All 4 tubes tested were unused New Old Stock (NOS), prepared for service by:

1. Running the filaments at 3.5 volts for 2 hours, then at 6 volts for another 22 hours.
2. Placed in the amplifier, and anode voltage was set to 900v for the initial tune-up.
3. Idle current set to 25 ma.
4. Driven at 2w, tuned for best input match and output power.
5. Anode voltage increased to 1200v, idle current reset to 25 ma.
6. Driven at 5w and tuned for max output.
7. Driven at 10w and tuned for max output.
8. Anode voltage and idle current reset to 1350v and 25 ma. 
9. Tuned for max output.

At this point, with the amplifier on standby, my no-load anode voltage was 1650, dropping to 1350 under full load.

The first 7289 was manufactured by G.E. in 1978, and turned out to be the best performer with the least thermal drift. The grid current could be run at 70 ma before it began to overheat and drift due to mechanical deformation. Reducing the drive restored normal operation. The foil packing envelope had been opened in 1986, and it had been stored that way until now (2004).

The second 7289 was also manufactured by G.E., but in 1974. It had a weaker grid, and overheated beginning right at 60 ma. At 75 ma., the tube would collapse completely in just 2 seconds. The packing envelope had been opened, and it had been stored that way for an unknown period of time. I bought this one on EBAY, so it's storage history was unknown to me.

The first 2C39WA was an HP part of unknown manufacture, date of manufacture unknown. It was ceramic like the 7289's, with a threaded anode post and screw-on anode heat sink, so it was of the more modern era. It was not in a sealed envelope, and had been stored that way for 18 years.

The second 2C39WA was the same origin and storage condition as the first.