AlexM
Rank: Duo
Posts: 30 
|
Post by AlexM on Jun 30, 2014 21:43:09 GMT
Hi,
Apart from an obvious fault such as hum, severe microphone or loss of gain that signifies that a valve's life is over, can one tell when it is past its best?
Does a valve have a burn in phase, and then consistent performance for most of its life until wear out, heater failure or something terminal happens?.
I know that the actual life of a valve depends on application, but how do you know if a driver or phase splitter is getting tired other than by using a calibrated valve tester or swapping them out randomly to see if things are improved? .
Cheers,
Alex
|
|
|
|
Post by Eduardo Wobblechops on Jun 30, 2014 22:19:54 GMT
I'd keep spares that you know are good and swap periodically, see if it makes any difference.
|
|
|
|
Post by Chris on Jul 6, 2014 18:52:28 GMT
Most valves usually burn in after about 100hrs then the lifespan varies - usually between 5/10,000 hrs. Having an amp biased properly is important as this ensures each valve maintains roughly the same current flow. There's a lot of nonsense with things like NOS valves(best avoided IMO) but getting them tested is pretty cheap.
From what I can gather they tend to end their lives with a small blue pop and then that's it.
Not the best of answers,I know but just my 2ps worth.
|
|
|
|
Post by lurcher on Jul 6, 2014 19:02:54 GMT
As a valve ages. its transconductance will fall. This is the number in mA/V (milli amps per volts) or mS (milli siemens). As to the effect this has on the amp the valve is being used in, it will vary depending on what the valve is used for. Eventually they will physically fail. The way they fail will vary from valve to valve. The may run away, passing more and more current until the (a) grid melts and comes into contact with the anode, when you will see more or less damage depending on the valve and how well designed the amp is. Or the heater will fail, in which case the valve will just stop working.
|
|
AlexM
Rank: Duo
Posts: 30
|
Post by AlexM on Jul 7, 2014 14:40:12 GMT
Nick,
Ok, so ignoring the terminal failure modes, what is the 'profile' of loss of emission - is it relatively stable after burn - in until the emissive layer of the cathode is exhausted, and what is the effect?
As a case in point, Would an output tube sound or less the same until it is 'used up'?. Would it be any different for a small signal valve used as a phase splitter, driver or buffer?.
I have had an output valve fail due to an internal short - a spacer finger dropped off when a spot weld failed and it dropped into the base. That was quite exciting, but essentially random and unpredictable. I am wondering if I am loosing performance continuously as the valves get used up. If so, at what point does the degradation require a replacement - only during the 'wear out' phase caused by a loss of emission, or at some point before then determined purely by subjective means or by testing.
I appreciate that these may well be questions that cannot be answered, but I'm very interested in your views. The KT120s in my power amp seem to be holding up quite well, and I'd rather not change them unnecessarily for cost reasons, especially if tthere is no real sonic benefit to be derived.
Cheers,
Alex
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 7, 2014 15:12:59 GMT
this seems to be a much asked question. Generally cathodes wear out over time because of the job they do. In such cases(normally) in my experience, the sound gradually loses it's sparkle(technical term). I've changed ageing output valves and the whole presentation has become brighter. I've changed them on the basis that I think they've done enough hours, not because they've presented any obvious flaws. I seem to rememeber reading that small sig valves and phase splitters generally last a lot longer than output valves....10,000 ish versus 3000 ish...check out the accepted publications, Morgan Jones etc. But.... it depends how the valve is configured and how hard it's working in the circuit. Then there is the circuit in regard to start up, slow start on heaters is meant to be good for longevity. Sharp temp changes can cause cracking on the heater element, or peeling of the emissive layer. most of these heater issues may not apply to kt120, it being indirectly heated. If you have a fixed bias circuit regularly measuring the bias voltage against the bias point will give an indication of change in transconductance..the objective metric... I know...no help at all...but realistically I've never read a definitive account of the change in sound. edit: here is what new sensor say, they use the word 'dull', another technical term: www.newsensor.com/faqs19.aspx |
|
|
|
Post by Chris on Jul 7, 2014 19:06:55 GMT
Good answers these fellas. It's stuff like this I like about forums.
Many thanks for sharing your knowledge.
|
|
|
|
Post by lurcher on Jul 7, 2014 20:32:12 GMT
Two things happen as they age, first. the transconductance falls, and this leads to a increase in internal resistance. Second, the peak current the cathode can pass will become limited. The first effect will often (depending on circuit) lead to a reduction in high frequencies, as the higher resistance acts with coupling caps to form a low pass filter. The second effect limits the peak signal the valve can pass, so starts to limit dynamics. Both of these together can be put under the "dull sound" banner.
|
|
AlexM
Rank: Duo
Posts: 30
|
Post by AlexM on Jul 7, 2014 21:41:39 GMT
Thanks all for your views.
Does this degradation happen in a linear progressive fashion from first power - on to wear - out, or is there a plateau where performance is stable, before a sudden, final decline?. An alkaline battery's voltage is pretty stable until the reagents are exhausted - does a valve behave in a similar way?
Cheers,
Alex
|
|
|
|
Post by lurcher on Jul 7, 2014 21:52:14 GMT
Well, ignoring catastrophic failure, I think its generally linear. However it may only be after the valve is about 60% down you start to notice.
|
|
AlexM
Rank: Duo
Posts: 30
|
Post by AlexM on Jul 10, 2014 11:36:46 GMT
Hi, For anyone interested in this topic, I have found an excellent book available as a PDF download titled 'Getting the most out of Vacumm tubes' by Robert Tomer, published in 1960. It covers the following chapter by chapter: Catestrophic failures (glass failures, Heater failures, arcing) Degenerative failures (Gas, Getters, spurious emissions, interelectrode leakage, cathode depletion) Subjective failures (Hum/Noise/Microphonics) . . . followed by chapters on characteristic variables, Selected and premium tube types, why so many tube types?, predicting tu.be performance, tube testers, special purpose types, and methods for lengthening tube life. It's a facinitating and suprisingly accessible piece of writing if you are interested in learning more about this technology, especially as it was written in a period where this was very much alive, rather than from a legacy viewpoint. Hope you find it interesting www.nutshellhifi.com/library/MostVacuumTubes.pdfCheers, Alex |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 10, 2014 15:13:20 GMT
Can valve failure cause damage to amps/speakers?
|
|
AlexM
Rank: Duo
Posts: 30
|
Post by AlexM on Jul 10, 2014 20:16:05 GMT
Yes it can. Damage to the amp is usually limited to coupling caps or bias resistors. You can get some loud bangs through your speakers when things go wrong too.
At least a valve amp is no going to send DC to your speakers like a directly coupled transistor amp can do. A well designed valve amp can be made quite safe but there a lot of not very well designed ones around and working voltages are high if there is a problem.
Having said all of that, I have had a push pull valve amp power tube fail and take out the plate fuse without any adverse effect or damage, and more recently an internal short on a kt120 where a mica spacer support finger dropped into the base of the valve, which was quite exciting in terms of visible arcing, but no damage was done to the amp or speakers. It didn't even blow one of the fuses, which was surprising.
It seems sense to plan for replacement before something breaks, and usually there is either an increase in hum or noise or a change in the sound to indicate that it is time.
Cheers,
Alex
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 11, 2014 2:34:56 GMT
Thanks that's useful to know.
Do you know what effect it could have in a valve dac? I've got a vdt1 DAC, hence the questions.
It uses one 5687 per channel for output and one ef86 for power regulation.
|
|
|
|
Post by lurcher on Jul 11, 2014 19:17:24 GMT
Everything around the ef86 in the power supply can stand seeing any other pin, so there is not much to go wrong if the ef went bad. The output stage is a long tail pair with a shared cathode resistor.in theory it could take out the cathode resistor, but I have never seen a 5687 go in that way. Its only a small valve, not a output pentode.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 12, 2014 2:28:19 GMT
Hi Nick, how's it going dude?
Thanks for letting me know!
|
|
|
|
Post by lurcher on Jul 12, 2014 8:40:46 GMT
No worries, by all means stock up on on 5687's but I think if the huge supplies out there start drying up, there are enough amps using them to justify the new suppliers marking them. In many ways they are just a ecc82 on steroids.
|
|
|
|
Post by gazjam on Jul 12, 2014 8:50:36 GMT
Speaking of valve aging, took your advice Nick on swapping out the year old KT88's for a new set.
Didn't know what I was missing, the old set sounded "nice" but even a year or so's use smoothed out the edges a bit, running not as good as new.
Good call!
|
|
|
|
Post by jandl100 on Jul 14, 2014 9:44:56 GMT
New power valves needed after just 1 year?  Where did I put that transistor amp .....? |
|
|
|
Post by itcosthowmuch on Jul 14, 2014 13:42:30 GMT
It very much depends on the valves application. There is one 300b amp I am aware of (Gamma) that eats 300b's at a rate of about two sets a year, yet there are other amps I have owned that valves last for 20+ years.
If I owned a 211 amp with RCA output valves in it, I would want 30 years life at £1,500 a pair.
|
|
