Cooking old tubes

[KB5MD]

I know this has been covered before on this board in one way or another, but I have a couple of 810 tubes that have been on the shelf for 40 something years.  Now that I have a use for them, what is the best method to bring them back to usefulness?  The filaments are good.
One person told me to bake them in the kitchen oven for awhile.  Has anyone else used this method or have a better method?

[k4kyv]

I think by "baking" they meant running the filament for a couple of hours, or maybe overnight, before applying plate voltage. Then, if you have any way to reduce plate voltage below normal, run them at reduced plate voltage, at near full dissipation for between 15 minutes to a couple of hours.  Then run them at normal plate voltage, and if possible close to full plate dissipation for another 15 minutes or so, before running them at  normal service.

In the final, this can be accomplished while feeding the xmtr into a dummy load. In this case add the intermediate step of running with grid drive but no plate voltage for a half-hour or so.  In the modulator, if you have a variable bias supply, you could run them with no audio applied, and adjust the bias at each plate voltage to just below the rated plate dissipation.

This is based on the recommendations in a transmitting tube booklet I have that was published by RCA, and which is quoted verbatim in a previous thread. I tend to prefer a little overkill with the time periods. RCA recommends only about 15 minutes for each step, but they were talking about spare tubes normally in storage at a commercial station, and I'm not sure it would be enough to clean up a tube that has been lying unused on the self for decades.

Baking tubes in an oven sounds more like audiophoolery or "CB expertise" to me.

[W4AMV]

Don, can you sucessfully "bake" a tube like an 811A, 814 and so on by applying the filament voltage only? Say at 1/2 value, current limited, and then slowly ramp up over time. Again, no plate V applied. 

[WA3VJB]

On the 300-G you can use the low power control to cut the plate voltage. Just make sure the relay actually works !

Also, I recommend burnishing and perhaps even using some Brasso to polish up the the grid and plate caps on the 810s, AND inside the push-on connectors that are standard on the 300-G.  This will minimize friction and mechanical resistance, and help address a common failure on the 810s between the cement that holds the porcelain or bakelite to the glass.

The transmitter's connectors here, while pretty, are physically hard on the tubes because they require significant torque to remove, and significant pressure to install.

[k4kyv]

Don, can you sucessfully "bake" a tube like an 811A, 814 and so on by applying the filament voltage only? Say at 1/2 value, current limited, and then slowly ramp up over time. Again, no plate V applied. 

I would slowly ramp it up, but then let it  run at full voltage for a few hours.

I tried to pull up the previous thread, but as usual, the search function yielded "no result".

[KM1H]

Running just the filaments only works with oxide cathodes and not tungsten filaments.

If a tungsten filament has been poisoned by running under voltage or in a bit of a gassy (not AIR) enviroment then running about 25% higher for 10-15 seconds may restore emission.

If the anode in a filament tube has the getter deposited on it then running at color often works but if the tube seals have let in AIR then the tube is junk. If the getter is not on the anode the only thing that may work is to RF heat the envelope where a getter film is deposited or heat the getter itself as done in production.

I had no difficulty regettering several 50-60's NOS 304TL's a year ago since the anode is the getter and its real easy to get that tube to glow a bright red at 900V and biased for 200W of dissipation.

Ive used that method since the 60's and it was old hat then and discussed on the air often, mostly about the big Eimac glass tetrodes that showed up as BC pulls.

Carl


 

[KB5MD]

Ran the tubes with only filament power for 6 hours.  That was sufficient, transmitter is fully functional.  Thanks for the tips

[KM1H]

Which proves nothing unless you ran them full bore first to set a baseline reference.

[k4kyv]

Running just the filaments only works with oxide cathodes and not tungsten filaments.

Why do you suppose RCA recommends running the thoriated tungsten filament-only, for a period of time as the first step of the break-in period following extended shelf storage?

If a tungsten filament has been poisoned by running under voltage or in a bit of a gassy (not AIR) enviroment then running about 25% higher for 10-15 seconds may restore emission.

I have always been careful not to run the filament at under-voltage, but I have tried that procedure a few times after accidentally severely overloading a tube, for example having a sustained arc occur in the tank circuit while I was outside checking something about the antenna, or inadvertently not being aware I was running full plate voltage with the tank circuit out of resonance   It never worked for me, even though RCA has recommended it in their literature.  Once the tube was severely overloaded enough to cause a noticeable drop-off in performance, mine were always shot for good.

I had no difficulty regettering several 50-60's NOS 304TL's a year ago since the anode is the getter and its real easy to get that tube to glow a bright red at 900V and biased for 200W of dissipation.

That apparently works. I once came across a pair of HK-354s (I think that was the number, the larger H-K triodes with the metallic (tantalum?) plates like the older 304-TLs) rated at 150w plate dissipation. I ran them filament-only for about 30 minutes, and then applied reduced plate voltage.  One of them turned blue inside almost like an 866A.  I went  through the complete RCA-recommended procedure: filament-only, reduced grid drive, then full grid drive, then full grid drive at reduced plate voltage, finally full grid drive at full plate voltage, into a resistive load. Following the RCA procedure, I left both tubes running with full filament voltage, about 1200 volts on the plate, with the grids biased so that the tubes were idling at just below the full rated plate dissipation. I let them glow away for about 4 hours. After that break-in, I could apply as much plate voltage as I wanted and the blue glow was gone.  The tubes would easily pull all the current the power supply would put out if I momentarily reduced the bias.  I never actually tried them out in a transmitter to see if they were fully functional; I think I traded them off for some different tubes.

I have found that graphite-anode tubes like the 211 and 810 are far less susceptible to gassing up on the shelf than metal-anode tubes like the 304-TL.  My experience with those metal anodes has been mostly with the older style tubes with the silvery plates that have heat radiating fins. This includes both Eimac and Heinz-Kaufmann tubes.  I'm  not sure what, if any, difference there is between the older style plates and the newer ones that are a dull grey without radiating fins.

My experience with metal-plate RCA tubes, namely 833-As and 8005s, has been that they are more like graphite anode tubes, very rugged.  I know that originally, 813s had graphite plates and were practically indestructible.  Then, RCA changed over to some kind of dark blue metal plate that appears flimsy by comparison. I have never used metal anode 813s, so I can't say how well they compare to the older ones.  Given a choice of good 813s of both types, I would opt for the graphite anode.  I suspect the metal anodes were a cost-cutting measure, much like discontinuing the little ceramic supports and the internal shield at the bottom of the plate structure, the latter-day 807s.

One of my very best 833As was overloaded by a previous owner to the point that there is a bright spot the size of a half-dollar on the plate, silvery deposit all over the interior of the glass envelope, and the glass envelope is sunk in with a sizeable dimple right where the bright spot is.  Somebody cut the plate voltage just in time, and apparently assumed the tube was shot and replaced it.  I got it in some junk and literally kicked the thing around the shack for several years before testing it in the rig.  Wonder it didn't get broken, because I assumed it was NG and made no effort to take care of it.  When I tried it out one day just for the hell of it, I was astounded to see that it functioned  like a new tube.  Whether installed in the modulator or the final, with that tube, I can get positive peaks as good as with any tube in my collection.  With several other pristine-looking tubes, the positive peaks are down.


 

[WA1GFZ]

Years ago a friend worked on a high voltage lasar system that used 3-400. EIMAC  burned them in at some crazy HV. NIB tubes came with holes in the plates as a result and worked fine.

[The Slab Bacon]

If a tungsten filament has been poisoned by running under voltage or in a bit of a gassy (not AIR) enviroment then running about 25% higher for 10-15 seconds may restore emission.

that almost sounds like the "old school" proceedure for "rejuvenating" '01As

Run them at around 7.5-8v for a few minutes, then up the voltage to 10 or 12v for a few minutes, then occasionally "flash" them up to 15 or so volts for a few seconds a few times, then leave them at 10v for a few minutes, than back them down to 7 or 8 for about 15 minutes and try them. I have actually done it a few times and brought them back to life, sometimes you just pop the filament. It is supposed to "boil" new thorium up to the surface of the filament. If they have low / no emission wadda ya got to lose.....................

YMMV   

[John K5PRO]

In 1977 I saw one that looked like this at a station, a suck in the glass in front of the plate. They said it was running very bright just before it quit. It happened in a Gates 1 kW, one which used 807 cathode followers direct coupled to the modulator grids, either a 1T or 1G most likely. I suppose one potential failure mode of that design is when the 807 fails, shifts bias to the modulator tube on that side and can push it to destruction.

BTW, some of the bigger Eimac, Thales and RCA tubes had getters attached somewhere near lead-in. It could even be applied to the conducting surfaces there. It is activated from adjacent thoriated tungsten filament. We're instructed to always allow those tubes a 15-30 minute warm up before applying HV, if they have been in storage for over a certain time.


k4kyv said:
"One of my very best 833As was overloaded by a previous owner to the point that there is a bright spot the size of a half-dollar on the plate, silvery deposit all over the interior of the glass envelope, and the glass envelope is sunk in with a sizeable dimple right where the bright spot is.  Somebody cut the plate voltage just in time, and apparently assumed the tube was shot and replaced it. "

[Ed/KB1HYS]

I had a 6146 once that had a blackened dimple on one side.  still ran and tested OK, so I left it in.  Considering that glass softens between 1200-1700 degrees it's Amazing how much abuse some tubes can take and still function.

[KM1H]

Why do you suppose RCA recommends running the thoriated tungsten filament-only, for a period of time as the first step of the break-in period following extended shelf storage?

Dunno, what was the date of that and what size tubes? Eimac and others say it does no good at least at up to the 4-1000A level.

My experience with metal-plate RCA tubes, namely 833-As and 8005s, has been that they are more like graphite anode tubes, very rugged.  I know that originally, 813s had graphite plates and were practically indestructible.  Then, RCA changed over to some kind of dark blue metal plate that appears flimsy by comparison. I have never used metal anode 813s, so I can't say how well they compare to the older ones.  Given a choice of good 813s of both types, I would opt for the graphite anode.  I suspect the metal anodes were a cost-cutting measure, much like discontinuing the little ceramic supports and the internal shield at the bottom of the plate structure, the latter-day 807s.

Metal plate 813's can burn thru at loads that wouldnt bother the graphite versions...Ive done it before realizing they were really different.

that almost sounds like the "old school" proceedure for "rejuvenating" '01As

Yup, Ive done it to many of them also as well as 45, 50, 210's, etc. Seems to work with most and likely the others had already had it done or were just run into the ground.

I had a customers SB-1000 where the 3-500Z had holes burnt in the anode in a dozen or more places where the fins contact the anode cylinder. It still works fine at the 500W output level (3100V) but at 600W the grid current drops on key down. Its real weird to see the filament glow thru all the holes.