Solid state/class d vs tube amps

[bucephylus]

I doubt that there is data confirming predictable shorter tube life. If tubes fail prematurely under normal usage it is probably due to sloppy QC.

Right; that’s what I would suspect. The perception is due to tube to tube variability vs actual construction.

 

[agedhorse]

Right; that’s what I would suspect. The perception is due to tube to tube variability vs actual construction.

Much of the perception comes down to marketing by various tube sellers combined with the feelings about the good 'ol days that maybe weren't as good as they were remembered.

Well built tubes are well built, no matter when they were built.

 

[Drgonzonm]

I doubt that there is data confirming predictable shorter tube life. If tubes fail prematurely under normal usage it is probably due to sloppy QC.

There's quite a bit of data out there, most of it 50 plus years old. Guitar amps typically have their power tubes with voltages above design Center voltages but below max values.
A design that is common is the voltage on the cathode is near or higher the design max between cathode and heater, resulting in arcing. Most common in preamps
While there is a debate whether standby switches are necessary, turning an amp off and on takes a toll.
Some body commented about amps and vibration, combos will have shorter life than a piggy back.
Placing power tubes too close together shorten tube life.
Choosing the wrong socket to build an amp with has caused more than one tube failure.
Some of the amps using EL34 tubes have shorter than expected life.
And then again tubes follow the mean time before failure curves.
Using tubes that require a warmup period, with out the warm up period shortens tube life.
Yeah the data is out there, it's typically published by reading western electric, mallard Telefunken and others.
On some of the older amps, the age of the electrolytic caps can mean tube failure upon startup. If you barn find an old tube amp, be careful, plugging it in and turning it on can cause some expensive damage.

 

[agedhorse]

I have plenty of data, plenty of analysis, and I can factually state that much of what's being said about short tube life (one or two or three years) is complete (and insulting to the general bass/guitar player consumer) bull-dookie.

Of course, if an amp gets dropped, mechanical damage can result (like a broken tube is the fault of the tube, especially when it's bouncing around inside the case), of course there are a few poor design examples that can result in premature failures (not the tube's fault), poor quality or defective tubes, but these are generally exceptions to the rule.

I've been working with tubes (at the engineering level not as a hobbiest) most of my professional design life, and my dad designed tube circuitry back in the 50's and 60's so personally I am well aware of what its true and what is fabricated. Also, the company I work for has been designing and selling tube amp products for ~45 years as well. After working with literally MILLIONS of tubes, we just might know a thing or two about factual reality. The parent company of the company I worked for previously has worked with TENS OF MILLIONS of tubes as well, virtually no difference in the validation of data.

This is a far cry from folks who may base their data on their feelings and maybe a a couple hundred tubes (giving the benefit of the doubt here).

 

[Drgonzonm]

I'll take rca's word.
Mesa boogie has an excellent reputation, and excellent quality. But IMO, MG dI'd not nor does'not push tubes like Jim marshall.
A quick perusal of mesa boogie indicates a lot of el84, 6l6 kt tubes, while MG does use el34's,
Most tubes have expected average life's of 10,000 hours. if operated within design Center specs. Guitar amps are typically run at higher voltages than design Center. If you build an amp whose tubes last 10-20% better than than the average guitar amps, life expectancy of the tubes is still lower than expected life stated by the original manufacturer. Is this bad, not really. We are willing to trade performance for life. And let's not forget mean time before failure. The more components, the shorter the life of an amp.

 

[Pocket4]

But will it make me feel groovy? I think it is still relevant.

 

[agedhorse]

The more components in an amp does not mean shorter life.

It's the designer's responsibility to consider the reliability/lifespan of each part and from all of the parts a complete life expectancy can be calculated. Each component has different life expectancy.

 

[JimmyM]

But will it make me feel groovy? I think it is still relevant.

Nope. All that matters is it weighs less than 2 lbs, has 5000w and never needs repair ever.

 

[Drgonzonm]

The more components in an amp does not mean shorter life.

It's the designer's responsibility to consider the reliability/lifespan of each part and from all of the parts a complete life expectancy can be calculated. Each component has different life expectancy.

No argument.
Just a question. How do you assign failure rate for an amp that is used in Orlando or Tampa FL compared to an amp played in Leadville CO. I chose these towns because of the different environmental factors. In central FL the quality of electricity shortens electrical component life. In Leadville the altitude challenges the designer's cooling assumptions.
While these may seem trivial, VW had a number of type I'V engine failures on there bus'es. When operating in the Rockies.

 

[agedhorse]

No argument.
Just a question. How do you assign failure rate for an amp that is used in Orlando or Tampa FL compared to an amp played in Leadville CO. I chose these towns because of the different environmental factors. In central FL the quality of electricity shortens electrical component life. In Leadville the altitude challenges the designer's cooling assumptions.
While these may seem trivial, VW had a number of type I'V engine failures on there bus'es. When operating in the Rockies.

Fortunately, I'm not designing engines.

There is manufacturer reliability data on all of the components that are used. There are also design metrics that specifically consider the variables within the operating environment. All of this data, along with the operating parameters of the circuits are then combined to create a composite overview of each component's operating conditions and predicted lifespan. While doing this, there will be times where a particular component show up as being unsuitable for the operation because its predicted lifespan within the circuit is significantly less than the surrounding parts, so this weak link must be addressed.

While it all sounds complex (and it really is), fortunately many components can be dropped out of the total reliability equation because the stresses on it are so low that it approaches the shelf life. For example, if the MTBF for a part is say 50,000 hours under the application conditions and it conforms to the typical aging/failure curves, it can probably be ignored completely because it's well beyond the expected lifespan of the product. Now, a smart designer might find that it costs no more to "over-design" because almost EVERYTHING drops out of the reliability equation and the practical reliability falls to a few key components that are out of the control of the manufacturer, or become standard repair/maintenance items over the extended life of the amp. I'm talking about things like fans, tubes, etc.

There's a reason why amps have become so much more functional, more powerful, more reliable and less expensive with a better warranty. It's things like reliability analysis that go a long way to give the manufacturer confidence that their product won't fail in unreasonable numbers. This is the sort of things that engineers in larger companies do when designing an amp, and things that boutique, or hobbiest based companies typically don't understand or can't afford to do. It's time consuming and costs money to complete.

 

[Bflat]

Are you familiar with Mil std 217 E ???

 

[Drgonzonm]

I've played with enough statistics to know to test the robustness of assumptions. I look at tube data and see the same graphs and data that was published fifty years ago.
I believe you are privy to some manufacturers info that those of us in the general public do not have. I also expect that your company has generated data for design purposes, that the manufacturer does not have.
I am familiar with motorla's statistical process control and the six Sigma principles, while I am part of the Choir , thank you for the explanation, as many of the TB'ers do not have your manufacturering and design. experience. I'very been away from statistical design applications for a little more than a decade, so I am a dinosaur.

 

[DavidEdenAria]

Regarding your statement about old production tubes vs new production tubes:

First, what construction factor do you think mainly limits tube life?

Second, do you have data that current construction methods systematically lead to predictably shorter life?

I just wonder sometimes with all the mythology surrounding tubes whether players sometimes understand that these are just engineered devices. They aren’t some kind of mystical device that issues forth from the psychic vortices.

My $0.019 without looking into the matter is that two things can lead to premature tube degradation. One is the glass to metal seal, which allows oxygen to leak into the tube. The other is susceptibility of the internal plates to vibration. On the latter point, the musician can do a lot of good to isolate the head from the cab, especially at high volumes and with cabs having less than optimal bracing. On the former point, I’m somewhat dubious that glass to metal sealing technology has gotten worse over the years; though I suppose there could be QC issues. Is any of that kind of data known or published? Or are folks just going on experience with a few parts?

Regarding your statement about old production tubes vs new production tubes:

First, what construction factor do you think mainly limits tube life?

Second, do you have data that current construction methods systematically lead to predictably shorter life?

I just wonder sometimes with all the mythology surrounding tubes whether players sometimes understand that these are just engineered devices. They aren’t some kind of mystical device that issues forth from the psychic vortices.

My $0.019 without looking into the matter is that two things can lead to premature tube degradation. One is the glass to metal seal, which allows oxygen to leak into the tube. The other is susceptibility of the internal plates to vibration. On the latter point, the musician can do a lot of good to isolate the head from the cab, especially at high volumes and with cabs having less than optimal bracing. On the former point, I’m somewhat dubious that glass to metal sealing technology has gotten worse over the years; though I suppose there could be QC issues. Is any of that kind of data known or published? Or are folks just going on experience with a few parts?

Metallurgy and sloppy quality control at times.

This info is from several tube vendors that sell NOS AND new production such as Mike at KCA NOS tubes......he has sold way over 20,000 and buys and screens both types.....feedback from customers and for myself, 17+ years at tube amp forums gathering info from 1000's of tube amp users in the guitar world, of which there are way more than bass amp tube users.

Many tube vendors today dont sell both NOS, strong NOS pulls (or used NOS types) and new production....if you talk to Mike or search his posts at the 'the gear page' or 'plexi palace' he describes the issues he has had with rejects of new production tubes thru the years and outright rejecting entire runs.

There is no doubt some new production has improved over the years....i havent talked to Mike in a while fwiw.

 

[DavidEdenAria]

Watford valves in Great Britain is another reliable vendor selling new and old production.....to loosely quote them, some of the Ei power tubes were absolute junk and unreliable....Ei 12ax7 have a long reputation of going microphonic and varying in gain....issues with early Chinese power tubes as well as early Chinese 12ax7 being noisy and fizzy sounding...inconsistent quality control for many years since the 1980's

Bias drift (which is normal to a degree) and the common complaint of certain new production 6L6GC in the last 20 years or so, "going flat" or "losing dynamics" when the tubes still tested strong with decent getter flashing remaining, despite rebiasing.

Tubes back in the day had to pass many stringent parameters to be deemed military spec or industrial graded....tubes did get rejected and were sent to dumpsters.

Look, NO DOUBT new production tubes have improved....but from some of the smaller vendors that still sell many 1000's of tubes, too many inferior tubes hit the shores that would have been rejected when tubes were a way of industrial life.

 

[bucephylus]

Metallurgy and sloppy quality control at times.

This info is from several tube vendors that sell NOS AND new production such as Mike at KCA NOS tubes......he has sold way over 20,000 and buys and screens both types.....feedback from customers and for myself, 17+ years at tube amp forums gathering info from 1000's of tube amp users in the guitar world, of which there are way more than bass amp tube users.

Many tube vendors today dont sell both NOS, strong NOS pulls (or used NOS types) and new production....if you talk to Mike or search his posts at the 'the gear page' or 'plexi palace' he describes the issues he has had with rejects of new production tubes thru the years and outright rejecting entire runs.

There is no doubt some new production has improved over the years....i havent talked to Mike in a while fwiw.

In fact, the glass to metal seal is mainly a ceramic problem; but, we can just generally lump it into being a Materials Science problem.

Regarding NOS vs new production tubes, I am rather more inclined to believe what I have heard from both Ron Wickersham and Andy Field (@agedhorse) regarding new production tube reliability than Mike at KCA NOS, who apparently makes his living by selling NOS tubes.

Nothing against NOS tubes. I’ve messed with a whole slew of 5751’s getting sounds I like. But, let’s be honest about where the differences are and aren’t.

 

[seamonkey]

There seems to be an unlimited supply of NOS tubes or there's a lot of counterfeits.
Could be semi-NOS. The telephone exchanges use to do regular maintenance and swap out all the tubes at regular intervals. Seemed this cut down on outages due to failing tubes. In their use case, tubes didn't last forever. Radio stations, Military and other industrial use of tubes swapped them out for maintenance intervals too.
Trade-ins or Dumpster diving and retesting probably saved a lot of these swapped out tubes for re-use.
So these stockpiles of old NOS tubes are a good thing to have.

 

[agedhorse]

With new production tubes, I have discovered over the years that if (as an amp manufacturer) you pay a fair price for a quality tube that's usually what you receive. If you try to get a "deal", you often get exactly what you pay for...

Also, from the engineering side of things, if you design according to the gold standard tube specs. and do not color outside the lines without understanding what and where the statistical boundaries are, you are not likely to run into the constant problems that some folks here seem to suggest.

If I hadn't been doing this for so long (pre Mesa), and didn't have many thousands of amps in the field to base my opinion on, I might not be quite as confident in my position.

 

[-Asdfgh-]

With new production tubes, I have discovered over the years that if (as an amp manufacturer) you pay a fair price for a quality tube that's usually what you receive. If you try to get a "deal", you often get exactly what you pay for...

Also, from the engineering side of things, if you design according to the gold standard tube specs. and do not color outside the lines without understanding what and where the statistical boundaries are, you are not likely to run into the constant problems that some folks here seem to suggest.

If I hadn't been doing this for so long (pre Mesa), and didn't have many thousands of amps in the field to base my opinion on, I might not be quite as confident in my position.

I presume all the 'tricks' to produce an all-tube sound from a solid-state or mostly solid-state design have been tried? Other than using filtering as previously suggested by another poster, is there anything we could be trying? I don't have an tube amps now to A/B against modelling, to see how well that is doing now in a reasonably objective way. I do wonder, at least for live sound, whether it's possible to distinguish, but then when I see a band I don't know what the original amp sound was in the first place, to know whether it has translated well.

 

[agedhorse]

I can't discuss specific "tricks" or approaches behind any of the solid state and/or hybrid technologies I might be doing research on for obvious reasons.

I can say that over the next 5 years thereay be some very interesting now technologies available as we (as an industry) better understand and develop alternative models to traditional circuits. Analysis tools and more advanced math models have been responsible for many of the improvements in the DSP based modeling world, and these tools can certainly be used in reverse to redefine (or refine) analog design approaches. I'm doing things now that I would not have thought possible because the math was just too complex and time consuming to even consider solving.

It's not just an exciting time to be a bass player but also to be an amp designer.

 

[Korladis]

With new production tubes, I have discovered over the years that if (as an amp manufacturer) you pay a fair price for a quality tube that's usually what you receive. If you try to get a "deal", you often get exactly what you pay for...

Also, from the engineering side of things, if you design according to the gold standard tube specs. and do not color outside the lines without understanding what and where the statistical boundaries are, you are not likely to run into the constant problems that some folks here seem to suggest.

If I hadn't been doing this for so long (pre Mesa), and didn't have many thousands of amps in the field to base my opinion on, I might not be quite as confident in my position.

Indeed.

It's been my experience that people vastly overstate how often you have to change tubes.

If I was in bands that played only at reasonable volumes, I probably would have only needed to replace my output tubes once, if at all, in the time I've owned my D-180.