My TV-3 tube tester shows a short on two of the three 42 tubes I just tested. Haven't seen this before, is it real?

Thank you,

It was on the pair of 6F6G(same tube, octal base) in my Silvertone 4586. Had a bad hum so I recapped it, still hummed, both outputs were shorted.

If it's a solid short, you can confirm it with an ohmmeter. If it's just bad (or not so bad) leakage, maybe not. Leakage can also only appear when the cathode is hot. The most common shorts I have seen in output tubes is from the screen grid to the control grid. These are frequently intermittent but can be disastrous when they happen. Fortunately, that isn't too often.

Sounds like I should throw them away.

Thank you both,

High resistance heater-cathode leakage won't cause a problem in an output stage (though a REAL short is a problem). Most tube testers will show "short" with 100K resistance - which is a problem on a grid, not so much for heater-cathode. I see heater leakage sometimes with 6F6 and 42 tubes.

I assume the high resistance short can further degrade over time?

All tubes degrade with use, and one of the ways this happens is by developing high resistance leakage paths between the elements and through the heater-cathode insulation. Most of the time, tubes get replaced for other reasons before these leakage paths get bad enough to become troublesome. On the other hand, if there are hard-shorts in a tube due to loose or deformed elements, the problem can't get any worse!

If you turn to the chart on page 4-4 of the TV-3 manual http://www.mcmlv.org/Archive/TestEquipment/NAVSHIPS91254-TV-3U.pdf (Courtesy MCMLV.org), you'll see a chart that can be used to determine which elements are causing the shorts light to come on. Heater to cathode leakage may be tolerable in output stages, but I would stay away from any tube that shows leakage from any element to the first grid.