HI ALL:
If a tube (25L6GT) tests bad for leakage, but strong for emission (on a Sencore TC109), what effect/problem will it manifest when installed in a set (Westinghouse H-130)? TIA.
BOB

Depends on what elements are leaking. Look at this thread.
viewtopic.php?f=8&t=81699

Depends which elements exhibit the leakage. Heater to cathode, maybe no problem, in a given circuit.

Large-cathode tubes like the 25L6 are prone to develop grid emission, and that often shows up on a leakage test, if the test is good enough. If that's the problem with the tube, you may get a pretty good case of distortion out of it, and it will likely cause other problems too.

With output tubes in general, leakage often involves vaporized material (mostly from the cathode) settling on the micas and establishing conduction paths between the cathode, grids, and plate. There'll be a short, sooner or later.

I wouldn't advise running any tube that can't pass muster on a leakage test, especially an output tube.

Larry

I won't use a tube that shows grid leakage as I don't want the tube to cause a problem with other parts in the radio. 25L6's aren't expensive, I recommend replacing it. The tube may work, but it could cause distortion or hum. Harry

That leakage also contributes to the "strong" tester result.

Carl

Cheap enough, toss it. If I had one I'd give it to you.

Thanks to all for the information/advice/opinions. Interestingly enough, the set works the same with the "leakage" tested bad tubes installed as it does with two tested good emission/leakage tubes. I have plenty of 25L6GT tubes that test good for leakage, so replacement is not a big deal. Just wondering what the ramifications of leaving the leakage tubes installed might be.
BOB

Bob

High dissipation tubes often have leakage over time. Notice carbon buildup in used tubes. The grid resistor is usually specified at not more than 500K for an output tube. This resistor can overcome some grid leakage. Filament to cathode leakage usually doesn't matter unless it gets real high.

This string begs the question, just what leakage does a tube tester report? I have a Hickock Cardmatic that has a leakage scale with an acceptable and unacceptable range (nominally scaled by the punch card installed) and it will find leakage in tubes that other tube testers will not find. Is the tube tester identifying cathode leakage, grid leakage, or what?

Norman

On some tube testers such as Eico 666 you can identify the pin with leakage and amount. The meter reads out leakage in ohms rather than a neon lamp.

Others are so sensitive many tubes fail. Sencore doesn't have much of an emission test but has an amplified leakage. Manufactures complained about this tester years ago as so many tubes were rejected.

25L6, 35L6, 50L6 all have the same leakage issues. They may operate fine at first but after they run for 5-10 minutes the control grid voltage starts to rise much like a leaky coupling capacitor. The tube conducts harder, runs hotter, and runs away into self destruct. The sound gets distorted and the high current taxes the rectifier.

I have had output tubes check leaky on a Sencor tester but when actually running in the set, the control grid voltage remains at zero and they sound just fine. I always monitor the control grid voltage for a while after replacing one of these just to make sure it's not creeping up.

It's called grid emission. All large-cathode output tubes are susceptible to it. 7591 and 7868 tubes are among the worst offenders, but there are many others.

A big cathode means lots of coating to vaporize, especially if the heater is overvolted and run too hot (as usual, with elevated line voltages now prevailing in many places). The vaporized material must settle somewhere. Some of it settles on the micas, creating leakage paths between elements. Most of it, however, settles on the control grid, and when it is heated up by that big cathode next door, it starts to emit like a second cathode, and things spiral out of control from that point.

Larry

If I understand Larry's comments correctly, excessive cathode voltage will contribute to "grid leakage" (which my Sencore TC109 measures). This particular set ( Westinghouse model H-130) has a line dropping "ballast" resistor listed as 72 ohm/10w (with a 12 ohm lamp shunt). The original one in the set measures a total of 52.1 ohms. Given my steady line voltage of 120VAC, seems like a cement 90 ohm/10-15W substitute would be appropriate, plus a 15 ohm for the lamp shunt. Would this type of replacement be appropriate, and will it reduce the possibility of future geid leakage problems? TIA for your thoughts/comments/opinions. I also hope that this discussion assists others.
BOB

Not excessive cathode voltage (unless you're working with filament-type tubes, in which heater and cathode are one), but excessive heater voltage, which boils off the cathode coating. Let's say the heater's meant for 6.3v, and due to excessive line voltage, you're actually getting 7 volts on the heater, or maybe more. A certain amount of cathode coating vaporizes at normal heater voltage; that's just the nature of the beast. But overvolting the heater, overheating the cathode, accelerates the process exponentially.

If you have excessive line voltage, you can add a bucking transformer or an outboard voltage regulator like a Variac for your sensitive gear. If your set has a ballast-and-lamp shunt that's working right, I wouldn't tamper with that; I'd try to control things externally.

Larry

A Sencor Mighty Mite 5 tester also refers to it as "Grid Leakage" but as Larry pointed out it is also more accurately called secondary emission and grid emission so I can understand some of the confusion.

Ive always found that the output tubes used in auto radios to have much higher leakage problems than the same type in home use. Its no wonder considering the typical 7.2 or 14.4V charging voltage from a generator at cruising speed.

The 6F6 is another Ive found to be especially susceptible to leakage and it may be due to the higher temperature of metal tubes.

Other tubes such as converters, RF and IF amps, also develop leakage that kills AVC action and results in distortion but Ive no answer to the cause as Ive found no pattern.

Carl

It's vaporization of the cathode coating, deposited on the micas and making leakage paths between elements. It's worst with output tubes, for they run the hottest, but small-signal tubes are subject to it also. It's worse with all tubes when their heaters are overvolted.



The tube on the right (a 6SU7GTY) fouled out with leakage that injected a loud "sputtery-spit" noise into the signal path. Compare its top mica with its twin on the left (NOS).

Tomer explains the entire matter in Getting the Most out of Vacuum Tubes (Gernsback, available online).

Larry