[PaulFolbrecht]

Was listening this evening when my right channel slowly faded out.  Something wrong.  Turned off left channel; right is playing, very faintly.  Ok, troubleshooting:

- Switched ICs - nope.
- Switched pre-amp channels - nope.
- Switched monoblocks - yep.  It's the amp.

These are Consonance Cyber 845 SET monoblocks (see my pic) that I've had for 8 months or so.  Hoping it was a tube, I start replacing them, power tube first (nope), then the 5867 input - yup.  Replacing that fixed it.  Yes, the tube was still lit - it looked completely normal, just like the other amp.

This is my first time experiencing a tube failure.  It ought to be very rare with small signal tubes, no?  This was a NOS Phillips that supposedly was new.  So, no more than 400-500 hours on it.  Of course, it ought to be good for 10-20x that.  Of course, it may not have been new - it's always a crap-shoot and I sure don't have a tube-tester.

Anyway - yes, maybe I am actually going somewhere - I got to wondering how common this is?  Premature tube failure in non-power tubes?  Extremely rare?  I have only been listening to tube year for 15 months or so.

[Pez]

I had a 12ax7 fail on me that was rather "new" it was a JJ ECC83 to be precise. I'm not sure if it was the fact that I had to box it up and move to a new place or if the tube just died young, but it started to have a really nasty hiss and the tube itself would light up and dim based on what the music was doing (bass pasages would make the tube glow brighter when it would hit). Needless to say I didn't take a risk and had to switch out the tube. Unfortunately that also meant having a mismatch of Sovteks and JJ's. .  Any way I won't say it's the same situation, but tubes have so many ways they can fail, just be glad your equipment was not damaged!

[PaulFolbrecht]

What was weird (to me) was how the amp was still working, music playing, sounding normal, just at a vastly reduced volume.  Had not heard of that failure mode before.  I guess the signal tube was doing no or little voltage amplification.  And I do mean "I guess".

I threw out the matching tube as well - one tube of a pair is useless.

[Pez]

I threw out the matching tube as well - one tube of a pair is useless.

Especially if you're like me and can never remember which tube failed in the first place. 

[slbender]

Hi Paul,

First there are many ways a tube can fail, and it depends on how hard they are driven, how high the voltages are, and how high the current levels are, and beyond that, perhaps just plain luck.  One of your 5687's could have been losing vacuum, or become gassy, or maybe the current was set too high through it.  To fail in a few months maybe it was the tubes' fault, or maybe the amps' fault.  Using another brand 5687 tube will tell.  If it fails within a year or so, then its likely it was  the amp.  If it survives, well maybe it just was that Philips tube. Tubes can last for tens of thousands of hours, or even decades if treated right, or they can fail in weeks or months if highly overclocked in terms of voltage and/or current.

I designed a SET Amp and I used the same tubes for around 9 years. About six months back I switched the input tubes (from NOS USA's to used W.German Telefunken's).  Happens, I prefer the sound of the Telefunkens, and these were used tubes which probably had 30 years of use already.  So you just never know.  So while there probably was no need to throw out your "other" Philips 5687, someday soon you might need a spare and regret throwing it out. 

Tubes don't last forever, but hey, transistors also have certain failure modes too.  Typically some 30 yo transistors need replacing with the same symptoms as some tubes, more hiss, lowered gain. And overloaded transistors blowup immediately, at least tubes usually tend to be more forgiving - glowing red and taking their time.

Interestingly a friend of mine recently had a 5687 tube (NOS Amperex World Logo) fail in a 211 based made-in-china amp. He also has had a problem biasing the NOS 211's, which are GE's, and I supect that his Chinese Amps are either driving the input tubes too hard, or pulling too much current through the 5687's because the 211's have got a problem, are gassy, or out of spec in some way.  In a simple circuit, things tend to interact with each other more so than in more complex circuits.

Currently, just for fun, I'm making an SET amp that is all 5687's - input, driver and output tubes.  It will be interesting, as I have a box of equipment pulls, more than a dozen 5687's that were used for many years, maybe 10, 20, or 30 years, or more, and I'll be using them in this amp.  So it will be interesting to see if I ever encounter any problems or failures.  aa


Steven

Was listening this evening when my right channel slowly faded out.  Something wrong.  Turned off left channel; right is playing, very faintly.  Ok, troubleshooting:

- Switched ICs - nope.
- Switched pre-amp channels - nope.
- Switched monoblocks - yep.  It's the amp.

These are Consonance Cyber 845 SET monoblocks (see my pic) that I've had for 8 months or so.  Hoping it was a tube, I start replacing them, power tube first (nope), then the 5867 input - yup.  Replacing that fixed it.  Yes, the tube was still lit - it looked completely normal, just like the other amp.

This is my first time experiencing a tube failure.  It ought to be very rare with small signal tubes, no?  This was a NOS Phillips that supposedly was new.  So, no more than 400-500 hours on it.  Of course, it ought to be good for 10-20x that.  Of course, it may not have been new - it's always a crap-shoot and I sure don't have a tube-tester.

Anyway - yes, maybe I am actually going somewhere - I got to wondering how common this is?  Premature tube failure in non-power tubes?  Extremely rare?  I have only been listening to tube year for 15 months or so.

[zacster]

It is amazing to me the misconceptions about tubes sometimes.  Tubes do not come in pairs.  You can always find another to match.  Yes they can last 30-40-50 years, or they can fail early.  So can light bulbs.  These things were made by the millions, remember, every early TV had them, and every radio, and a lot of instruments.  Some will fail.  The military had higher specs for their tubes to minimize failure amid the rugged use.  Their was an army of TV and Radio repairmen that didn't do much more than replace tubes all day.

As for new production tubes, they do have a high failure rate. QC just isn't what it used to be.  We're lucky someone is still making them.  Certain brands of certain types are notorious for failure.  I've had 2 Sovtek 6B4G tubes fail before I had a quad of good ones that have lasted over a year now.  They should last for a good long time.

[Pez]

I'll agree on the level of "useability". Yes a tube can potentially last for a very very long time, but unlike a light bulb which will either work or not work a tube is infinitely more complex. The sound will change drastically over that period of time.  So much so that for the hifi enthusiast tube changing is often times a must.  It sort of reminds me of guitar strings.  Yeah you can use a guitar string for a long long time without one actually failing, but have you listened to a set of strings that are played on every day for 6 months? YUCK!

[SET Man]

As for new production tubes, they do have a high failure rate. QC just isn't what it used to be.  We're lucky someone is still making them.

Hey!

  You are right about that. But yes we are lucky that tubes are still being made today.

   I've never had any tube really just die on me. But I did had some new production tubes... JJ 12AX7 and EI Elite 12AU7... went noisy after a few days of use But I other than it fine.... I'm currently have 4 EI Elite 12AX7s that work fine since late 2002.... hmmm maybe I need to replace them again soon.

   If the new tube is a bad tube it is likely to act up after a few days of use. Otherwise it should be fine and work well until it run out of gas.

   I guess these are pretty much small part of the fun or frustration of owning tube equipments 

   But! That is okay with me... so far so good. I just can't see myself going back to an all SS system anymore

Take care,
Buddy

[aerius]

Anecdotal evidence of course, but based on what I've seen over the last few years, tube related issues in Chinese tube amps is uncomfortably common.  Whether it's bad circuit design, poor assembly & QC, or just plain bad luck I don't know, but out the roughly 20 pieces of Chinese tube gear I've seen, about 7 of them have suffered from various problems ranging from severely limited tube life, channels going dead, to the extreme of a tube arcing over and frying.

[PaulFolbrecht]

Anecdotal evidence of course, but based on what I've seen over the last few years, tube related issues in Chinese tube amps is uncomfortably common.  Whether it's bad circuit design, poor assembly & QC, or just plain bad luck I don't know, but out the roughly 20 pieces of Chinese tube gear I've seen, about 7 of them have suffered from various problems ranging from severely limited tube life, channels going dead, to the extreme of a tube arcing over and frying.

Brands?

There has been and still is some junk coming out of China.  But, now, also very high-quality stuff (EE, Cayin, PL), and I think Consonance is at the very top of the heap quality and sound-wise.

[aerius]

Brands?

There has been and still is some junk coming out of China.  But, now, also very high-quality stuff (EE, Cayin, PL), and I think Consonance is at the very top of the heap quality and sound-wise.

A couple ASL's, a Consonance, a DarkVoice, a Dared or Dared clone, and I can't remember the others off the top of my head.  Mostly towards the lower end of the price spectrum.

[PromitheusAudio]

Paul

If you are getting a replacement hunt for a tungsol 5687. i used them sonically one the best i compared so far. I tried various NOS. Stay far way from most JAN PHILIPS tubes. Sonically they aren't that grade. Can't comment on life span. I generally don't used them for so long.

BUt i believe your tungsol 5687 could be the driver tube for the 845? if that is the case the 845 tube might be pulling too much current when going into class a2 for the 5687 to handle.

Or important check and see the voltage output of the power transformers.Some of this power trans may not be made for 120v but perhaps 115 or 110. This could put out too much voltage and kill the tubes faster. If you don't check on this putting more tubes is like sending them to their death bed slowly. Like here, in Malaysia, many chinese amps are rated for 220v but mains is 240volts. That would certainly spell doom for the tubes and the life span of the tubes.

Also chinese companies have this funny idea, they will spend on boutique parts and good/decent opt but in the end compromise on the power transformer manufacturing part. Even though the sound of the amp comes a huge part from the power transformer.


Or you are just plain unlucky to have a tube that failed QC and escape the test. In sense you had an "escapee" tube.  No matter how much of tube checkers or testers some tubes and semiconductors often escape in sense, up to 9 types of testing phases. I have worked and thought at 2 ex-motorolan companies and experience many times ic's escaping multiple testers phase.


Hope this info helps

[PaulFolbrecht]

If you are getting a replacement hunt for a tungsol 5687. i used them sonically one the best i compared so far. I tried various NOS. Stay far way from most JAN PHILIPS tubes. Sonically they aren't that grade. Can't comment on life span. I generally don't used them for so long.

Don't think I have a pair of tungsols.  I'll try to grab a pair.

BUt i believe your tungsol 5687 could be the driver tube for the 845? if that is the case the 845 tube might be pulling too much current when going into class a2 for the 5687 to handle.

Yes, it's the driver (6922 is the input).  These amps do NOT drive the 845 into a2.  I don't have an engineer's understanding of amplifiers, but I know that a driver with much more voltage swing is necessary for that.

Or important check and see the voltage output of the power transformers.Some of this power trans may not be made for 120v but perhaps 115 or 110. This could put out too much voltage and kill the tubes faster. If you don't check on this putting more tubes is like sending them to their death bed slowly. Like here, in Malaysia, many chinese amps are rated for 220v but mains is 240volts. That would certainly spell doom for the tubes and the life span of the tubes.

I'll check but I think that is highly unlikely.

Also chinese companies have this funny idea, they will spend on boutique parts and good/decent opt but in the end compromise on the power transformer manufacturing part. Even though the sound of the amp comes a huge part from the power transformer.

Of course.  Any amp is 9/10ths power supply, right?

When I was researching these amps, I talked to a gent in England who owns a pair.  He is pals with one of the designers at Art Audio, not Joe Fratus, but the other one, whose name escapes me.  Anyway, this guy tore the amp down to look at the parts and concluded that they "hadn't skimped" on the all-important power and output trannies.  They weigh about 50lb each, which is decent for an 845 monoblock, so that sounds fairly accurate.

Also, I do not think it's likely they'd sound as huge as they do, with quite impressive macrodynamics, on 90dB speakers if there were serious compromises in this area.

FWIW, they draw 150W from the wall.  Rated at 28W output (more likely more like 22W), that is a healthy power supply, no?

Or you are just plain unlucky to have a tube that failed QC and escape the test. In sense you had an "escapee" tube.  No matter how much of tube checkers or testers some tubes and semiconductors often escape in sense, up to 9 types of testing phases. I have worked and thought at 2 ex-motorolan companies and experience many times ic's escaping multiple testers phase.

That's what I'd like to think.  If another driver goes within another few hundred hours I'll start to wonder.

Paul