I routinely rip apart military surplus for the tubes and other useful components, so I looked in some of my goodie boxes, and found these. It looks like there are about 6 black strips inside that contact the glass and there are little chimneys between those strips and the outside wall. If air was forced up through the chassis it would flow through these little chimneys removing heat.
It is interesting that all of these have the same numbers on them, but no two are the same color. I don't know if all of these came from the same piece of equipment, or each came from a different boat anchor. Whatever it came from was likely stored outside in the Florida rain.
These fit an EL84. I think I have seen smaller ones in my warehouse. Anyone willing and capable of valid temp measurements?
Attachments
I think there are some special pencils or "crayons" that are made for checking temperature of surfaces. They either melt or change color at some specified temp. I can't recall at the moment what they were used for or what temp ranges they cover, but it was for HOT stuff. Might be a problem to get them to "draw" on glass though.
Steam Punk Coolers!!
How about some "steam punk" coolers. Use several of those flexible metal lamp necks (like say 3/8 inch diameter), mounted to the chassis around each tube and bend them to point at the tube(s). Pressurized chassis air flows up thru them and gets directed at the tube from several angles. Maybe could put an artificial smoke generator below deck to simulate steam, or just drop some dry ice down a chute into the chassis once in a while. A new meaning for "cryo cooled tubes". Maybe put brass mini fire engine hose nozzles on the ends too. Paint the flex red. Heh..., I just realized, you can buy these things off the shelf already, oil/water mist-er flex fixtures for a lathe or milling machine or diamond saw.
How about some "steam punk" coolers. Use several of those flexible metal lamp necks (like say 3/8 inch diameter), mounted to the chassis around each tube and bend them to point at the tube(s). Pressurized chassis air flows up thru them and gets directed at the tube from several angles. Maybe could put an artificial smoke generator below deck to simulate steam, or just drop some dry ice down a chute into the chassis once in a while. A new meaning for "cryo cooled tubes". Maybe put brass mini fire engine hose nozzles on the ends too. Paint the flex red. Heh..., I just realized, you can buy these things off the shelf already, oil/water mist-er flex fixtures for a lathe or milling machine or diamond saw.
Last edited:
"How you gonna keep that pressurized nozzle quiet?"
Could make it whistle at ultrasonic frequencies. No, just kidding. I would just use the low pressure from a sub-chassis fan to push the air thru. Could still recess the socket some to get additional cooling air if necessary. (and a big hole in the end of the nozzles, not the tiny hole that mist-ers use)
How 'bout some LEDs in the nozzles to color the smoke too. Now if we really wanted to get fancy, some nylon strings inside the flexes could be pulled by some servos to make the hoses move around, maybe time share them around on the different tubes. Or a bent rod inserted into the flex tube could be simply rotated around from below by a clock motor to make the hose rotate around between tubes.
Or how about a little wood maniken man with a chopping axe (controlled by strings like in a cuckoo clock) banging away on a tube? OK...., I knew that wouldn't go over. How about the maniken operates an old fashioned bellows that blows air on the tube.
All right..., I've really got it down now, I promise... no more after this one:
The tubes get mounted on swiveling sockets that rock back and forth to the music.
Could make it whistle at ultrasonic frequencies. No, just kidding. I would just use the low pressure from a sub-chassis fan to push the air thru. Could still recess the socket some to get additional cooling air if necessary. (and a big hole in the end of the nozzles, not the tiny hole that mist-ers use)
How 'bout some LEDs in the nozzles to color the smoke too. Now if we really wanted to get fancy, some nylon strings inside the flexes could be pulled by some servos to make the hoses move around, maybe time share them around on the different tubes. Or a bent rod inserted into the flex tube could be simply rotated around from below by a clock motor to make the hose rotate around between tubes.
Or how about a little wood maniken man with a chopping axe (controlled by strings like in a cuckoo clock) banging away on a tube? OK...., I knew that wouldn't go over. How about the maniken operates an old fashioned bellows that blows air on the tube.
All right..., I've really got it down now, I promise... no more after this one:
The tubes get mounted on swiveling sockets that rock back and forth to the music.
Last edited:
There are sticks of this but liquid Tempilaq works on glass
Temperature indicating liquids. Quick application.
It's used to indicate temperatures for heat treating, as well as over-temperature indication. Used in R&D labs, on race car engine parts like turbochargers etc. You can find out peak temperatures by painting on a series of strips or dots of successively higher temps
I have some Penta transmitting tubes that have yellow painted logos on the glass that turn brown at operating temperatures, then back yellow as the glass cools. I wonder if it is permanently affected by >220 degrees C or so...
Last edited:
There is always an absorbed and reflected part - of course, it always emits on it's own since it's temperature is over absolute zero
With respect to the tube, it's the absorbed vs reflected part that's relevant, what the cooler really does is attempts to increase the radiating surface at the 'expense' of lower temperature. A cooler has to have a lower temperature by definition, if it is to cool. Assuming that there is adequate airflow, transfer of heat from the tube to the cooler is by convection and a relatively large portion is by radiation. The convection part is there because air flowing over the cooler flows also over the bulb surface, and since flow is not ideally laminar, it will transfer some heat from the higher temperaure tube to the lower temperature cooler. Also, at the same time, this same airflow cools both the tube and the cooler. The tube innards are also quite a bit hotter than the bulb glass, but since there is vacuum inbetween, the transfer of heat is almost exclusively by radiation (with exception of parts where inner workings of the tube touch the glass, such as supports, pins, etc, but this is a small portion of the total for regular glass tubes). The radiation is partly absorbed by the glass, and also partly reflected back into the tube, which lowers the cooling effect. However, unlike the glass, the cooler is made specifically to absorb this radiated part, by virtue of choice of material, color/coating, and geometry. In the case of the 'harmonica' fin coolers, the angle of the fin is an optical trap of sorts. The cooler is made much better at absorbing IR than the glass but still not perfect so part of it is reflected back, and also radiated by the very fact the cooler is hot. We want to prevent this radiation going back onto the tube innards, so the angles of the fins are such that the reflected or radiated part mostly ends up on the neighboring fin, where it is again largely absorbed but partially reflected - many times, this rises the absorption level to almost ideal, and thus more heat is transferred onto the cooler, to be taken away by convection.
Story is plausible, backed by some documentation, and I havn't a whole
lot of reason to doubt. "Seller?" have you seen the prices, not like he's
attempting to make a mint on those coolers. Why so angry about it?
Why a getter expert shouldn't know something useful about processes
inside an overly hot vacuum tube? The connections do seem obvious.
Also, this "seller" didn't barge in to sell something. From post #3 onward,
we ouselves have kept up referencing, seconding and thirding the Pearls
and asking specific questions about them. He answered, so big deal...
I'm a cryo skeptic, but not above having it proved to me. Getters and
heatsinks, I don't need further proof. Those things at least are real.
I'm only worried he might sell me a handful of getters, what would I do?
Last edited:
Even if the contact coolers work, there is still an issue with the uniformity of the cooling. If part of the glass bulb is cooled down and another part is not, there is going to be significant thermal expansion stress in the glass:
"I tried this on an EL34 and after a hardish session the tube packed up with cracks in the glass where the sink was fixed. Slight expansion/ contraction in the zone where the sink doesn't quite make contact all the way round is asking for trouble. "
Cool the sides and the top cracks off. I would at least supplement such a contact cooler with fan air to cool the other parts. The RCA 813 data sheet says "nothing touches the glass!" And for tubes that don't run red plates, this may all be purely academic, even the soft glass will handle 600+ C... Clearly the tube has already failed if it gets that hot.
"I tried this on an EL34 and after a hardish session the tube packed up with cracks in the glass where the sink was fixed. Slight expansion/ contraction in the zone where the sink doesn't quite make contact all the way round is asking for trouble. "
Cool the sides and the top cracks off. I would at least supplement such a contact cooler with fan air to cool the other parts. The RCA 813 data sheet says "nothing touches the glass!" And for tubes that don't run red plates, this may all be purely academic, even the soft glass will handle 600+ C... Clearly the tube has already failed if it gets that hot.
Last edited:
Smokin's comment above is interesting and germane. Hot spots are a big deal and one of the 'deals' with coolers is that they equalize bulb temp to a great degree.
And absolutely NO, soft glass will NOT 'handle' 600ºC and parts getting that hot expire forthwith. Did you ever see a power tube whose envelope had caved in due to extreme heat. I have a PYREX™ 6550C around here that suffered such a fate, and the stupid thing still works after a fashion. It's not a happy camper but it makes gm none the less.
And just BTW, I am NOT a gettering 'expert,' a ridiculous comment, read the post. What I said was that I have worked at the implementation of continuous gettering in tubes for audio use so as to ensure a long, long working life. To ensure, in other words, that my customers get -value- when they buy parts in whose design I've had a hand. Capiche ? ?
And absolutely NO, soft glass will NOT 'handle' 600ºC and parts getting that hot expire forthwith. Did you ever see a power tube whose envelope had caved in due to extreme heat. I have a PYREX™ 6550C around here that suffered such a fate, and the stupid thing still works after a fashion. It's not a happy camper but it makes gm none the less.
And just BTW, I am NOT a gettering 'expert,' a ridiculous comment, read the post. What I said was that I have worked at the implementation of continuous gettering in tubes for audio use so as to ensure a long, long working life. To ensure, in other words, that my customers get -value- when they buy parts in whose design I've had a hand. Capiche ? ?
IR Thermography & Tube Cooling
Something I forgot to mention above is that I've lately been talking with FLIR, makers of industry-standard IR thermographic equipment, IR cameras in other words.
Doing bulb temp measurements with thermocouples is difficult and the proverbial picture being worth . . . it seems to me a good idea to take some snaps.
The matter though is that it's not a 'point and shoot' gig, there are matters of emmissivity and glass transmission loss to learn about and factor in. Also, there's a small matter of cost in that the camera needed runs some $US17K and rents at about $300/day. So one's ducks need to be all lined up beforehand, and I'm working on that. I should have some very interesting pix on my site in a couple of months.
Something I forgot to mention above is that I've lately been talking with FLIR, makers of industry-standard IR thermographic equipment, IR cameras in other words.
Doing bulb temp measurements with thermocouples is difficult and the proverbial picture being worth . . . it seems to me a good idea to take some snaps.
The matter though is that it's not a 'point and shoot' gig, there are matters of emmissivity and glass transmission loss to learn about and factor in. Also, there's a small matter of cost in that the camera needed runs some $US17K and rents at about $300/day. So one's ducks need to be all lined up beforehand, and I'm working on that. I should have some very interesting pix on my site in a couple of months.
Well, I wouldn't argue with keeping soft glass tubes well below 600 deg C. Especially as the typical Sweep tubes spec out a 200 to 260 deg. C. max operating temp. which would presumeably be soft glass.
I mentioned the 600+ figure since the RCA TT3 Transmitting Tube data handbook mentioned 625 deg. C as the softening temp. for soft glass. And 750 deg. C for hard glass.
Out of curiousity, I've pulled out "Building Scientific Apparatus" 4th ed. to see what it has on glass properties. It lists Soda-lime (presumeably the "soft" glass) as softening at 700 deg C and Pyrex 7740 as softening at 820 deg. C. I assume that is what is being referred to as "hard" glass as there are some even higher temp glasses listed. Vycor 7900 lists as softening at 1500 deg. C and Fused Silica (Quartz) at 1600 deg. C.
Softening temps:
Soda-lime: 700 C
Pyrex 7740: 820 C
Vycor 7900: 1500 C
Fused Silica: 1600 C
The annealing temps may be more relevant data since these would be the lowest temp that the glass can relieve stress by "cold" flow.
Soda lime: 550 C
Pyrex 7740: 555 C
Vycor 7900: 1020 C
Fused Silica: 1100 C
And strain point temps, which are the highest temp the glasses can be rapidly cooled from without cracking:
Soda-lime: 500 C
Pyrex 7740: 510 C
Vycor 7900: 890 C
Fused Silica: 950 C
Huh, I would have thought Pyrex would be a lot higher than soda-lime on that one. I'm a little disbelieving on the soda-lime at 500 C.
edit: I just looked it up on a specialty glass mfg website and they do say 514 C. I don't think any of my drinking glasses would handle that. (WalMart, made in China)
Then there is linear expansion coefficient (cm^-1 K^-1 10^-6), which would give an indication of resistance to temperature gradients:
Soda lime: 8 to 10
Pyrex 7740: 3.3
Vycor 7900: 0.75
Fused silica: 0.55
Then some interesting info is given on thermal conductivity of glasses. Apparently they are all similar at 0.008 J cm sec^-1 cm^-2 K^-1
For reference, this thermal conductivity is stated as an order of magnitude less than that of graphite, two orders of magnitude less than metals, and about an order of magnitude greater than wood.
Optical transmission for soda lime is stated as variable depending on composition. (but generally not so hot for IR) Pyrex is clear down to about 2200 nm wavelength (IR) and fused silica down to about 3500 nm. IR range starts at about 670 nm and longer.
I mentioned the 600+ figure since the RCA TT3 Transmitting Tube data handbook mentioned 625 deg. C as the softening temp. for soft glass. And 750 deg. C for hard glass.
Out of curiousity, I've pulled out "Building Scientific Apparatus" 4th ed. to see what it has on glass properties. It lists Soda-lime (presumeably the "soft" glass) as softening at 700 deg C and Pyrex 7740 as softening at 820 deg. C. I assume that is what is being referred to as "hard" glass as there are some even higher temp glasses listed. Vycor 7900 lists as softening at 1500 deg. C and Fused Silica (Quartz) at 1600 deg. C.
Softening temps:
Soda-lime: 700 C
Pyrex 7740: 820 C
Vycor 7900: 1500 C
Fused Silica: 1600 C
The annealing temps may be more relevant data since these would be the lowest temp that the glass can relieve stress by "cold" flow.
Soda lime: 550 C
Pyrex 7740: 555 C
Vycor 7900: 1020 C
Fused Silica: 1100 C
And strain point temps, which are the highest temp the glasses can be rapidly cooled from without cracking:
Soda-lime: 500 C
Pyrex 7740: 510 C
Vycor 7900: 890 C
Fused Silica: 950 C
Huh, I would have thought Pyrex would be a lot higher than soda-lime on that one. I'm a little disbelieving on the soda-lime at 500 C.
edit: I just looked it up on a specialty glass mfg website and they do say 514 C. I don't think any of my drinking glasses would handle that. (WalMart, made in China)
Then there is linear expansion coefficient (cm^-1 K^-1 10^-6), which would give an indication of resistance to temperature gradients:
Soda lime: 8 to 10
Pyrex 7740: 3.3
Vycor 7900: 0.75
Fused silica: 0.55
Then some interesting info is given on thermal conductivity of glasses. Apparently they are all similar at 0.008 J cm sec^-1 cm^-2 K^-1
For reference, this thermal conductivity is stated as an order of magnitude less than that of graphite, two orders of magnitude less than metals, and about an order of magnitude greater than wood.
Optical transmission for soda lime is stated as variable depending on composition. (but generally not so hot for IR) Pyrex is clear down to about 2200 nm wavelength (IR) and fused silica down to about 3500 nm. IR range starts at about 670 nm and longer.
Last edited:
We have a Flir camera here at work. The lens we have is unfortunately suited for taking pictures of IC die in operation to see where the losses are.
It would seem that merely pointing the camera at a tube would result in an image not unlike those used in some forum members avatars. You would get an image of the metal elements inside the tube with the colors representing various temperatures. There would be an offset due to the transmission loss through the glass. Getting a reading on the glass itself would require at least a small patch of thermally emissive paint. Getting pictures of the benefits of fitting a tube cooler would require at least a small hole in the cooler to allow for photographing the plate structure.