• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Why does red plate spell early death?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Biasing tubes too hot typically results in red spots glowing on the plate. It is common advice to run tubes below maximum dissipation, or their life will be shortened.

Exactly what physical or chemical phenomenon causes the accelerated failure under high dissipation conditions? Does the cathode just burn off its emissive coating that much faster? Does the heat warp the plates? Or do the plates offgas at high temperature, ruining the vacuum?
 
It can ruin the vacuum in tubes by releasing trapped gases in the metal that weren't baked out.

Alot of infra red radiation being pushed inward the the grid and cathode and if the grid cant take it then it could very easily melt.

You can also start boiling off metal causing it to sublimate and condense on everything else in the tube.




Nick
 
Red plate -> failure

All metals have gas "dissolved" in them when they are made. Part of the processing of a tube is to heat-up all the internal elements as hot as possible (without damaging them) to drive out the gas while the tube is still connected to the vacuum pump. However, this process of diffusion takes time, and for mass-production can only be done for so long. At the end of this out-gassing phase, the tube is sealed and then the getter is "flashed" which captures nearly all the remaining gas as well as leaving a coating on the bulb that will absorb some gas in the future.

When a tube is over-heated, usually evidenced by red or orange plates, more gas is liberated from the hot metal. Reactive gasses, such as oxygen and water vapor, combine with the barium oxide on the cathode and "poisons' it, thus reducing its electron emission capability. At some point, the emission isn't enough to supply what the tube needs, and it goes "flat".

The presence of gas in a power tube can also cause positive grid current to flow. This happens because at high voltages, gas molecules ionize (an electron separates, leaving a positively-charge atom), and these positive ions get attracted to the negative grid. As they accumulate, a grid current starts flowing. If there is a high resistance in the grid circuit (i.e. the grid resistor), the grid voltage starts to go more positive, causing the tube to draw more current, causing the plate to get hotter, thus releasing more gas. The end result can be a spiral to violent death. This is why it is important not to use too high a grid resistor value in power amp circuits.

The positively-ionized gas molecules are also attracted to the cathode, and because of their high mass, can actually physically damage the barium oxide layer. This is called "cathode stripping".

As mentioned before, the getter flash can absorb some gas molecules, but as it only has a certain capacity to do so. That is why power tubes that are driven hard or have been used a lot have an eroded flash.

The getter flash can cause problems in really high-voltage tubes, such as transmitting tubes, so many of them, such as the Eimac glass tubes, have no getter. This requires that they be brought to incandescence while on the vacuum pump, and be pumped for a very long time (sometimes 24 hours or more). This costs money, so isn't considered feasible for consumer tubes.

I hope this helps,

- John Atwood
 
Heat => tube death

no problem Nick. I started writing my post before I saw yours posted.

You bring up a good point - the sublimation of metals. From what I have read, this usually happens first from the getter flash. The effect of this is to liberate more gas and to cause leakage paths where the metal gets deposited. Another reason why flash getters aren't used in high-power transmitting tubes. Many high-power tubes use materials such as zirconium on the plate, which absorbs gas when hot.

- John
 
John's description is indeed correct. I will add that many new production tubes, and some old ones were not built with the best attention to component purity and vacuum quality. It takes less "glow" to contaminate the vacuum.

The grid current issue is very real and will lead to runaway, dead tubes, and blown parts. Pay attention to the "maximum grid circuit resistance" spec in the tube manuals. Reduce the resistance far below the maximum if you intend to push a power tube near or above the limit.

Just how much contamination can be generated by a glowing plate? Several years ago I proved that you could "outgass" enough ions from a glowing plate such that an ionized plasma could be generated inside the tube. By careful manipulation of the current limiting I kept the plasma going until the glass began to soften. When the glass began to soften it bulged OUT!

I have since lost this tube, but I plan to repeat the experiment again some day.
 
Ty_Bower said:
Biasing tubes too hot typically results in red spots glowing on the plate. It is common advice to run tubes below maximum dissipation, or their life will be shortened.

Exactly what physical or chemical phenomenon causes the accelerated failure under high dissipation conditions? Does the cathode just burn off its emissive coating that much faster? Does the heat warp the plates? Or do the plates offgas at high temperature, ruining the vacuum?

Unlike transistors, VTs like to run hot. However too much of a good thing is a bad thing. In addition to what's already been mentioned, excessive heating can also stress the glass-to-metal seals, leading to the vacuum's escaping. Not a good thing.

Unless the spec sheet says it's OK, red plates are best avoided.
 
Ex-Moderator
Joined 2003
Another source of water vapour (for poisoning the cathode) is the mica wafers that support the electrodes. All mica necessarily contains water, so if you heat it excessively (with a red-hot anode) it will outgas - that's why high reliability valves use ceramic electrode supports.

As Wavebourn points out, a very few anodes are designed to operate dull cherry red because they are coated with getters such as zirconium. Usually, they're graphite anodes, and they always have ceramic electrode supports.
 
Administrator
Joined 2004
Paid Member
Take a look at the 8005 sometime - this glows cherry red in operation and the first time I saw this in a MAC MI-250A (IIRC) I was more than a little disturbed even though the service manual stated quite clearly that this was normal. In this tube running with some plate glow actually allows the plate chemistry to capture and recombine gas floating around in the envelope. I think there are also certain Eimac triodes that operate this way - a bit counter-intuitive if you ask me..

Makes me wonder if whether with careful operation long out of service power triodes like these might slightly improve their vacuum once restored to operation? I have heard this alleged, but have not experimented to find out if true or not. Baking at high temperatures "allegedly" helps too, but I am wondering whether this is all urban legend as I can't imagine you could get things hot enough for this to work without compromising the glass and seals?

Comments?
 
Re: Red plate -> failure

JohnAtwood said:
grid current starts flowing. If there is a high resistance in the grid circuit (i.e. the grid resistor), the grid voltage starts to go more positive, causing the tube to draw more current, causing the plate to get hotter, thus releasing more gas. The end result can be a spiral to violent death. This is why it is important not to use too high a grid resistor value in power amp circuits.

- John Atwood


Well written- expanding- this is one of the reasons why not to exchange a KT88 designed o/p stage with 6550's unless well below the power/B+ ratings. On paper the tube power ratings seem identical but the 6550 has a softer vacuum than the 88. So that's why in fixed bias high power stages, make sure the 50K grid leak resistance isn't increased. Tough work for driver stages.

If you force the 6550, the symptoms of anode current runaway become noticeable regardless of the amount of neggy volts applied and the grid (g1) is no longer in control. Most tubes posess this when pushed beyond ratings, some more than others.

richj
 
kevinkr said:

Makes me wonder if whether with careful operation long out of service power triodes like these might slightly improve their vacuum once restored to operation? I have heard this alleged, but have not experimented to find out if true or not. Baking at high temperatures "allegedly" helps too, but I am wondering whether this is all urban legend as I can't imagine you could get things hot enough for this to work without compromising the glass and seals? Comments?

From personal experience I can tell you that to some extent this is true. Small transmitting type tubes can sometimes be degassed by simply running them. It all depends on how gassy they are. Some, like the one Tubelab George put up pictures of, that are deep blue/purple inside are way too far gone. Others less gassy can often be brought down by running them for a time. Sometimes it will take hours to recombine the gas, and other times it will take only a few minuets. I've actually watched the grid current fall (using my big rack tester) as they ran. Blowing forced air across them helps in the process. And sometimes cycling them through hot runs and then cooling them down to be heated again will be needed to get the gas down.

Certain Eimac tubes like the 4-250A, 4-65A, 304TL, 4E27 and many others, big and small, were built with a tantalum plate that was designed to run cherry red during normal operation. The red hot plate is the getter in these tubes. Eimac even coined the term "Pyrovac Plate" for this action.

Some tubes like the 813 and European versions of the above mentioned types (and others) employed a carbon or graphite plate that when hot provided gettering action. I have even been able to cook out gas in some external anode tubes like the 4CX250B and 8122 for example. While these tubes have only a copper anode, baking them under power can sometimes help. It really all comes down to just how much gas you are dealing with and how much time and effort you're willing to spend. Several times I wasted hours trying to degas a tube only to find that it was weak or the cathode had become poisoned.

Victor
 
Kevin, Morgan Jones wrote an article some years back (Electronics World, Nov 2000) where he did some experiments on baking tubes that have been long in storage. IIRC, he also mentions this in VA3.

FWIW, I used some D3a that he had baked and they seemed a bit quieter than "fresh from the old box" ones, but they suffered a loss in cathode emission at a rather young age. It may not have been the baking, I have no controlled tests to indicate that and the baked tubes were from a different batch than my unbaked ones, so I'd want to try it on something cheaper and easier than more of my precious D3a.
 
hey-Hey!!!,
The Tantalum anode tubes designed to run red/orange hot can indeed be de-gassed. This process needs some careful attention, as the cathode is also hot and it seems to me that that element can also get wrecked by gas( Thoriated Tungsten ). I've fired the cathode, and rapidly applied plate voltage....lots of blue glow and kicks in the current limits of my supply but quickly the anode glowed, and the gas stuck to it.

For the grid circuit issue, seems that the lower the value the better, ending at a grid choke or an IT secondary. The 813 limit is 30k, and for my apps, a grid choke will do much better. Same with 50's, Eimac transmitting tubes, etc.
cheers,
Douglas
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.