How long do tubes really last?

[Steve Conner]

I've always thought of power tubes as having a lifespan of a few thousand hours, similar to lightbulbs. Then I saw this:

http://lists.contesting.com/_amps/2003-06/msg00151.html

This radio ham claims to have over 40,000 hours on a 4CX250B tube, and it's still good!

This raises a lot of questions, such as:


Do you really need to retube your amp every couple of years?

Do modern tubes have a pathetic lifespan compared to NOS?

Do glass power tubes have a short life compared to ceramic RF tubes like the 4CX250B?

Do musical instrument amps abuse the tubes so badly that they go to an early grave?

Are there easy design changes we could make to said amps that would make the tubes last 5-10 times longer without compromising tone?


I've got a big collection of used tubes that I've pulled from scrap equipment, and I rarely see a bad one. I think I saw one GEC KT66 with a damaged grid that caused a red spot on the plate, an EL37 that had thermal runaway, and a pair of metal US Army 6L6s that were gassy. Every other tube I've ever pulled has been full of life and sounded good. I have some Sovtek tubes too, but I've not put enough hours on them to draw any conclusions.

[Old Tele man]

...check the MIL-STD RELIABILITY Handbook for vacuum tubes, there're equations and tube-life estimation data...yes, the MIL/NOS tubes lasted quite awhile!

...RDH states: with power tubes, "end-of-life" when gm falls below 70% of original spec.

[MWJB]

Environment that the tubes are in, the amount of abuse they get affect longevity...in most cases (semi regular use, realistic bias, not run "balls to the wall" all the time) they seem to last more that 24 months. The thing is you get accustomed to how they sound and don't realise small amounts of degredation...it's a bit like looking in the mirror and thinking, "Who's that fat grey haired bloke?"...it creeps up on you.

Unless you have a bunch of the same brand of tubes (new, used, just functioning) how can you rate their tone...yes that old used GEC KT66 probably sounds great, but how about compared to a NOS tube? Tubes will continue to "function" well after they have stopped sounding like they did in their prime.

A lot of harp players I know don't run their power tubes as hot current-wise as guitar players, 4-5yrs of regular gigging (2-3 nights a week) is no problem for current prod 6L6.

NOS vs modern is probably more relative to brand than age, I've had St Petersburg Svets let go at dissipations they should handle easily (but still recommend them in less arduous situations because they sound good), the venerabe Sovtek 5881 just trundles on & on (no help to you if you don't like the tone though!).

Guitar amp tubes are relatively cheap, accessible - changing them every couple of years is not going to break the bank on your average 4xEL84, 2x6V6 or 2x6L6 amp. Having said that, despite my nagging, I don't know anyone who does regularly retube for reasons of reliability, some do it out of curiosity though. Most wait til their amp barely functions/has trouble finishing a set/starts screaming & humming at them before they bring them to me.

"Do musical instrument amps abuse the tubes so badly that they go to an early grave?" Some do, some don't. There's probably millions of little SE amps with the original tube in them. Can't be many vintage AC30's with original tubes?

"Are there easy design changes we could make to said amps that would make the tubes last 5-10 times longer without compromising tone?" - equal reaction for every action and all that, but humour me - if I told you that running your Super Reverb at 15mA per tube with Sovtek 5881s would make them last 10 or more years would you...or would you fit the tubes you liked the sound of best? If biasing hot gave you a magical tone but meant that tubes only lasted 12 months what would you do?

Perhaps note on a bias pot min & max setting for the tubes you have & set by ear within those perameters...you *may* possibly find that your favourite sound may not be at max dissipation limits? I think though in most cases, guitarists will find the hotter end more pleasing.

Amps are musical instruments, they're ONLY purpose is to make a pleasing sound and be reliable enough for you to be comfortable knowing that they'll make it through the night, night after night for a period that you are happy with.

[dai h.]

I've heard some anecdotes about certain tubes lasting very long such as the GEC KT66 and Mullard GZ34. I don't think the new ones are worth as much compared to the old ones, though of course sometimes you get duds with the old ones sometimes. From a quality control perspective today's stuff doesn't make sense if the tubes need to be sorted after they leave the factory. The old ones had etched or printed codes that said when and where they were made so they could track problems down. Built at a better time for tubes. More, bigger markets, more serious and continuous tube production, more testing on some of the MIL, industrial, special purpose old stock you can buy today, more competition, etc. As an example of a superior old tube, Bendix 6106 ($15-30 or so?). Sort of a super 5Y3GT rectifier tube. If you saw one (special glass, ceramic spacers instead of mica, ceramic cathode sleeve?, etc.) I don't see how anyone would see a modern version as a better value.

[MWJB]

"I've heard some anecdotes about certain tubes lasting very long such as the GEC KT66 and Mullard GZ34." With respect, I have to say thast I have heard anecdotes and had practical experience to the contrary regarding the Mullard GZ34. I am not alone. Even the boutique builders were mainly using the Chinese Ruby a few years ago, before the JJ & Reflector came along.

I think generalisations based on "new" & "old" are really too broad to have any validity today, especially after some stocks have been picked over to the point that it's just the ink on the tube that sells it. Undoubtedly, back in the heyday there were tubes of a consistency and quality that surpasses much (in some cases all, in others not) of what is available today...and many commonly used amp tubes (such as 5Y3 that you mention - what production amp came with a Bendix 6106 by the way?) are still so readily available, at a reasonable price, that there is no real reason to pursue a modern equivalent...if indeed there is such a thing? (The high voltage Sovtek/Reflector "5Y3" clearly isn't, reports on the Reflector 5751 are not favourable either, but then I only like the GE out of the NOS choices).

I have thrown away countless noisy/short lived, military grade, NOS 12A#7 tubes (a friend of mine has dozens of Brimar 12AX7 NIB - ALL useless)...in the end, even after screening by a reputable dealer, cost vs reject rate/lack of longevity means that I'm generally better off buying current production, unless specifically requested otherwise. If the odd current prod needs replacing a few weeks down the line, then it costs less than when the same thing happens with NOS (which it does). Admittedly, for all I have just said, I do succumb every now and then to bit of NOS bait.

I'm sure that we can bat anecdotes back and forth regarding "this NOS tube" and "that current production", but for a lot of folks it's just more practical to use the better current production choices. It makes no more sense to pursue a tube simply because it is NOS than it does to buy a crappy sounding tube because it is current production.

Whilst there are amp manufacturers (hundred+ amps per year) who offer NOS upgrades, are there ANY that can rely on NOS exclusively? Shouldn't be a problem for smaller scale production/keen hobbyists.

[dai h.]

yes I guess it would depend on your circumstances whether old stock is a viable choice and sure they aren't all good or were so and some of the new ones seem okay (Svet EL34, JJ) but (while there seems to be some improvement and promise) the manufacturing stance as a whole seems fundamentally compromised in the sense that there seem to be no standards. As I mentioned with the etched codes, how do they track down defects? How much do they care if they don't have such a system? Can you get a datasheet with genuine specs? I see things like old famous trademarks being bought up and "Mullard" put on "EL34s" which don't seem to have that much to do with the original, etc. Personally I could care less about the brand. The performance should matter more than whether it's Russian or Chinese, or whatever brand it has right? But maybe all of that is inevitable because markets are too small, labor costs too high for skilled manual labor in the "West", not enough competition, etc.

such as 5Y3 that you mention - what production amp came with a Bendix 6106 by the way?

the point there is that a civilian couldn't buy a "super tube" like that but now you can for what must be a whole lot less than it would cost if they were made today. Another example: you can buy those Russian "5881"--original MIL USSR versions--sometimes quite inexpensively. Tubes that because of what they were going to be used for that had extra quality control, testing for reliability, improved designs, parts, chemistry, etc. which new production probably will not go through.

Whilst there are amp manufacturers (hundred+ amps per year) who offer NOS upgrades, are there ANY that can rely on NOS exclusively? Shouldn't be a problem for smaller scale production/keen hobbyists.

yes, not really viable for mass production and more popular stuff gets harder and harder to find, but for individuals, hobbyist DIYers, etc. there seems to be plenty of old stock as whole (but smaller for popular types) to play with. Good used works also. Alternate choices. Same thing with different branding, no brand. Same thing but different package (12AX7 = two 6AV6), or basing, pin outs, "overlooked" tubes. Of couse price may vary A LOT depending exactly where you buy(individuals, dealers, ebay, etc.) plus it makes a BIG difference if you have some knowledge (if you can tell who made them by looking at them regardless of branding, tell the difference between older and newer, etc.).

But back to the original post, I do agree that it would be good to come with practices and/or methods that help extend the life of tubes. Soft start, not dissipating more than you need to, shock mounting, and whatever else this involves, etc.

[Enzo]

Was that 4CX tube running in a CW rig in class C?

Life of a tube is a slippery concept. I just replaced the quad of Tesla EL34s in a MArshall last night. AMp "worked" but sounded terrible. DUll and lifeless, but it worked. A fresh set of tubes and it sounds great. In fact the bias setting coincidentally was already right on, even.

When the military specs 70% or Gm or something for worn out, they are thinking field radios, not screaming guitar solos. They don't care if the "tone" is not quite what it once was. or if there is a little noise in the audio. or microphonic. Does the radio communicate or not?

I used to work with an organ repairman, who went on service calls to old ladies' homes to service their old Lowery organs - or whatever brand. Some of those things had a module for each of the 12 notes of the scale, each module with an oscillator tube and a few divider tubes. Could be a hyndreed tubes in one, literally. All those tube had to do was oscillate. Didn't much matter how weak they got. They could be down in the yelow in the tube tester and still work fine. WOuld have sounded terrible in a guitar amp.

I think "tube life" is about as specific as something like "NOS tubes are better (or worse)."

[MWJB]

Hi Dai,

I agree with just about everything you post here. However...

"Can you get a datasheet with genuine specs?" JJ have data sheets for all their tubes at their website, perhaps not as in depth as some of the classic tube data sheets. I'm not sure what a "genuine spec" is. Most classic amp designs completely ignore limiting values of manufacturers specs anyway.

Indeed, the practice of resurrecting well known brands (in name & cosmetics mainly) should really be recognised for the marketing ploy that it is. That's not to say that there is anything wrong with these tubes in their own right, it's just even if they were made by the same company, after years of non-production, by a different workforce...they probably wouldn't compare directly with the originals. Move the manufacturing operation halfway round the world and source different materials....?

The name gives the consumer a comfort level and the manufacturers (not just tubes either...speakers, whole amps, cars...you name it, the practice is everywhere) realise this. Whilst I don't care much for the ethos, it does mean that people are paying attention to our market and evolving the product range...which is a better situation than say 20yrs ago, with respect to current production tubes.

[dai h.]

my point basically is if they are forthcoming with genuine specifications about their tubes as you see with old manufacturer datasheets ("GE6550", "Tung-Sol 6550", etc.) or manf. datasheets for transistors, ICs, parts with dimensions, max. this and that, etc. Will they say, "this 'KT88' of ours is the same thing as our 6550 in a different bottle"?, or this "12AX7" is a re-based tube which doesn't have exactly the same characteristics of a "12AX7", or "this 5AR4/GZ34 may blow up if you try to use it as such", or "this is BS designed to maximize profit" and so on. I guess the situation is getting better but it seems a bit of a wild west situation. But again perhaps it is about what to expect given the situation. So (given the situation) in some cases finding old stock may give better sound, higher performance, more reliability, longer life, and better value.

Most classic amp designs completely ignore limiting values of manufacturers specs anyway.

You mean like screen and plate, cathode to heater voltages? Maybe the tubes survived because they were better than spec and rated conservatively. Maybe the philosophy was different and they wanted to make good products that lasted (and they could because of the different circumstances) and the ideal tube for someone else is something that fails immediately after the warranty runs out...lol.

[Steve Conner]

Originally Posted by MWJB

if I told you that running your Super Reverb at 15mA per tube with Sovtek 5881s would make them last 10 or more years would you...

The thought of it makes me feel ill... point taken

As far as I know that 4CX250B was in a linear amp that spent most of its time idling at a small plate current.

FWIW, I've been using a pair of the New Sensor "Tung-Sol 6550s" for a while and like them. Even though they look like the standard Russian 6550 guts in a coke bottle, they are nicely made with heavy glass, lots of thick mica spacers, and triple getters, and I like the tone. (Biased as hot as the power supply can stand of course... :P) I compared them against a pair of used-but-good GEC KT88s, and I think they hold up well. I'm tempted to get a set of six if I ever get my 300W bass amp project to work.

I've always used shock mounts in my home-made amps, too, and observed the screen dissipation ratings for EL34s.

[Paul P]

I wish current tube manufacturers would offer accurate and detailed
datasheets for their tubes. I find it hard to believe that they don't,
how hard could it be for them to do so ? Is it perhaps because their
tubes are so inconsistent that they wouldn't meet the specs most of
the time ?

You'd think that someone would have taken the time to measure some
tubes and make their own datasheets. Why hasn't anyone done this ?

Paul P

[dai h.]

it seems some people have. A tech on alt.guitar.amps apparently did trace curves or whatever and from what I gather found say a Sovtek 6550 equivalent to a US 6L6GC (must have made appropriate mods such as supplying extra filament amps, etc.). What I'm referring to above w/the re-based 12AX7, the Sovtek "12AX7WA" internals actually a near-equivalent with bit lower gain (90), internal shield between sections (not connected to anything I think in the Sovtek). Perhaps if they were more honest and said "this is different and isn't exactly a direct sub but is a decent tube and can be made to work better in 12AX7 circuits with the following modifications blah blah blah...", then that would seem more acceptable than "better than Telefunken!" hype and BS-filled advertising. Or with Groove Tubes and their "made in USA" (mostly Chinese made?) reissue GE6CA7s or whatever those were supposed to be. Again, maybe it was actually quite a technological feat to make the original tubes. They had the whole history of inventing and using them for electronics for years(WWII--all tubes), no sub(pre-transistor) to use or develop, lots of effort, money, continuous development and refinement, same people working for years and years, etc. I just think of situations like buying a single French Mazda 12AX7 from the 1950s which I didn't get to cherry pick or anything, taking it home and trying it, and it has no problems with microphonics, sounds good. Kind of makes you think WTF!?

[Enzo]

One reason, if not the main reason, part manufacturers release data sheets is for engineers to use in designs. They want an engineer to design THEIR parts into new consumer products. There is in th overall scheme of things not much new product design using tubes. We think tube amps are important, but to the electronics industry we are a fleck of dust on the annual report.

As to the tube specs themselves,they never represented the absolute max for the tube. The specs were there for table radio designs - don't design your product with over this or that voltage if you expect it to perform on the table top for 10 years reliably.

[dai h.]

We think tube amps are important, but to the electronics industry we are a fleck of dust on the annual report.

yes that's probably something a lot of musicians don't realize. I do see people occasionally (basically) opining, "if someone could make a kick ass tube, they would be rich" but I would guess the reality is the opposite and if they spent the money to really do it to the old standards they might end up poor real quick.

[Steve Conner]

It's true. I bet that even in the musical instrument amp business, most of the money comes from shifting solid-state products.

I used to look down at Marshall's Valvestate amps. They have a nasty cheap feel, it just doesn't seem right when you pick up a Marshall head box and it feels as if it's empty, and then you take the back off and it actually is empty apart from a puny transformer and a few silicon cooties!

But then I realised they are actually very clever and are delivering exactly what the majority of the market wants. Something affordable that kind of sounds vaguely like a Marshall, can rock out in the garage and has a headphone socket so you can practice without waking up the whole block.

I think of tube amps more like hot rods or choppers. They're a kind of culty minority thing, and that's part of what makes them cool.