History:
Early 1968 at GEC Chelmsford UK as a passing technician, I distinctly recall the longevity spec for the KT88 inhouse manufactured; soaktesting racks filled with valves til the eye can´t see anymore with timers set for 2000 hrs all valves with cherry red anodes, and after time-up, batch samples shock/parameter tested., glass cleaned, fresh logo/date and after QC passing put into boxes. The electric required and heat produced was quite a show. Now today, I wouldn´t daren´t believe on economic grounds any such hefty test would be done, (maybe still in Russia/China where there is ample electric) but those days DID produce tubes/valve with excellent consistent reliability. I still use 807´s in an old guitar amp dated 1954.
"Use´em but don´t abuse´em"
Bench Baron.
Early 1968 at GEC Chelmsford UK as a passing technician, I distinctly recall the longevity spec for the KT88 inhouse manufactured; soaktesting racks filled with valves til the eye can´t see anymore with timers set for 2000 hrs all valves with cherry red anodes, and after time-up, batch samples shock/parameter tested., glass cleaned, fresh logo/date and after QC passing put into boxes. The electric required and heat produced was quite a show. Now today, I wouldn´t daren´t believe on economic grounds any such hefty test would be done, (maybe still in Russia/China where there is ample electric) but those days DID produce tubes/valve with excellent consistent reliability. I still use 807´s in an old guitar amp dated 1954.
"Use´em but don´t abuse´em"
Bench Baron.
@benchbaron Hello, that's interesting, but perhaps I have misunderstood. Are you saying they ran the anode red hot for 2000 hours and then sold them if not destroyed? Or perhaps you are saying only a small sample size from each batch would be tested that way and then discarded, if they survived the test then the untested valves from that same batch would get sold, and if the samples failed then the whole batch got destroyed. Interested to learn more about how they were made back then; a lot of knowledge never gets recorded and is lost with time.
2000h is 83 days or nearly 3 months, and that's a looong time.
Maybe that explains why my Gold Lyon/M-O/GEC/KT88 all have a (small) brown border around the getter flash. I got them brand new in the 70's, and they stayed unused in a cool dry closet since then.
Generally speaking, the parameters drift a bit during the first few hours and stabilise, but the sound continues to evolve over a few days at least. But then I can only measure the anode current and the transconductance. As usual in audio, there are other factors in play that we don't measure for lack of knowledge.
Maybe that explains why my Gold Lyon/M-O/GEC/KT88 all have a (small) brown border around the getter flash. I got them brand new in the 70's, and they stayed unused in a cool dry closet since then.
Generally speaking, the parameters drift a bit during the first few hours and stabilise, but the sound continues to evolve over a few days at least. But then I can only measure the anode current and the transconductance. As usual in audio, there are other factors in play that we don't measure for lack of knowledge.