar EVO 120 (3000rpm) https://www.ekwb.com/shop/ek-furious-vardar-evo-120
can dissipate just a bit less than 1000W/10°C coolant rise.
With 2500W you can expect the coolant temperature 25°C above ambient temperature.

EK-CoolStream CE 560 (Quad) https://www.ekwb.com/shop/ek-coolstream-ce-560-quad
with four EK-Furious Vardar EVO 140 (2500rpm) https://www.ekwb.com/shop/ek-furious-vardar-evo-140
can dissipate ~1050W/10°C coolant rise so expected coolant temp is 24°C above ambient temperature.

With D5 based pump and ZMT tubing you can safely run the coolant temperature in the range of 50-60°C.


This is about the ballpark I would expect from their small sizes. There are some short Russian videos on youtube of water-cooled LDMOS amps with PC-CPU radiators, and I think they are mostly relying on the thermal inertia of the water when run at their limit.

I think a window-air-conditioner sized coil, or a small transmission cooler, would get closer to the 2500W level.

This raises some interesting prospects for cooling multi-2 or multi-multi contest stations without the heat actually being dumped into the room. In my area in a big contest we just open all the windows in wintertime and even with all the windows open, its like a sauna in summertime. For starters the radiator(s) could be placed outdoors but you'd need antifreeze for winter if there wasn't a contest going on :-).
Logged
WA7PRC
Member

Posts: 1837


WWW

Ignore
« Reply #22 on: July 29, 2017, 07:06:04 AM »

IIRC, at the world's largest OEM of RF-driven sealed CO2 Lasers, a water-cooled kilowatt RF source (Class E, 40.68 MHz) used two ARF1500 FETs and was cooled with ≥ 1 gl/min of water. The cold plate used 1/4" OD copper tube pressed into the bottom of the aluminum cold plate. The tubes made "S" bends under the FETs. The mounting surfaces for the FETs were specified for flatness and finish. We "lapped" the FETs into place, resulting in using the tiniest amount of thermal compound. We never had cooling-related failures (using the sensing scheme I previously mentioned).

Bryan WA7PRC
Logged
KB2WVO
Member

Posts: 382




Ignore
« Reply #23 on: July 30, 2017, 02:23:56 AM »

car or truck heater cores lot cheaper. do the same thing..
Logged
K6AER
Member

Posts: 4666




Ignore
« Reply #24 on: August 01, 2017, 04:42:00 PM »

I like that idea.
Logged
W0MLD
Member

Posts: 9




Ignore
« Reply #25 on: August 02, 2017, 04:05:37 AM »

Am I nuts here or are you not actually needing to handle 2500w of heating?  Aren’t you going to be running say 2500w but 1500 of that will go out as RF?
Logged
N2WQ
Member

Posts: 99




Ignore
« Reply #26 on: August 02, 2017, 04:06:20 AM »

I like that idea.

I looked at automotive radiators and found out that they are not that much cheaper. You can get a quite large computer radiator from eBay for about $100. The benefit of the PC radiators is that they have a shape that works well for the amp's enclosure and, more importantly, already have a design that allows for easily attaching 120mm fans.

Separately, we learned that these types of amps are reasonably efficient, much better than my worst case assumption of 50%. We have a 2m version running at over 70% efficiency. If this efficiency translates to HF, then the radiator needs to dissipate about 1,000 Watts worth of heat.
Logged
N2WQ
Member

Posts: 99




Ignore
« Reply #27 on: August 02, 2017, 04:10:30 AM »

Am I nuts here or are you not actually needing to handle 2500w of heating?  Aren’t you going to be running say 2500w but 1500 of that will go out as RF?

Going back to the original design objectives- the goal is to produce an amp that can do 2500W out with absolutely no time limit. As a RTTY contester I'd like to have plenty of headroom to ensure this amp will last literally forever. I don't want to ever have to repair another amp.

Thus the starting point was sizing for worst case scenario- if we want to produce 2500 out and assume 50% efficiency then we need to dissipate 2500W worth of heat. As we just learned from the 2m version of the amp (single LDMOS chip), the actual efficiency is a bit over 70%. Of course it is yet to be seen if this type of efficiency will stay at HF.
Logged
W9IQ
Member

Posts: 1707




Ignore
« Reply #28 on: August 02, 2017, 04:59:10 AM »

I would recommend the use of a radiator designed for this purpose. In addition to built in fan mounts, the plumbing connections are more suitable for the application. There is also less concern about laminar flow reducing the cooling capacity.

- Glenn W9IQ
Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
KH6AQ
Member

Posts: 7718




Ignore
« Reply #29 on: August 02, 2017, 05:17:33 AM »

A SS amp designed to deliver 2500 watts with 50% efficiency will dissipate 2500 watts. Reduce the RF drive so the amp delivers 1500 watts and the efficiency drops to 39% while dissipation is 2350 watts. Set the amp up to deliver 1500 watts at 50% efficiency and it will dissipate 1500 watts.
Logged
Pages: Prev 1 [2] 3 4 5 6 7 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!