Technical Bulletin TB-11 Part 1
Cathode Issues and sparking tubes
by Jac van de Walle
For amplifier manufacturers: Please click here for basic design rules. The text below here is for the interested end user.
The white Spark Problem.
Sometimes, an amplifier can cause a tube to spark. These are little white sparks during start up of the amplifier, sometimes causing speaker noise too. The spark is useually small. For the tube itself, this spark is not dangerous, but it is not the best for maximum life. If a spark occurs, it gets visible with mesh tubes, while solid plate tubes spark invisible at the inside. The effect for the tube cathode is the same.
The first thing to say, this is always an amplifier problem. Even when a tube sparks only in particular amplifiers, or situations, it stays an amplifier issue. So yes you can sometimes solve it by exchanging the tube. Or in reverse, a tube exchange caused the sparking. But when the design issues are solved, any tube will work in any amplifier, and not spark.
What causes the spark physically?
If the local current density of the cathode is exceeded, it this generates also heat locally. The cathode has a powdered surface, and if we would look at that under a microscope it looks like a moon landscape, with craters and pieces of rubble. Because of this, some microscopic small parts and places emit locally more electrons that others. Specially when the heater is in the warming up phase, the most active spots begin to work first. This is the moment, any tube may possible spark. If at this particular moment, the amplifier already tries to pull maximum current from the tube, or exceeds it even, this full current will flow through this small spots, which initially began to work. It should be clear, this concentrates a lot of current on a recuded number of small places. This causes local over heating of those spots. To make things worse, these spots were the better ones, even capable of this high current. It is why they began working at first. Only, these spots will now locally over heat, which cause tiny fractions of cathode powder to become loose. These very fine dust particles, blasted out into the vacuum, get charged negative by the free electrons. Once charged, they moved away from the (negative) cathode, and want to fly to the anode. Before they get there, they get massively bombarded by all of those electrons, which equal the electric current of the whole tube. The dust particle gets white hot, and evaporates. This gives the white spark you can see. This evaporation for a very short moment causes the audible effect. In very rare, severe cases, it causes an avalanche effect, and then the whole tube lights up. This is called a plasma effect, and it causes a short, and the fuse may blow. Some 100 years ago, OSRAM called it the Trigger Effect, which is a good name for it.
First the diagnosis, then the treatment.
Fully understanding this issue requires a high level of technical understanding. I will try to avoid referring to this, when possible. It is a common, some people just swap the "bad" tube with a "good" one, and from that they can tell for SURE, if the cause is within the tube or within the amplifier. Myself I do not have this magic ability.
There is a common misunderstanding, tube manufacturers are supposed to build rectifiers which can withstand a certain l level of abuse. Amplifier manufacturers load rectifiers with the maximum specifications. They do not do so with output tubes, because this not good for the lifetime. But with rectifier tubes, that doesn't seem to matter.
Jean Claude Verdier (Yes from the French VERDIER company) could not have explained it better here: Please read his words.
Luckily, tubes and amplifiers will survive the attacks, but we must say, good product quality is not defined like that. Some products do not accept any abuse at all, while they can last for ever if you use them properly.
The nicest example I once had, was by a tube dealer in Rome, Italy, who mistakenly tested a 2A3 tube on 300B settings, with his tube tester. Note, 300B has 5Volt heater voltage, and 2A3 has 2.5Volt. Apart from that, 2A3 will test well on 300B settings. He had been doing so ever since, and did not notice. Until, one day it went wrong. The first tube of an very expensive AVVT 2A3-Mesh pair survived, and test results were fine. The second 2A3 he tested, blew up the heater wire. We had a lovely discussion about tube quality afterwards. His reasoning was, if the "good" tube can survive it, the "bad" tube has a quality issue. Those two tubes he said, were not the same. For him, one was good and the other was bad. He never changed his opinion, and quit the relationship for ever.
What does it mean when we have a white spark issue?
The good news is: The tube is still good. The flashing may seem terrible. but in vacuum a short is not possible. The tube takes no damage, but rather the amplifier could, or the SE transformer, or the tweeters. Though this risk is small, and most of the time it will only give a small crack noise in the speaker.
The bad news is: The amplifier has a design problem. To say it clearly up front, a white spark is always an amplifier problem by definition.
What causes the white spark?
This is explained at some other places in this text as well, but quick and simple: It is caused by too much current, with a not fully warm tube.
The most seen cause is excessive in rush current with tube rectifiers, far above the maximum peak value. There are a few ways to do this wrong.
The second cause is DC coupled circuits, which work good by coincidence, and sometime by some other coincidence, they fail with only one tube of a pair..
The effect gets larger when the tubes have better emission, so it can happen after a tube exchange.
To prevent any damage do not use the amplifier any more, and contact your amplifier manufacturer to have it checked up, or better say repaired.
For good amplifier designers, the below links are not really new, but don't feel ashamed to read it anyway.
1) Warning by OSRAM, about design issues,
which causes sparking tubes
2) Spark Warning by RCA, about design issues,
which causes sparking tubes
3) Another Spark Warning by RCA.
4) Table of lowest allowed transformer
windings resistance
5) This Application Note, Part 2
Important Notes for amplifier manufacturers
6) Rectifier tube killer schematic
7) Circuit to calculate minimum series resistance of heater winding.
