Question about Amplifier Tube Cooling

[W8ER]

Guys, I ran across this great article on some cooling modifications made to a Hanry 5K and was unable to find it until a few minutes ago. In light of the discussion, I thought it might be interesting reading. Some of the techniques for both measuring the airflow and curing the "problem" was way over just interesting.

Cooling Mods for a Henry 5K

--Larry W8ER

[Rob K2CU]

A neat item to get to monitor your air temp is a low cost, dual display, indoor/outdoor digital thermometer.  Just for fun I used one to compare twmperatures at the top of my T4X-C with and without the little PC processor cooling fan that sits on top of the output tubes. Without fan, the air rose to 114 degrees F in a room with 68 degree air. ...and that was in standby.  with fan turned on, the temp dropped to 74 degrees.  OBviously, more testing is in oreder, such as comparing top mounted exhaust vs. rear mounted fan and blowing vs. sucking.


Another idea for comparing the color of the plate glow might be to build up an IR measuring circuit using a black plastic encapulated IR photo diode/transistor.  it could be mounted to look at the tube plate through the glass.  If cooling the plate stem actually takes heat out of the anode, then a drop in IR output should be measurable.  Multiple detectors could compare tubes.  This could also be used to verify if paiinting the inner walls of the Rf cage actually reduces back IR radiation heating of the plate.
It wouldn't give you any calibrated measurement, but would be good for comparisons.

[W2VW]

So back to the original question - from what you've read, do you feel that the plate structure in the vacuum glows less brightly when more air is passed by the outside glass envelope?  

T

Only is the inlet air is cooler than the outlet air....

[Steve - WB3HUZ]

This is purely speculation on my part, but I think there is a point at which increasing the air flow across a tube will not improve the dissipation or more effectively cool the tube. My reason is that for a given inlet air temperature, there is a finite limit at which the heat can be radiated by the plate, transferred across the vacuum to the glass and then reradiated by the glass into the air steam. Once the air flow pulls away all the heat the plate/vacuum/glass system can provide, the cooling limit is reached.

Once again, this is just speculation and "thinking out loud" on my part. But, if my speculation is basically correct, and the tube manufacturers knew this limit, their CFM requirements for a given tube would hopefully reflect it (probably be somewhere reasonably close to the limit). So, the question becomes, in the case of a 3-500Z, is the 13 CFM air flow specification at the "cooling limit" number I speculated above? I don't know. But, if it were, then passing more air across the tube would not change the color of the glowing plate.

How's that for some BS?

[W2VW]

Let the freakin' anode glow red. It was designed to work that way. Keep the bulb seals cool and you are set. If the anode is turning white due to over dissipation then you can blow all the cool air in Hoboken on it and still not help cathode bombardment.

[K1JJ]

This is purely speculation on my part, but I think there is a point at which increasing the air flow across a tube will not improve the dissipation or more effectively cool the tube.

Yes, it is probably a point of diminishing returns when the flow is big. But the air flow must get to the optimum point first. I found that when running the normal recommended air socket on an 8877 or 4X1, I had a very difficult time getting enuff air thru it to adequately cool the plate structure when piss beating the tube.... :lol:  

With my 8877 I put on a huge blower and it still seemed restricted/turbulent, the air flow was noisy and the air was hot. After drilling the holes the air cooled down noticably and was quieter.  So the bottom line is I had not reached the critical air flow in these examples and the holes helped to get there...  Again, I think the pins/seals need very little air being below the major heat source and the plate is where increased air helps. This indirectly keeps the pins cooler too by pulling would be plate conduction heat away from them.

This can all be solved by building a class E rig or picking up an SB-230 with a conduction cooled final, caw mawn.

T

[w3jn]

The SB-230, however, is one amp you *don't* wanna piss-beat.  The 8873 toobs are rare and pricey, and don't take well to a good beatin'.

You fergot the other option - go with a sand-state leenyar.  'S what I've done... but I must say it is NOT silent due to xformer lamination hum.  Also after a long old buzzard the fans will suddenly come on, sounding like a B-29 spooling up for takeoff.  Fortunately, that's a rare occurrence, but it's enough to startle a guy.

73 John

[WA1GFZ]

The cooler the air passing over the tube the more you can dissipate
up to the point where the plate melts. Heat travels to cold and wants to go up since heat contains upsadasium.  Remember the old days when guys
would mount a 6L6 upside down in a pan of water.
I bet you could really push a tube to the limit of emission if you keep it cool. Look at the pulse ratings of a tube. Heck the pair of 6DQ5s in my viking 2 have been switching a half amp peak at 1550 V since 1983 and show no signs of wear.  A plate can run Red as long as the glass doesn't melt. Yellow or white may be a concern.

[KL7OF]

If you are into chimney cooling...Go to the sporting goods store and get a pyrex glass for a coleman lantern....They come in many sizes and cost less than 10 bucks ......

[K1JJ]

The SB-230, however, is one amp you *don't* wanna piss-beat.  The 8873 toobs are rare and pricey, and don't take well to a good beatin'.

You fergot the other option - go with a sand-state leenyar.  73 John

Yep, the SB-230 is a popular amplifier lately for the local group on 6M...  One guy has converted his over to a mono-bander and is doing two more. Yes, the tube is hard to find, though they are buying the amps intact for reasonable prices right now.  They put out 600W silently on 6M when set up right.  And you don't wanna drill and inhale the tube's berlium? insulation. These days I doubt Heath wud take that risk with ham kit guys..

Yes, a solid state linear is a great way to go too. Price can range all the way up to the Icom PW-1 ?  that costs upwards of $7-10K+ ?...  Or luck out and get a surplus unit ot build one.   Push the botton on your new ricebox, it changes the antenna AND changes the linear's band. Call CQ on each band 160-6M every 2 seconds until your finger gets tired, caw mawn.

Then again, it's easier to get on the cell fone or internet, but many of us like to row around town with our vacuum variables or bread slicers.

T

[Art]

the GI1b russian triode is a good substitute for the 8873 in the sb-230. . .yes, you will need to pressurize the cabinet with a muffin fan but it is a nice conversion. . . 'bout 500W out with 30W drive . . . the tube is about $20 on epay. . .

[W1RKW]

Something I've always thought about doing with my car was to insert a couple of those solidstate peltier junction thermal cooling devices into my intake of the engine and reduce the intake temp during the summer and squeezing more hp out of the engine.  I've always noticed the cars performance changes when it's hot out but always ran best in the 70 degree range or lower.

One could do something similar to this and reduce the ambient temp by 10's of degrees and blow that cool air over a tube.  I'm sure a drop in ambient temp across the envelope would make a significant difference in performance compared to normal room temp especially during the warm months of the year.  

http://www.allelectronics.com/cgi-bin/category.cgi?category=450&item=PJT-4&type=store

[WA1GFZ]

Yea that sounds great but you might need 100 amps at 12 volts to generate enough cold air to make a difference.  They are cool parts but power hogs.

[K1JJ]

Yep, agreed.

I used one to cool my homebrew CCD camera for astronomy photo use.  It took some time to cool down to -20F, and  the current was about 30A at 13.8V.   The area it cooled was maybe a coupla square inches of aluminum slab-bacon that the tiny CCD chip mounted on.. [cold= low thermal chip noise]   I ran antifreeze thru the aluminum/Peltier to pull the heat away. The tank was about 20 gallons and warmed up at the end of the night from the Peltier "cold finger" heat transfer.  Not a very efficient system, but at that small scale it was FB for the job.

T

[W1RKW]

maybe pumping cold air from an AC unit would be more efficient for tube cooling...

[WA1GFZ]

How do you plan to turn the compressor?

[W1RKW]

20Amps at 13.8Vdc for a 100watt pep slopbucket signal. Yep that's not to effiencient either but what the heck.   I guess BTU's is the issue here.  An AC unit probably would be best for tube cooling.  Modify a deep freeze unit to cool the tube if you're gonna use external duct work. Modify it with the plumbing then fill it up with some thermal mass and you get yourself the best tube cooling unit around.

[WA1GFZ]

1KW of heat is a bit over 3000 BTU if I remember

[Steve - WB3HUZ]

Looking at the Eimac data sheet, it appears moving more air over the 3-500s would not change the color of the plates because the air flow is solely for cooling the base and plate cap seals. (See the highlighted section below)

COOLING - Forced-air cooling is required to maintain the base seals at a temperature below 200º C, and the plate seal at a temperature below 225º C. Air flow requirements to maintain the above maximum temperatures are tabulated below. (For operation below 30MHz)

Code:

Base to Anode Air Flow

Anode                         Air Flow                     Pressure
Dissipation                    (CFM)                         Drop
(Watts)                                                     (in-H2O)

300                             6.6                         0.023
400                            10.3                         0.052
500                            13.0                         0.082

The anode of the 3-500Z operates at a visibly red color at its maximum rated dissipation of 500 watts.

In all cases, air flow in excess of the minimum requirements will prolong tube like. NOTE: Two 3-500Z tubes in a single amplifier, chassis mounted, may be adequately cooled by use of a fan so mounted as to pressurize the space below the sockets. Fans suitable for use ...... <cut>

In all cases, the only criteria of proper cooling is the temperature of the tube seals. (Emphasis added) Tube temperatures may be measured with the aid of temperature sensitive paint, spray , or crayon.

Your friend with the modified SB-220 may want to up the air flow, since the specifications given above are for 30 MHz and below.

[Steve - WB3HUZ]

1KW of heat is a bit over 3000 BTU if I remember

I don't think there is a direct conversion because BTUs are a units of energy and Watts are energy per time (joule per second). So, one BTU is equal to one kilowatt-hour.

[WA1GFZ]

We looked it up at work when I was wondering how much KWH it would take to match the heat of my 20,000 BTU kerosene heater. I saw the number wiritten on the wall yesterday so I wouldn't forget.
It is a bit over 3400 BTU for 1 KWH so I would need about 6 KW of base board to match the heater.

[Steve - WB3HUZ]

Yep, 1 kW is 3415 BTU/h

[W8ER]

Yep, 1 kW is 3415 BTU/h Give a man a fire and he'll be warm for a day; set him on fire and he'll be warm the rest of his life.

Steve, you seem to have a thing for heat!

 :evil:


--Larry

[VA3ES - Piss-Weak Ed]

 Steve, you seem to have a thing for heat!

Yeah.... especially "cash heat" :badgrin:  :p .

[K1JJ]

Interesting summary on the 3-500Z cooling, Steve.

According to a Google search, "Soft solder melts at 200 - 250 C and consists of tin and lead in varying proportions, with sometimes a trace of antimony. "

Some hits said about 180C - 200C.  

So looks like when the filament pin solder melts, the critical 250 C  degrees maximum is about to be reached on the seals according to the information.  So, a good thing it melted and shut down the tube.

Well, it looks like the majority thought [including me] that the color of the tube gets less bright with more air flowing. Now the Eimac info seems to infer that it has no effect on the plate color.  If so, the plate structure must rely solely on radiation cooling.

So it's back to my original first comment that the plate [in the vacuum] does not know what the outside air is doing.

However, with an external anus tube, air cools both the seals and plate structure.

This is what makes a good topic for discussion - we all learn something new.

T