Should I Leave My Tubes on 24 x 7?

[watchnerd]

I have an a tube phono stage (Fi Yph), one of these:

https://6moons.com/audioreviews/fi3/yph.html

It uses two 6DJ8-type tubes. The default are Electro Harmonics 6922, but I'm currently running a pair of NOS Matsushita 7DJ8 (7 volt version of 6DJ8). I'm assuming it should be a gentle load given the voltage rating difference?

https://www.upscaleaudio.com/products/matsushita-national-pcc88-7dj8

I don't really need any of this, as my Devialet has a good phono stage already, but now that I have a SUT (2, actually), I thought it would be fun to dabble with a tube phono stage for a while.


I've read that preamp tubes can have 5,000 to 10,000 hours life, and at this low price point, I'm not all that worried if I have to replace them every 18 months; it's a pittance compared to cartridge replacement costs.

So....

Should I leave the tubes on 24 x 7? Or only turn them on when needed?

What's the trade off between turn-on wear-and-tear vs time under voltage?

And does sitting there powered on, with no music, wear the tube out the same as when there is a signal?

 

[DonH56]

I'd turn them off but your call. There are 8,760 hours in a year if I did the math right (365 days * 24 hours/day = 8,760 hours). Low-level tubes tend to last longer than power tubes but yes they wear out whenever they are on, signal or no.

I do not know how much 14% under voltage will extend the life (if any). However, since tube stages tend to use low feedback, you may also be changing the gain and other parameters since you're probably reducing the bias current and gain a bit.

 

[watchnerd]

I'd turn them off but your call. There are 8,760 hours in a year if I did the math right (365 days * 24 hours/day = 8,760 hours). Low-level tubes tend to last longer than power tubes but yes they wear out whenever they are on, signal or no.

I do not know how much 14% under voltage will extend the life (if any). However, since tube stages tend to use low feedback, you may also be changing the gain and other parameters since you're probably reducing the bias current and gain a bit.

I've heard some say that the shock of turning the tubes on and off is as bad or worse than leaving them on longer.

Which, I understand the change in temp and voltage, but.....

When I think back to half a century ago, when tubes were the only option, surely people didn't leave electronics on 24 x 7. And it seems like tubes often lasted multiple years, despite that.

Thoughts?

 

[Blumlein 88]

I wouldn't leave them on.

There are delay circuits. These will apply filament voltage, but not plate voltage. After 30 or 60 seconds when the filaments are warm, you turn on the high plate voltage. Not sure this is right as letting voltage go to both the filament would heat slowly getting current flow slowly. The reverse happens if you slap full plate voltage on a warmed up tube.

Soft turn on circuits applied plate voltage slowly over a couple seconds which makes more sense to me.

Various ways you can do all this.

Then I read somewhere that in the days of ENIAC computers (17,000 tubes in these), they tested soft turn on because every time they powered up the ENIAC they lost some tubes. They built one test section with soft turn to test it over months. They didn't find any difference in how many tubes were lost or the lifetime of the tube. They found three things made a big difference in reliability. Firstly, they burned in tubes for 100 hrs before installation to weed out early failures. Secondly they used the tubes well below the rated plate voltages. Thirdly, they found insulation that rodents hated and wouldn't chew on so they used it to insulate the wiring in the latter times of the ENIAC. Rodents damaging wiring was a big problem early on.

So if you are okay replacing them leave them on. I think you would find no degradation after they've been on 30 minutes. Usually it only takes that long as tubes age and get weak in my experience. If you are sure you aren't using them certain times of the day you could get one of those fancy programmable timers and have the stuff turn on and off with a timer. That way you aren't just burning them all the time, yet they can be warmed up when you need them, and you are limiting the number of times they are turned on and off.

I would think never turning them off and on would result in an extended number of hours the tube would last. But just as a for instance suppose this makes them last 15,000 hrs instead of 7500 hrs. You still aren't coming out ahead if in one situation they are being used 90% of the 7500 hrs, but only being used 20% of the 15,000 hrs.

 

[Blumlein 88]

OH and a story about my longest lived tube device. I received a black and white 12 inch portable vacuum tube television in 1966 made by GE. It had a switch that had Off-Standby-On. Standby put about 2/3rds voltage on the filaments at all times. From off it took about 3 minutes for the TV to come on and have its full quality picture. From Standby it took about 30 seconds.

I used this TV pretty extensively taking it to college in the 1970's where it was my only TV. While in college the switch for Off-Standby-On became glitchy. I bypassed it with a simple on-off toggle switch. I used it for a spare bedroom TV for years. Gave it to my mother who used it watching soaps in the laundry room for many years. I ran across a TV repair store shutting down and purchased nearly all the tubes in it in case they ever needed replacing. I did have to replace one tube in the 1990's. The volume control had become scratchy so I replaced it at that time too. It finally was no longer used after 2006 (40 years). Now I don't know how many hours it had, but it was plenty. I don't know how some other components like caps and such didn't go bad. I turned it on and it still worked earlier this year when I sold it to a collector along with the spare tubes.

 

[watchnerd]

There are delay circuits. These will apply filament voltage, but not plate voltage. After 30 or 60 seconds when the filaments are warm, you turn on the high plate voltage. Not sure this is right as letting voltage go to both the filament would heat slowly getting current flow slowly. The reverse happens if you slap full plate voltage on a warmed up tube.

If the Fi Yph has delay circuits, I can't tell.

It seems to just come on, albeit at an output voltage way below line level,

 

[solderdude]

I've heard some say that the shock of turning the tubes on and off is as bad or worse than leaving them on longer.

Filaments operate at much lower temperatures as glow bulbs. For this reason there is no evaporation like there is in halogen and filament light bulbs.
I would switch it off when you are sure you aren't going to switch it on again within 1 or 2 hours or so.

 

[Deleted member 9685]

Condensers have lifespan.

 

[Blumlein 88]

Condensers have lifespan.

Yes indeed they do. Some aren't much more than low voltage vacuum tubes.

 

[Julf]

The Colossus (early computer) rebuild at Bletchley Park still has some of the original WWII-vintage tubes in it. They do turn it on and off (the electricity and cooling bill would be scary if they kept it on 24/7), but ramp the voltage up/down in slow steps, so turning on or off takes half an hour.

 

[watchnerd]

Condensers have lifespan.

How long?

 

[DonH56]

How long?

Varies widely with type and application. The rule of thumb is ~10,000 hours these days with different types and materials ranging from maybe 2500 to >20,000 hours. Performance may deteriorate much sooner, especially for electrolytic capacitors, and applied voltage, ripple current, and temperature are major factors in how long a capacitor will do its job. And some ceramic and film caps last much longer. Life testing is usually at 1000 hours under various stress conditions (typically higher voltage and temperature) and lifetime extrapolated from that.

 

[Tom C]

I’ve been through this with my McIntosh C2500 and Mc275. What I’ve settled on is that I turn the system on when I’m ready to use it, then don’t turn it off until I turn in for the night. I don’t use it every day.
The Western Electric website has schematics and technical papers from back in the day available for download. In my opinion, there is no more expert or authoritative reference in existence. I slogged through the technical manual, where the author indicated that an idle tube, turned on but without a signal present, will accumulate charged particles on the cathode. Once a signal is applied, the tube output will be degraded until that accumulation is dissipated. So, leaving the tubes on when not being used decreases performance, but temporarily.

 

[watchnerd]

I’ve been through this with my McIntosh C2500 and Mc275. What I’ve settled on is that I turn the system on when I’m ready to use it, then don’t turn it off until I turn in for the night. I don’t use it every day.
The Western Electric website has schematics and technical papers from back in the day available for download. In my opinion, there is no more expert or authoritative reference in existence. I slogged through the technical manual, where the author indicated that an idle tube, turned on but without a signal present, will accumulate charged particles on the cathode. Once a signal is applied, the tube output will be degraded until that accumulation is dissipated. So, leaving the tubes on when not being used decreases performance, but temporarily.

Huh, very interesting.

Did Western Electric have anything to say about tube break-in?

There is a lot of audiophile lore about how many hours tubes take to break-in.

 

[solderdude]

I’ve been through this with my McIntosh C2500 and Mc275. What I’ve settled on is that I turn the system on when I’m ready to use it, then don’t turn it off until I turn in for the night. I don’t use it every day.
The Western Electric website has schematics and technical papers from back in the day available for download. In my opinion, there is no more expert or authoritative reference in existence. I slogged through the technical manual, where the author indicated that an idle tube, turned on but without a signal present, will accumulate charged particles on the cathode. Once a signal is applied, the tube output will be degraded until that accumulation is dissipated. So, leaving the tubes on when not being used decreases performance, but temporarily.

That would only be the case for class-B output stages. Not in this particular case.

 

[watchnerd]

That would only be the case for class-B output stages. Not in this particular case.

Does this mean Class A circuits have enough heat and current, even with no signal, to not accumulate particles?

 

[Tom C]

Huh, very interesting.

Did Western Electric have anything to say about tube break-in?

There is a lot of audiophile lore about how many hours tubes take to break-in.

No, it wasn’t mentioned specifically. Probably not a thing then.

 

[Tom C]

That would only be the case for class-B output stages. Not in this particular case.

Not sure what you mean. In the book, the point was made during a discussion of tube characteristics in general, not in a specific application. I don’t have time at the moment, but later I’ll dig out the reference. The text is available for free download on the WE site, so anyone who cares to refer to the original will be able to.

 

[Tom C]

I have an a tube phono stage (Fi Yph), one of these:

https://6moons.com/audioreviews/fi3/yph.html

It uses two 6DJ8-type tubes. The default are Electro Harmonics 6922, but I'm currently running a pair of NOS Matsushita 7DJ8 (7 volt version of 6DJ8). I'm assuming it should be a gentle load given the voltage rating difference?

https://www.upscaleaudio.com/products/matsushita-national-pcc88-7dj8

I don't really need any of this, as my Devialet has a good phono stage already, but now that I have a SUT (2, actually), I thought it would be fun to dabble with a tube phono stage for a while.


I've read that preamp tubes can have 5,000 to 10,000 hours life, and at this low price point, I'm not all that worried if I have to replace them every 18 months; it's a pittance compared to cartridge replacement costs.

So....

Should I leave the tubes on 24 x 7? Or only turn them on when needed?

What's the trade off between turn-on wear-and-tear vs time under voltage?

And does sitting there powered on, with no music, wear the tube out the same as when there is a signal?

BTW, I really like that chassis design. Reminds me of the old Cray supercomputers.

 

[watchnerd]

BTW, I really like that chassis design. Reminds me of the old Cray supercomputers.

I can see the resemblance to the old Crays.

Don Garber, the maker (now passed away), was very artisanal in his amp designs. I'm not sure how good his circuits are (the Yph has pretty pathetic gain), but the chassis always looked cool.