I don't think that is the case, I have seen used ones, and I am very careful to avoid ones that are sold as used. New ones have quite a lot of gettering, as they run they loose a large percentage of the getter and the appearance of the glass changes as well from evaporating tungsten coating the glass.
I've found the Eastern European sellers of electronic parts on eBay so far to be extremely ethical and trustworthy. I've never had a problem with any transaction even when the goods were destroyed in shipping. They've always either been what they were represented to be, or worked hard to make me happy if they weren't. I have bought more than 20 pairs of the GM70 and a few hundred of various other types generally in bulk from these sellers. I haven't had any problems with purchased tubes, I did have a small number that got destroyed in shipping, the sellers always made good and quickly.
You should preheat the tubes. It will extend their life significantly. It doesn't have to be 15 minutes, I do it because I have time, but 2 - 3 minutes is much better then nothing.
Also, when turning them off, you should turn off B+ first, wait until the caps are discharged and then turn off the amp.
Another thing, your anode disipation is too high. Graphite version can withstand 110W, no problem, but it shortens their life. 100W is max for them if you want them to last. I know it's a little bit of a problem when you use fixed bias, but it can be done. My amps use fixed bias to, but it's fixed but variable
It slides with B+, so I have constant anode disipation and current.My amps were measured with R&S Analyzator, Pout is 32W.
Operating point 1050V - 1100V 95 - 90mA! I'm going even higher this year, 1200V 80mA. Regards
Thanks for sharing! Is your GK-71 triode strapped? If yes, does it withstand 800V easily on G2?
You should preheat the tubes. It will extend their life significantly. It doesn't have to be 15 minutes, I do it because I have time, but 2 - 3 minutes is much better then nothing.
Also, when turning them off, you should turn off B+ first, wait until the caps are discharged and then turn off the amp.
Another thing, your anode disipation is too high. Graphite version can withstand 110W, no problem, but it shortens their life. 100W is max for them if you want them to last. I know it's a little bit of a problem when you use fixed bias, but it can be done. My amps use fixed bias to, but it's fixed but variable
My amps were measured with R&S Analyzator, Pout is 32W.
Regards
Unfortunately preheating the tubes is not an option because of a long ago design flaw, bias and B+ come up together. It's not likely going to get fixed as that would require an additional power transformer, at $50 every 12 - 14 months or so for graphite it is not too big a big deal. (Coppers are more of course)
The operating bias is easily adjusted (there is a bias pot that allows adjustment over an 80mA - 140mA range with a typical tube), but I've not found a significant correlation between life and operating point. I'm currently running 90mA with copper plates at 1kV.
How many hours would you estimate you have on the amps over the past 4 years - are you approaching 1000 hours or are you well past it? A rough estimate is good. Is it possible that you are nowhere near 1500 hours which is the best I've ever done? If you have achieved 3000 - 4000 hours then I will have to consider making some changes, otherwise there is no evidence to justify additional investment in time and effort in these particular amps. (But noted for the next one)
I have meet several other people running GM70 in some cases much more conservatively than I am, and with lower voltage and cathode rather than the fixed bias I am using and they report very similar life times.
I think a very important consideration in the GM 70 is filament voltage.
http://www.tubebooks.org/tubedata/HB-3/General/Types_of_Cathodes.pdf
If you read second paragraph second page: "a reduction of 5 percent in the filament voltage of tubes with pure tungsten filaments will double their life"
Looking at the GM 70 tube spec sheets I get that they should be run between 19 and 20 volts, which is a 5 percent difference. They should last twice as long at 19 volts than they do at 20 volts. I am running mine at 19.5 volts and would guess I always get more than 1500 hours from them.
http://www.tubebooks.org/tubedata/HB-3/General/Types_of_Cathodes.pdf
If you read second paragraph second page: "a reduction of 5 percent in the filament voltage of tubes with pure tungsten filaments will double their life"
Looking at the GM 70 tube spec sheets I get that they should be run between 19 and 20 volts, which is a 5 percent difference. They should last twice as long at 19 volts than they do at 20 volts. I am running mine at 19.5 volts and would guess I always get more than 1500 hours from them.
I almost made the same mistake and bought additional power transformer for bias. Heating is with modified SMPS and it is 20V. I think that also reduces life of a tube that big. If you don't have constant/regulated voltage on heaters it could shorten the tubes life.
I can't tell for sure how many hours. I think between 1000 - 2000 hours. Another thing I've done is that I've switched 'polarity' of SMPS for heaters and every month or so, I change tubes between channel's. GM-70 doesn't have center tape in heater so one side is degerating faster than the other, if you know what I mean.
You probably have more knowledge than me, but I've studied GM-70 pretty much and I still have much help from colleagues that are great with tubes.
If I can help here a little bit, it would be my pleasure.
Regards
The comment on the 19V vs 20V filament supply is interesting wrt life.
I had originally set the CCS for 20V operation after some warm up period, but have noted they drift downwards a bit during the course of the first hour of operation. I am actually not sure where they end up, I suppose I should check..
I had given some thought to uneven wear in the filaments over the life time of the tube, and the possibility of reversing the connections periodically, my concern initially being about filament breakage which was also one of the motivations for CCS based heating. This was actually the first design where I used Rod Coleman's filament CCS, they have worked reliably over all these years, but are well heat sinked.
I had originally set the CCS for 20V operation after some warm up period, but have noted they drift downwards a bit during the course of the first hour of operation. I am actually not sure where they end up, I suppose I should check..
I had given some thought to uneven wear in the filaments over the life time of the tube, and the possibility of reversing the connections periodically, my concern initially being about filament breakage which was also one of the motivations for CCS based heating. This was actually the first design where I used Rod Coleman's filament CCS, they have worked reliably over all these years, but are well heat sinked.
I have measured my gm70 with 20V and 17V Heater voltage. The difference in values is very small.So, under heating is not changing the tube characteristics a lot. Typically smps can be regulated, but of that does not work, why not a dropper resistor ? 1V less is not such a big thing, I does this with the 4V rectifier all the time.
The comment on the 19V vs 20V filament supply is interesting wrt life.
I had originally set the CCS for 20V operation after some warm up period, but have noted they drift downwards a bit during the course of the first hour of operation. I am actually not sure where they end up, I suppose I should check..
I had given some thought to uneven wear in the filaments over the life time of the tube, and the possibility of reversing the connections periodically, my concern initially being about filament breakage which was also one of the motivations for CCS based heating. This was actually the first design where I used Rod Coleman's filament CCS, they have worked reliably over all these years, but are well heat sinked.
Haven't tried with CCS or Rod Coleman's filament regulator.
I have very good experience with SMPS.
I'm not talking about tube characterictics, I'm talking about tube life. Take in consideration that isn't small driver tube, it's big output tube and under heating shortens it's life.
Understood.
Do you know this article ?
https://p02.de/wp-content/uploads/2013/02/Regeneration_T510-1.pdf
Even though in German, the graphic on p. 4 is interesting: If you go to 90% filament power, you can get much more than 200% life time out of a thoriated tungsten filament while if you go down further to 85% it is dead after 20% of its lifetime...
Do you know this article ?
https://p02.de/wp-content/uploads/2013/02/Regeneration_T510-1.pdf
Even though in German, the graphic on p. 4 is interesting: If you go to 90% filament power, you can get much more than 200% life time out of a thoriated tungsten filament while if you go down further to 85% it is dead after 20% of its lifetime...
Perhaps of interest, CPI Microwave Power Products Eimac Operation APPLICATION BULLETIN no. 18 document about Extending Transmitter Tube Life for TT filaments.
Please note previous AB18 document statement : "When a new power tube is first installed in a transmitter, it must be operated at rated nominal filament voltage for the first 200 hours. This procedure is very important for two reasons. First, operation at normal temperature allows the getter (a device that chemically binds tube contaminates) to be more effective during the early period of a tube’s life, when contaminants are more prevalent. This break-in period conditions the tube for future operation at lower filament voltages. Secondly, during the first 200 hours of operation, filament emission increases. It is necessary for the life extension program to start at the peak emission point.", is not stated explicitly in attached revision.
Please note previous AB18 document statement : "When a new power tube is first installed in a transmitter, it must be operated at rated nominal filament voltage for the first 200 hours. This procedure is very important for two reasons. First, operation at normal temperature allows the getter (a device that chemically binds tube contaminates) to be more effective during the early period of a tube’s life, when contaminants are more prevalent. This break-in period conditions the tube for future operation at lower filament voltages. Secondly, during the first 200 hours of operation, filament emission increases. It is necessary for the life extension program to start at the peak emission point.", is not stated explicitly in attached revision.
Just thoughts passing through my mind.
I know that service life of regular tungsten incandescent bulb is extended way more than 400% when operated in series through a single diode. The brightness, however, is greatly reduced.
Sign of imperfect vacuum? Regettering using induction heater when tube is intentionally operated hot to induce outgassing after some use (say 200 hours) could help?
Sign of filament temperature being pushed too high?
I know that service life of regular tungsten incandescent bulb is extended way more than 400% when operated in series through a single diode. The brightness, however, is greatly reduced.
No, it is in pentode mode. I don't think there would be any problems with G2 when triode strapped, look at graphs.
http://tec.org.ru/_bd/21/2115_71-.pdf
I believe that is what the document suggest, but at 95% rated voltage until emission is found to be unsatisfactory. Then increase filament supply to rated voltage. Lastly increase to 105% rated voltage when needed.
But I have also seen another document (currently escaped my memory) stating 90% as a good target to extend service life of TT filaments.
Have you heard about Miller-Larson effect?
Thoriated Tungsten Filament power cycling
Every turn on/turn off shorten TT filament life.
It's independent of the filament voltage or current.
www.r-type.org/pdfs/thoria05.pdf
Thoriated Tungsten Filament power cycling
Every turn on/turn off shorten TT filament life.
It's independent of the filament voltage or current.
www.r-type.org/pdfs/thoria05.pdf
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.