| klm1 |
Gentlepeople:
being of an inquisitive nature, as well as a tube noob;
could someone tell me why I should not pursue this line:
[www.tomshardware.com/2006/01/09/strip_out_the_fans]
Knowing as i do that liquids cool better than air,
could someone advise me on this? i.e.,high frequency, high voltage,
possible fire hazards,etc.,etc.,etc. . .
I can see potential for very long lifetime of tube,
possibly sweeter(?!?) sound? |
|
|
| SY |
Karo syup will definitely give you a sweeter sound. So will Hershey's, albeit with a certain darkening.
In reality, something like a silicone fluid will work, preferably one of the fluids made for diffusion pumps. |
|
|
| tubelab.com |
Sweeter sound? Smells like french fries to me.
Seriously, I know some ham radio operators that cool their tubes with water! One of my friends is running 2C39's with their plates DIRECTLY submersed in water. These tubes are made of ceramic with the metal plate exposed. There is a heat sink attached to the plate which is normally cooled with forced air. They claim that with water cooling they can almost double the 100 watt plate ratings. The water is circulated through a heat exchanger before recirculation.
You would think that water and 2000 volts would make a bad combination, but PURE water is not conductive. They use distilled water and change it when there is any cold tube current. About once a month. I don't think that I would try this, or recommend it, but I have seen the amplifier, and it has been running for a couple of years. We are talking about almost 1KW of power at 1296 MHz.
I have seen high end gamers PC's with water cooled Pentium chips. There is a radiator on the back of the case. This allows operation of the chip in excess of its speed ratings (overclocking). |
|
|
| klm1 |
Yes i know about transmitter tube cooling. :hot:
The site i referenced says they used "pure" water,
it caused problems after a while;
says they had to "isolate" the RF capacitance on the main CPU chip,
when they tried oil.
I was just interested in possibly submersing the tubes upside down in the oil, as a cooling substitute.
I can guess the hazard from grease getting on hot surfaces-
I don't want to "guess at" RF/HF complications w/hot oil vapors:hot: :hot: :hot: |
|
|
| DigitalJunkie |
HAM's used to dunk Metal 6L6's in oil,and crank the power up!
I guess they could get 80-100W from one 6L6! :hot: |
|
|
| Dave Cigna |
I think it should work fine, though I don't recommend vegetable oil as it will go rancid eventually. I once worked for a company that made thick film resistors. We would do some testing with the resistor in an aquarium full of light motor oil. It worked fine and there was no problem even at several kV. (No, this is not how maximum rated power dissipation was determined. :xeye: )
There's some indication that microphonics and resonances of the glass bottle can be heard in very revealing systems. Some people (whom I trust and respect) swear that globe tubes sound better than identical tubes in an ST bottle. Dunking the tubes - or even the entire amp - in an oil bath ought to put an end to that issue.
-- Dave |
|
|
| Brian Beck |
Here's one commercial liquid-cooled amplifier:
Von Gaylord Uni (Sea Urchin)
At $48K a pair, I doubt there are many of these in circulation.
Liquid cooling of large RF power amplifier tubes goes way back. For example, in the 30's, WLW cooled its 500,000 watt AM transmitter tubes with water that was sent to an outdoor pool with fountains for evaporative cooling. |
|
|
| beamnet |
Unfortunatelly, i am at a university with too much computer NERDS.
before there were commercial water coolers availeable, they used old camping refridgerators. They actually froze their processors in a container with silica beads to avoid water condensation and ice forming.
you could cool tubes by wrapping a copper tube around it with cooled liquids from such a freezer...i think... |
|
|
| amperex |
| Oil cooling is really a waste of effort on glass tubes. If one was to use a 3CX100 that has a metal exposed plate, it would be an option over forced air cooling. |
|
|
| poobah |
klm1,
I used to build undersea brushless motors. To keep the seawater out (and from crushing the motor) we filled them with mineral oil. Conductivity, or lack thereof, and all that was good... even great.
1) All wiring must be teflon as the oil strips the plasticizers (guessing) from normal PVC's... This leaves the PVC brittle and the oil polluted. Polyethylenes, polypropylenes, nylon, amides, imides, and most importantly, epoxies seemed to be happy.
2) Caps are problematic... you have to examine their construction. Orange Drop types worked well as they are epoxy and polypropylene based. Electrolytics need to be dipped in epoxy several times AFTER being soldered... the oil plays havoc on the rubber end seal... we gave up on electrolytics anyway... they don't hold their shape well at 5000 PSI. Caps wound or covered in tapes need to tested to ensure the adhesives don't break down.
3) Resistors seemed to be no problem across the board... though we never used a carbon comp (wouldn't anyway).
4) Potentiometers are a no-no unless they are wirewound and even then we didn't trust 'em.
5) Connectors in general performed well, although we limited ourselves to high-contact-pressure box or socket types... no molex KK and the like.
** Silicone oils might work well as far as compatibility with plastics/ polymers is concerned. Can't say much though... we had to glue stators assemblies and other gizmos... so silicone was not allowed in the facility.
** Any elastomers must be tested for swelling, shrinking, hardening, or oil pollution... many fail... use Viton where you can.
** Oil does cool things well, but not as well as intuition tells you. It does not convect (flow) as much as air does.
Go for it, and lemme know how things work. I am still working out my 350 kW diesel-turbine-powered-vacuum-tube-SE-bass-guitar amp. I have suffered years for my music; it's time everyone else did.
;) |
|
|
| Brian Beck |
| Olive oil for that organic sound! This is a pretty clever implementation of oil damping, but I have to wonder why this Zenn fellow didn't keep the jar upright, with tube pins pointing out the top. The sockets and PC board (or point-to-point wiring) could surely be inverted easily enough. I wouldn't want to risk a leak with gravity pulling on the oil against the seal 24/7, with temperature and pressure cycling. |
|
|
| poobah |
Brian,
The holes in the upper mica insulator will allow the electrons to fall to the bottom of tube if used upside down... leaves the bottom heavy and the midrange hollow. |
|
|
| Giaime |
| quote: | Originally posted by poobah
Brian,
The holes in the upper mica insulator will allow the electrons to fall to the bottom of tube if used upside down... leaves the bottom heavy and the midrange hollow. |
Great Poobah,
with this you have explained all the differences between Fender and Marshall amps :D :D :D |
|
|
| poobah |
You're a smart man Giaime!
Both amps have good treble... treble electrons are faster and don't fall through the holes...
:D |
|
|
| klm1 |
Gentlepeople:
1) Obviously, I have not done my homework!:o
2) DigitalJunkie-yes, I have had Ham friends tell me this(!)
3) Dave Cigna-I wonder how much cooling by metal ends
of the resistor vs. body of resistor...also,harmonics of tube
resonating, overlap into audio bandwidth, horn design tells
us shape matters much(!)
4) Brian Beck-waaaayyy out of my league :bigeyes:
5) Beamnet-kool(!) idea
6) Amperex-that was the idea of this thread, wondering if it
(WAS) a good idea. Not being an engineer myself(although
my family has had several), it just seemed practical.
7) Poobah-OK, here i go;
A)Yes, PVC is not chemically as cross-linked as epoxy/fiber-
glass resins;
B) Fascinating-high pressure does wreak havoc in
strange places:eek:
C) Check out [http://www.surplussales.com] for your hi end
amp needs; especially heatsinks,enclosures,meters.
8) commstech: that's more in line w/what I had in mind.
Thanks for the lift up; next time I will do my research
first:clown: |
|
|
| Miles Prower |
Liquid cooling will work only with external anode, xmtr VTs such as those used in UHF power amps, or that Gaylord "Sea Urchin". Tubes in glass bottles won't be helped by this at all since the main means of cooling these types is by radiation. Sure, you might be able to keep the glass from softening, but the innards will still burn out.
Liquid cooling isn't going to get you 100W from a pair of 6AQ5As, at least not for very long.
This isn't like SS, where you can add a bigger heat sink and horse up the power. |
|
|
| commsysman |
| Many high power TV transmitters date back to the 1950s and have special tubes that actually have water cooling designed into the VERY LARGE high power transmitter tubes. This requires PURE non-conductive water, and one either needs to change the water when a monitor indicates too much conductivity OR a de-ionizing filter system to keep the water from accumulating impurities. In my opinion, trying to use the sort of scheme you descibe is foolish and unsafe. I suggest you use finned aluminum heat sinks on the tubes in combination with forced airflow; this will do a good job without the safety issues and expense of liquid cooling. |
|
|
| klm1 |
Please don't think I have any purpose/intention of doing this-My only intention was to raise the possibility of its being done PROPERLY!
That being said, I don't think the proper way is to"burn"
up tubes.
I have read the safety info on this site, and Gary's tubelab.com
site(Thanks, Gary!) has really opened my eye's to the dangers.
I understand the idea of water cooling, and the chemistry of it-
it doesn't take massive amounts of ions to cause technically pure
water to conduct !!! |
|
|
| Tweeker |
Pearl Tube Coolers. Not sure how well they work. The seals are one of the more important parts to cool if you are using forced air cooling.
Heres a pic of a water cooled GU23A triode, along with a high vacuum HV Machlett diode, a couple of GM-70s and what is likely a radar choke. The pipe connections are for grid cooling water.
| quote: | I understand the idea of water cooling, and the chemistry of it I understand the idea of water cooling, and the chemistry of it
doesn't take massive amounts of ions to cause technically pure
water to conduct !!! |
Your right, it doesnt take many at all, and things may get interesting once DC electricity is added. Water doesnt stay technically pure all that long in most enviroments either.
Setups where water is used as a dielectric are often heavily and continuously filtered, it is an undertaking really only for specialized physics laboratory use etc. |
|
|
| gorgon53 |
convection cooled tubes - use a chimney, you can also use a fan SUCKING the air from top (DONT blow, sucking the air away works better and reduces thermal stress).
watercooling tubes with a metal can - circulate the water at 0.5m/sec-1.5m/sec in a copper coil wound around the tube.
water cooling of tubes intended for watercooling - use the original coolerassembly and use a polyethylenehose for isolation, use a min. length of 0.8m/kV for destilled water or at least 1.4m/kVp
for tap water (tapwater MUST be drained, dont circulate), consult the datasheet or use at least 1L/min/kW.
contactcooling - has high straycapacitance (even with expensive isolators) but works for tubes with external anode,
consult the datasheet for the highest hotspot temp (usually around 200degC)
oilcooling of glasstubes - have no expirience with it, but my guess is it wont bring much (low thermal capacity, thermal stress, dielectric losses), i think it is better to use a chimney instead, but if you want to try use at least transformeroil, use pressure regulation and overpressureguards and prevent aircontact. |
|
|
|
