>From a newbie to SSPA heat management:
Are there any on-line references with engineeriing information to design
retrofitted water cooling hardware to existing large Heat sinks? Are there
such things to buy, as aluminum tube extrusions, say 1/2" ID, thick wall,with
an outside flat?
Like:
http://images1.hellotrade.com/data2/WN/WF/HTVENDOR-3105392/type1-250x250.jpg
If these could be retrofitted to the flat heatsink, close to existing power
devices, would bonding surfaces have to be de-anodized? Polished? How to
determine heat removal numbers? Would, say, half gallon per minute room temp,
through a 1/2" tube, with flat, say 4" long, bonded to a heat sink remove a
huge amount of watts?
Hmmm my questions about cooling RF amps, is very similar to the Overclocking
enthusiasts solutions to cooling CPUs. So here is some information from the
QRO CPU engineers:
http://www.overclockers.com/forums/archive/index.php/t-210844.html
I picked up a bunch of Pineapple Technology SS PAs, mostly bench prototypes,
but supposedly in working condition. lots of BLF 872 and 861s mounted on 7"
x 7" x 2-3/4" or 7 x 14 ,etc.
Many were individual heatsinked board prototype fixtures from this (or
similar) Pineapple KW module:
http://www.ptibroadcast.com/pdf/pa1ku.pdf
Whole sink is extruded, anodized, with fins that are 2", standing proud of the
3/4" thick aluminum base. Many of the loose boards that are in the ~ 150 watt
category, others are in the 600 Watt category ( BLF 872s) and are mounted on
anodized ultra flat plates, drilled tapped, to mount on the large heat sink.
Is there a huge difference in thermal conductivity between active packages
screwed directly to machined landings in heat sinks, vice screwed into machined
landings on matching aluminum flat panels the size of the board, which are
screwed to the heat sink?
Best Regards,
Pat Barthelow; Echoes of Apollo
apolloeme@gmail.com apolloeme@live.com
"The most exciting phrase to hear in Science, the one that heralds
new discoveries, is not "Eureka, I have found it!" but:"That's funny..."
----Isaac Asimov
MOONBOUNCE YURI'S NIGHT
> From: jd0@broadsci.com
> To: amps@contesting.com
> Date: Mon, 30 Apr 2012 10:19:46 -0400
> Subject: Re: [Amps] Tubes vs. Solid State (was) Expert Amps 2K-FA:
> AnyOpinions ?
>
>
> > Based on my experience, I agree with your comments above. First, the
> > solid-state amps I've used are not very power-stable with PA deck
> > temperature. In the broadcast industry, output power must
> > generally stay
> > within a reasonably tight tolerance.
>
> In solid state broadcast transmitters, there is always APC (automatic power
> control). The exciter or IPA power output is varied dynamically, usually
> under microprocessor control, to keep the final TPO at the setpoint. It is
> not uncommon for the final amplifiers to drop in gain as they come up to
> normal operating temperature. 1 dB wouldn't be unusual, 2 dB probably would
> be.
>
> > OTOH, solid-state devices have
> > been used in
> > broadcast transmitters since the late 1970s (e.g., Harris MW-1).
>
> The Harris MW-1...now there's a transmitter I'd like to forget...
>
> > Seeing output power change with temperature is not that big
> > of a deal as
> > long as the amp remains reasonably linear. But it's an
> > extreme annoyance to
> > me as is the sound of jet noise coming from the amp when I'm
> > engaged in a
> > 3-5 CW exchange. IMO, the greatest improvements to be made
> > to ham-class
> > amps is in the area of cooling and thermal power drift.
>
> The obvious solution here is liquid cooling.
>
> > In VT amps, we can rely on high-volume designs to keep air
> > moving at low
> > sound pressure levels. Perhaps someone can create a similar
> > design to
> > create a finned heat that allows for a high volume, but low
> > velocity of air
> > to keep the PA deck cool.
>
> High volume/low velocity only works when you have a suitably large surface
> area of heatsink. Heatsinks aren't cheap, and, if the goal is to maintain a
> small form factor, you have design criteria that are mutually exclusive.
> Again, liquid cooling makes good sense.
>
> > Another factor affecting users of some VTs is that SS devices
> > are just as
> > prone to go obsolete.
>
> Very true. However, modern SS manufacturers are usually very responsible
> when it comes to putting out EOL (end of life) notices for devices well in
> advance. That doesn't seem to be the case with hollow state.
>
> > If the MRF150 was to go obsolete
> > tomorrow, probably
> > enough existing Asian sources can pick up that demand
>
> Be very wary of buying SS RF devices from overseas. There is so much
> counterfeiting, re-labeling of devices, selling of e-waste, etc., going on,
> that you never know what you're going to get. I've been burned several
> times.
>
> When it comes to selecting devices, learn from what we've seen in the tube
> realm. The devices that have a high usage in commercial fields will likely
> be around longer than those that only the hams are using. 8877's are still
> being made, and likely will continue to be made until the commercial world
> stops using them. Same with LDMOS FET's. MRF151G, BLF278, SD2932, etc. are
> still used in large quantities in broadcast, medical imaging, etc., so
> expect them to be available for quite some time. Picking the latest and
> greatest kilowatt-class ruggedized FET from Freescale or NXP might seem like
> a good idea initially, but unless the device gains acceptance in the
> commercial world, don't count on it being around 10 years down the road...
>
> --- Jeff WN3A
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|
