The AM Forum
March 28, 2024, 10:42:49 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2 3   Go Down
  Print  
Author Topic: Question about Amplifier Tube Cooling  (Read 36843 times)
0 Members and 1 Guest are viewing this topic.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« on: March 16, 2005, 01:21:03 PM »

The 6M guys are axing good questions lately.  One of them is running a converted SB-220 [pair of glass 3-500Z's] doing 1300W out on 50 mc - not bad.

He axed about cooling. The SB-220 does not use chimneys, rather some fans blowing and sucking air. I told him that chimneys would be better for the tube seals, glass, etc.

But then he axed whether the plate structure inside the tube would run less red with better air flowing around the tube. At first I said "yes", but now I'm not sure.  I've realized that the plate structure is in a vacuum and does not know what the outside air is doing. It uses mostly heat radiation for cooling and the hot air outside gets swept away providing better outside glass/seal protection..

Any ideas on this?   Maybe it depends upon whether the tube is in an open area or is surrounded with a glass chimney. ie, if the air hangs around the tube getting hot with little air flow, does if re-radiate back into the plate structure causing it to get hotter/redder?  Or, maybe it's just a matter of degree - stagnant hot air will  result in less heat radiation transfer by blocking/buffering the heat vs: cool air? A thermal dynamics question I think.

I seem to remember that when running 4X1's with chimneys, whenever I increased the Variac for more blower air, the plate structure ran cooler [less red] but maybe  it was when I was in the Bizzaro Whirl where everything is backwards.  

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W2VW
Contributing
Member
*
Offline Offline

Posts: 3489


WWW
« Reply #1 on: March 16, 2005, 01:43:09 PM »

Nice subject to generate interest Tom. Keep the red plates. They will eat any gas. It's the little metal thingys on the bottom which have to pass through the glass that are the worry. Double that for the SB220/221.
Logged
W8ER
Guest
« Reply #2 on: March 16, 2005, 04:00:04 PM »

Tom,

I'm no expert on tube cooling but I stayed at a Holiday Inn Express last night.

The 3-500Z is cooled by radiation of the heat from the plate structure. Increasing the air flow around the tubes will help, obviously. There are two theories about cooling the 3-500Z tubes. That may be due partially because increasing the airflow around the tubes also removes heat from the metal parts of the enclosure immediately around the pubes also!

First is using a muffin fan, the way it was done in the Heath SB220/221. An article praising the cooling design by the Heath engineers said that not only did the method satisfy cooling around the plates of the pubes but the area of the chassis that was cut out beneath the chassis, and the proper orientations of the filament pins, allowed them to be properly cooled as well. This apparently was not an accident! My only comment here is that I have never heard of the solder melting and running out of the filament pins on a Heath SB220/221! Harbaugh Electronics offers an improved fan blade for that beast and if I had my old SB220, I's pop for one of those puppies.

Second, the old chimney and air socket combination. I'm not certain that this would be my choice, if I were designing one. Besides a problem with chimney design and a dearth of different shapes for 3-500Z chimneys, it don't seem to work da best. For instance my Henry 2K4 had a problem with the solder melting and running out of the filament pins. This can be unnerving in the middle of an old buzzard! The tubes were being run properly, within ratings but the indication was that the cooling wasn't working! If I remember properly, and that is tough with increasing age, the chimneys tops were slightly belled in and my belief is that that shape restricted air flow. Eimac chimneys, much more expensive, had fairly straight sides and the opening at the top was the same size as the opening at the bottom.

I did wish to point out that most "HAM" amps, that use 3-500Z's, use the Johnson "cheapie" ceramic sockets, lots of ceramic material with a few small hole to allow air to pass through. . Henry does .. Ameritron uses sockets made from glass baord (real cheapie)! The Eimac "air sockets" are very different and very expensive. they use an insulating ring and the contacts are fastened to the ring. The pins are open, sort of hanging in air. As I recall the Eimac sockets run upwards of $100 each! The use of "other" may account for the poor cooling that we have seen in the ham amps. Remember the full ratings are counting on proper cooling!

This picture was of a tube used in my Henry:

!

In both cases, the painting of the surfaces that are immediately adjacent to the pubes with black non-reflective paint has been reported to help reduce the tube color.  High temperature black paint is readily available and maybe even powder coating might prove to be a great solution. If noting else, it sure looks pretty when you take a picture of the pubes in full color!

--Larry W8ER
Logged
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4410



« Reply #3 on: March 16, 2005, 04:04:26 PM »

T,
I'm sure chimneys would be a definite benefit, max air flow around the tube is a must. You're dealing with heat transfer.  Because heat is energy heat will always move towards the cool zones. For example open the window of your house during this time of year, the cold doesn't come in but the heat goes out. Cold is the lack of energy or heat in this case.  One may not see a perceptible decrease in redness of the plates but I'm sure from the physics stand point the plates are somewhat cooler.  Even though the elements of the tube are in a so called vacuum there's no perfect vacuum better than space and yet heat is transferred from the sun to the earth in the form of visible and invisible light energy.

The glass of the tube and chimney do absorb heat but they also reflect some so removing the heat off the glass is a good thing with air flow much like a car engine. Even though your dealing with a less dense medium like air it still has heat transfer characteristics though they are not as efficient as a liquid.  

Painting any metal in close proximity to the tubes with flat black paint will eliminate or reduce infrared energy reflected back to the tube.  I had a tube audio amp that had a cage around the output tubes. I painted the cage flat black and the cage absorb more heat compared to when it wasn't painted.  So eliminating heat reflection and moving heat from the surface of the glass will help with heat transfer from the tube plate.
Logged

Bob
W1RKW
Home of GORT. A buddy of mine named the 813 rig GORT.
His fear was when I turned it on for the first time life on earth would come to a stand still.
W2VW
Contributing
Member
*
Offline Offline

Posts: 3489


WWW
« Reply #4 on: March 16, 2005, 05:53:34 PM »

Quote from: W8ER

My only comment here is that I have never heard of the solder melting and running out of the filament pins on a Heath SB220/221!


You have now.
Logged
W8ER
Guest
« Reply #5 on: March 16, 2005, 06:44:45 PM »

Quote from: Dave Calhoun W2APE
Quote from: W8ER

My only comment here is that I have never heard of the solder melting and running out of the filament pins on a Heath SB220/221!


You have now.


Interesting! Never have before Dave. Were you running it within spec .. meaning 300 or under and what kind of color were they showing?

--Larry
Logged
W2VW
Contributing
Member
*
Offline Offline

Posts: 3489


WWW
« Reply #6 on: March 16, 2005, 07:06:48 PM »

Not my amp. The guy said it was putting out 1500 Watts of slopbucket. 2500 Watts input verses the 2000 it was designed for. Another fella I know did the same thing. Those amps run a long time at design levels but most people beat the crap out of hammy amps.
Logged
KA8WTK
Contributing
Member
*
Offline Offline

Posts: 874



« Reply #7 on: March 16, 2005, 07:18:49 PM »

Tom,
  I have been thinking about the 6m guy's question and I would think that you are correct in saying "Yes".

Just thinking about the physics involved, because I have no direct experience with tubes like that, leads me to the following thoughts:

1: The plates are going to radiate heat across the vaccum of the tube until it hits the glass.
2: The glass will absorb some of the heat and in turn radiate some.
3: Some of the heat will be reflected back to the plates by the glass.
4: Some of the heat will pass through the glass as raidient heat to be reflected back to the glass and plate by objects around the tube.

So, if you can keep the glass envelope cool, the glass will act like a heat sink for the plate making it run cooler. Painting the surrounding area black will absorb the heat that is radiated by the glass and dissipate it into the chassis. You are dealing with radiated and conducted heat. The radiated heat should act just like visible light and reflections seem to  come into play here.

There was mention made about the different styles of chimneys. Because the air is going to heat up as it passes the tube envelope, it makes sense that the top of the chimney should be "bigger" so the heated/expanded air can go out the top and not restrict flow.

In short, I agree with all the ideas presented in the thread so far and combining them all would lead to longer tube life for a tube that requires that much cooling. They all make good sense and I will keep them all in mind if I ever get or build a rig with a tube that requires a critical amount of cooling.

Good discussion, thanks!
Logged

Bill KA8WTK
W8ER
Guest
« Reply #8 on: March 16, 2005, 08:05:16 PM »

Quote from: Dave Calhoun W2APE
Not my amp. The guy said it was putting out 1500 Watts of slopbucket. 2500 Watts input verses the 2000 it was designed for. Another fella I know did the same thing. Those amps run a long time at design levels but most people beat the crap out of hammy amps.


Oh!

A friend bought an old AL82 from me and swore that he was getting 3400 watts out of it. He did for about 2 weeks on the first set of 3-500Z's. I don't think he got as much out of the second pair!   :lol:

--Larry
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #9 on: March 16, 2005, 08:59:37 PM »

Thanks for the info, guys...

Bill/WTK-

Your info looks good and agrees with a conversation I had tonight on 75M with Stu/AB2EZ. He's a professor of elec engineering and quite sharp.  He agrees with you that the cooler the outside air is, the faster the heat transfer is from the internal plate structure - and the cooler the plate will be. For example, if the air is the same temp as the plate and not moving, then little or no heat transfer takes place. If it is 700 degrees cooler, then a large transfer occurs. There's more, but that's the main idea.


BTW, Larry, I had solder drip out of my 3-500Z's that had just a fan on them.  It was a Sigma XR-3000D.  Dino had the same thing happen.  I cut a 4" hole in the bottom of the cabinet and inserted a blower duct that pressurizes the bottom and the tube pins. This solved the problem.

I was able to repair the pins by QUICKLY heating the pins with a propane torch and soldering them back. Best to put on lots of heat quickly - not low heat for a long time or the glass/seals could blow out.

73,
Tom, K1JJ
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
wa1knx
Contributing
Member
*
Offline Offline

Posts: 451



« Reply #10 on: March 16, 2005, 11:24:21 PM »

hi too voo,
      my sigma, sits on the shelf. i'd sicken if i broke two 3-500z
from solder melting - chicken - i might send the toobs out to you
to fix, and shipem back if you could help - - -
      so I'm stuck using a sb-200 - but working the wurld from
here!! btw, do you use silver solder or regular?

deano!
Logged

am forever!
Steve - WB3HUZ
Guest
« Reply #11 on: March 16, 2005, 11:41:23 PM »

It's no wonder the solder melted out. He was running the amp well in excess of its ratings. I don't think this incident is an indictment of the design used in the SB-220.

I don't think it matter whether a chimney is used or not. As long as enough air is moved across the pins and the envelope, everything will be OK. I don't think there is only one way to do this.

So, if the guy runs the converted SB-220 at the ratings of the original SB-220, I don't see why he would need to do anything further to it.

The 3-500Z is rated for full output to 110 MHz, so getting good numbers on 6 meters shouldn't be surprising.


I agree with Larry on the sockets. The Eimac air sockets are the best if your are going to use a pressurized subchassis and chimneys.

Heat flows (be it via radiation or conduction) from hot to cold. So the colder the air on the outside, the better/faster the heat flow from the tube.


3-500Zs getting a work out at HUZ Radio.




Quote from: Dave Calhoun W2APE
Not my amp. The guy said it was putting out 1500 Watts of slopbucket. 2500 Watts input verses the 2000 it was designed for. Another fella I know did the same thing. Those amps run a long time at design levels but most people beat the crap out of hammy amps.
Logged
W8ER
Guest
« Reply #12 on: March 17, 2005, 12:20:02 AM »

Just in case anybody was wondering about the socket types, I was able to quickly dig up these pictures and the prices of each.

This is the ceramic type that you will see in a lot of the amps like Henry's and other "better" manufacturers amps. Cost is $35 each:



Then there's the cheapie fiberglass socket. Used in the AL80B and the AL82 Ameritron amps at $22 each. Notice how little air each of these allows to flow around the seals of the tubes. Just little holes here and there:



And then there are the Eimac "AIR SOCKETS". These allow air to reach every part of the base of the tube. Notice that it is totally open. These could only be used in a closed chamber type of cooling system. The price of this socket is $79 however or nearly three times the cost of the others:



If you consider how much the more expensive "air socket" are and the price of the tubes, the air sockets may not be such a bad investment after all!

--Larry W8ER
Logged
W3BYM
Guest
« Reply #13 on: March 17, 2005, 07:45:40 AM »

Hi Tom......Might as well put my two cents in the hopper.  I have had my SB-220 for 25 years and still using the same finals.  For the last 8 years it has been used on AM in the hi voltage (2700v SSB ) position. About a 200 watt resting carrier.  Oh the amp will do much more but you start to run out of head room cause of the PS.  At this point Pos Pks above 100% start falling.  Yes the plates glow a nice orange near their centers. The fan does a good job of blowing out all that hot air and keeps both the bottom and top seals cool enough...Obviously I've never had solder melting from the pins like them pricy amps.

73
Logged
W8ER
Guest
« Reply #14 on: March 17, 2005, 09:03:11 AM »

Hey Tom ..

So that's why you're always so piss weak here in Ohio .. trying to keep them toobes cool!

--Larry ER
Logged
W3BYM
Guest
« Reply #15 on: March 17, 2005, 09:18:21 AM »

At least I get out...BTW you better be carefull I know the "agent"

73
Logged
W8ER
Guest
« Reply #16 on: March 17, 2005, 09:26:08 AM »

You best be careful .. I know "the Betty" and she likes me enough to make me cheesecake!

--ER
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #17 on: March 17, 2005, 09:45:13 AM »

The 3-500 does not have the same base as the 4-400. The old johnson sockets were made for the 4-125 through 4-400 base with the holes lined up with the tube base holes. This restricted the air flow requiring higher air pressure under the socket to move enough air through the small holes to cool the jug.
Eimac then built the 3-500 and 3-1000 tubes without this base to restrict
air flow so the tube could be cooled with a lower pressure under the socket because there was less restriction under the tube. The newer socket was open to allow more air flow.
A 3-500 in a 4-400 socket puts you back into needing more air pressure because the socket has the small air passages. One could drill holes around the socket to move more air over the glass.
SSB service may not be a problem but AM linear is making those
plates glow.
The heat on the plate travels up the stem to the plate cap. The higher the temperature of the seal the closer you are to failure. I had a stick on temperature sensor on the plate cap of my old 4-1000A rig that turned color if the radiator ever got too hot .
A red tube needs a lot of air past the radiator cap. Vertical fins would have been better. fc
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #18 on: March 17, 2005, 10:47:50 AM »

Yes, agreed, Frank.  It's standard procedure in all of my amplifiers to drill 1/2" chassis holes in a circle around the tube socket.  The glass chimney JUST covers them at the edge. Maybe 8-12 holes depending upon the size of the tube. This makes a huge difference in air flow through the chimney.
Works outstandng for 4-400's, 3-500Z's and especially well for 4X1's!

I even did this for my 6M external anode, ceramic tube 8877 HB amp. At first I had no holes and I went to a larger blower  - not much help, just raw restriction. Used an "official" 8877 air socket. After drilling the holes I needed a Variac to throttle down the air flow!  It passes so much air now it's like a hurricane and cool. Air still passes thru the tube fins but gets the additional flow of air on the edge. The blower works better cuz of less turbulance.  The best part about this is that the blower room noise of air is quieter because of less turbulance too. It consists of of mostly white noise rather than rumbling and whistling.

The ultimate cooling system is what I described above PLUS mount a BIG blower outside the wall (or outside the house) with a duct connected to the rig in normal fashion. Connect the duct to the blower so it SUCKS air out. The air comes in thru the TOP of the tube and down. VERY quiet with no turbulance or hiss. In addition, the whine of the blower is gone and the room stays cool in the warmer months. Reverse the connection outside and you can push air in and warm the room. But I found moisture from outside a problem to rusting parts, so I just suck air out now.  This is the best cooling system I've ever used for big tubes.  I can run QRO on AM in 90 degree WX and not even know the big rig is on.

Warning:  Do not use plastic dryer duct, etc for sucking air out - it will melt or deform. And do not use a solid material like PVC pipe, as it will act like a trumpet and conduct the blower noise into the room rather loudly. Rather, use a fiberglass reinforced flexible hose (4" diameter) that will withstand 350 degrees F . Available from McMaster on the web by the foot.

73,
Tom, K1JJ
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #19 on: March 17, 2005, 11:38:14 AM »

I just reran the dryer hose because the plastic crap traps lint and clogs the hose. Home depot sells the aluminum 4 inch cheap. I stick a sheet metal screw at each joint. You have to be careful with holes around the tube. You want enough air to flow across the pins. It is best to make a U tube and measure the pressure as you add bypass holes to verify the chassis pressure is still high enough. I did the same thing with the old 4-1000A rig.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #20 on: March 17, 2005, 12:15:34 PM »

Quote from: WA1GFZ
I just reran the dryer hose because the plastic crap traps lint and clogs the hose. Home depot sells the aluminum 4 inch cheap. I stick a sheet metal screw at each joint. You have to be careful with holes around the tube. You want enough air to flow across the pins. It is best to make a U tube and measure the pressure as you add bypass holes to verify the chassis pressure is still high enough. I did the same thing with the old 4-1000A rig.


If you do a remote blower mount, the aluminum 4" will also "trumpet" and amplify the motor sound back to the rig. You need a fabric of some kind like fiberglass, etc. that will withstand some heat and is a poor conductor of sound.

As for the holes, you will find that usually enuff air finds it's way by the pins even with the extra holes. The path is very broad in the journey from the blower to the socket/pins. In fact, I usually use a sealed/pressurized  sub-chassis so that the tube compartment is maybe 6"X6"X6"  and becomes very efficient for air flow. I usually keep the bi-filar choke in there too so that it gets cooled.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #21 on: March 17, 2005, 12:23:52 PM »

Hey Tom,
I see botox didn't work for you.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #22 on: March 17, 2005, 12:48:36 PM »

Quote from: WA1GFZ
Hey Tom,
I see botox didn't work for you.


ih knarF,

thaW si a xotob??

[Bizzaro World message]

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Steve - WB3HUZ
Guest
« Reply #23 on: March 21, 2005, 09:18:01 PM »

Been doing some reading on this. Seems the amount of air flow over and away from (through the air system) the tube the important thing. You can determine the amount of power dissipated from the following.

Pd = 169Qa[T2/T1 -1],

where Pd is power dissipated,
Qa is air flow in CFM
T1 is air system inlet air temperature (K)
T2 is air system exhaust air temperature (K)

As you can see, the more air flow through the air system (Qa) the greater the potential power power dissipation.

You can get an idea of how much power your tube is dissipating by measuring the inlet and exhaust air temperatures. Let's use a 3-500Z as an example. The tube requires 13 CFM for proper cooling. If the inlet air was at room temperature, what would the exhaust air temperature be if the 3-500Z was dissipating 500 watts?

Rearranging the above formula to solve for T2 we get the following.

T2 = T1(Pd/169Qa +1)

Plugging in the numbers yields:

T2 = 294 [(500/(169 * 13)) + 1]        {70 degrees F equals 294 K}

T2 = 294 (500/2197 +1)

T2 = 294(0.228 +1)

T2 = 294(1.228 +)

T2 = 360.9 degrees K = 189.9 degrees F

That's getting pretty warm. Actually the exhaust air would be a little hotter than that because we didn't account for the power dissipated by the filament.

It's interesting to note that if we increase the air flow, the exhaust temperature decreases (e.g. for 20 CFM the temperature is 147.8 degrees F).

Has anyone here done any temperature measurements? If so, do they compare with these calculations? I have a 3-500 amp. I think I'll get a thermometer and see what's happening.

I also thought about the trick of drilling holes in the chassis around the tube to increase air flow. Check my thinking here, but if you are using a blower that can supply the required CFM AND handle the back pressure, I don't see how drilling holes helps. Given the proper blower, ALL the air  must pass through the socket and over the tube. It seems drilling holes just bypasses some of the air that would normally pass through the socket and over the seals on the pins.

Am I smoking dope here or what?
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #24 on: March 22, 2005, 12:06:40 AM »

Good info, OM.  Nice to see it reduced to a formula.

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?  I think so myself...

As for the holes around the socket... After they are drilled I've seen perhaps three times the air flow passing thru the tube's chimney. I think it's either a matter that my cheap  blowers do not work well with back pressure, or perhaps the turbulance generated by the difficult path thru the socket causes flow problems.

I also think that just a small amount of air is needed by the seals at the bottom, and the majority is needed by the glass envelope up on top where the plate structure resides.  Whether or not the holes are there the pins get enuff air. So, I think any way to increase the overall flow thru the chimney past the plate is good.  Well, at least I've found with my latest 8877 6M linear that the air temperature is now luke warm, where before it was hot for the same blower/variac setting - perhaps triple the air flow now coming out of the tube's fins.

I realize that if Eimac wanted it run that way they wud have recommended it, but never mentioned it.  But it has worked for me with 4X1's, 4-400's, 3-500Z's  too. The Tron is who turned me on to it years ago. [The Guru Trogladite can't be wrong!  caw mawn -   :lol: ]

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: [1] 2 3   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.081 seconds with 19 queries.