So, in a recent debate with some friends, I am really wondering how much grid current is acceptable in a new and unused power triode like 300B. My very old and used specimens that I tried, their grid currents seem to be from below 1uA to 20uA at most, while on new eastern production tubes their grid current seems to be in the region of 1 mA! What value should be considered as ”acceptable” ?
The Western Electric 300B does not state a limit for grid current. But a maximum value is easy to work out.
The maximum allowable grid-leak resistor for the 300B (fixed bias) is 50kΩ.
Page 3 of the WE data sheet gives a plot of anode current against grid voltage, with anode voltage as a parameter.
If we locate the standard operating point for the 300B (Ua=350V, Ug = -74V; Ia=60mA) on the graph, we can see what happens if the grid-to-cathode voltage decreases. The horizontal axis is 5V/division, so we can quickly see what effect one division has: just move right to the next division from the operating point on the 350V curve: -74V to -70V. The anode current increases to 82mA! This is much too high, and would cause overheating and other problems.
Actually, even 1V would be unsatisfactory, as it would raise the anode current by ca. 10%.
How does this affect the grid current? We know that the grid leak resistor is 50k, so the leakage current that puts an extra 1V on 50k is 20µA.
Keeping the grid current well below 10% of drift would mean limiting the grid current to 2µA or even less; and anyway I would expect a good quality 300B to leak <1µA in any situation. More than 2µA of leakage, IMHO is too much for any sample to honestly call itself a 300B.
Excessive leakage current might indicate gas contamination in the vacuum; sometimes this gas can be chased into the getter with a little heating (to ca.100 °C).
The maximum allowable grid-leak resistor for the 300B (fixed bias) is 50kΩ.
Page 3 of the WE data sheet gives a plot of anode current against grid voltage, with anode voltage as a parameter.
If we locate the standard operating point for the 300B (Ua=350V, Ug = -74V; Ia=60mA) on the graph, we can see what happens if the grid-to-cathode voltage decreases. The horizontal axis is 5V/division, so we can quickly see what effect one division has: just move right to the next division from the operating point on the 350V curve: -74V to -70V. The anode current increases to 82mA! This is much too high, and would cause overheating and other problems.
Actually, even 1V would be unsatisfactory, as it would raise the anode current by ca. 10%.
How does this affect the grid current? We know that the grid leak resistor is 50k, so the leakage current that puts an extra 1V on 50k is 20µA.
Keeping the grid current well below 10% of drift would mean limiting the grid current to 2µA or even less; and anyway I would expect a good quality 300B to leak <1µA in any situation. More than 2µA of leakage, IMHO is too much for any sample to honestly call itself a 300B.
Excessive leakage current might indicate gas contamination in the vacuum; sometimes this gas can be chased into the getter with a little heating (to ca.100 °C).
This is exactly my point! Thank you very much my dear Rod! I also red an article in old sound practices magazine from J.C Verdier where he measured 7 specimens of old 300B tubes with grid currents ranging from 7 to 135nA ! But in todays troubled world of globalized economies there are “professionals“ who are trying to sell “wonderful blue glowing tubes” with 1mA of grid current as “normal” and “designed that way” !
Many drivers that RC couple to the 300B grid . . .
At power-up, if the driver tube is cold, and the 300B is cold, the unloaded B+ voltage can be applied through the driver RL to the coupling cap, to the cold 300B grid resistor, Rg to ground.
(With cold driver and cold output tubes, B+ is larger than normal).
Suppose the unloaded B+ goes to 475VDC, and suppose you use a 400V coupling cap.
475V, RL, 400V coupling cap, Rg to ground. The capacitor says OUCH!
That is why I always use 600V or 630V coupling caps.
At power-up, if the driver tube is cold, and the 300B is cold, the unloaded B+ voltage can be applied through the driver RL to the coupling cap, to the cold 300B grid resistor, Rg to ground.
(With cold driver and cold output tubes, B+ is larger than normal).
Suppose the unloaded B+ goes to 475VDC, and suppose you use a 400V coupling cap.
475V, RL, 400V coupling cap, Rg to ground. The capacitor says OUCH!
That is why I always use 600V or 630V coupling caps.