Power Tube Lengths

Hi.

It is normal for a power tube to be short than the others?  I'm attaching a few pictures showing 1 tube on the far left is noticeably shorter than the other 3.







Thanks,
Ron

There will be a natural production variation just as in any manufactured product.  Are these all the same make and type?  I see you have JJs in your preamp valves, (good choice for me), but we can't see what the power valves are.  Are they all the same make of EL34 and with no variations such as E34L?

And the 6 million dollar question, do they measure up as balanced when you apply the rear TSC bias test and do they sound right?

At the end of the day, if I'm honest that wouldn't worry me too much as long as the amp is sounding as it should.  Good engineering involves making your solution not only reliable but also repeatable which is a forgotten aspect to non-engineers.  That said, I know and value a little "organic fertiliser" in the mix in some areas of guitar playing.  Slight natural differences like this in components can lend a wonderful natural sweetness to the sound if they are in areas where they can genuinely influence it.  I would say that, as the valve is an enclosed vacuum making the internal chamber size pretty much irrelevant, this is unlikely to have any real influence on how well it works as long as it is a matcvhing type which meets production tolerances.

I used to own the GM36 and was crazy into swapping tubes to tailor my sound. With all of the different tubes I tried, I never recalled seeing the power tubes with different lengths. Good to know that manufacturing could cause the same brand/model of tube to be be slightly different. I wasn't aware of that.

This is a brand new GMD40 which came shipped with JJ preamp and power tubes. I was surprised they came shipped this way, but then I read an article from 2 or 3 years ago that stated H&K made this change.

I performed the TCS test and learned that the tubes are ok. Here are the number of flashes for each LED from left to right while looking at the back:

1 - 12
2 - 10
3 - 11
4 - 13

I understand the paring is 1/4 and 2/3. So the pairs are with in the 4 volt tolerance as mentioned in the manual.

I appreciate your response.

Thanks,
Ron

If they are now supplied with all JJs throughout that is news to me too YMM.  Thanks for the heads up on that.  H&K's stance was always, "we supply cheap valves as stock because every guitarist is just going to toss them out as a matter of course".  A more expensive valve set at manufacture, even by a small amount, means increases in the amp's cost at each and every stage of the supply chain as each person adds on their profit as a percentage of what it costs them.  This will mean a significant price increase to the purchaser at the end.  And if they are blindly going to immediately throw them away anyway in favour of a magic set...

[EDIT:]  I just worked out that if the valve set increases in cost by 5% (that's 1 -> 1.05 relative), and everyone is adding on 60% profit (that's 100 -> 160, 1.6x), over say two stages, H&K -> Import Agent -> Wholesaler, that cost is now 2.7x the original. That is a really significant price increase to the guitarist overall.

The modern obsession with valve rolling is something I have always spoken out against as you probably have seen here.  It is a myth that different makes of the same valve have marked differences in tone.  That goes against everything that the electronic industry has set up as a way to keep performance the same across all makes of a valve type.  If it is claimed to be an EL84 then it MUST meet the specifications laid down to be labelled an EL84.  What is forgotten is that there are permitted differences in any parameter within a spread of tolerance allowed on a production line, I believe it is about 20% for valves, but that is nothing to do with differences between manufacturers.  People put in a JJ with an amplification factor towards the top end of the tolerance range, (that's mu, not "gain"), and say a Sovtek at the lower end of the tolerance range and hear a slight difference.  This can't be due to the natural production spread difference in amplification factor, it MUST mean that JJ and Sovtek valves sound different by their nature.  That is non-engineering thinking.  This has been magnified by the many average uninformed (and overly gullible) guitarists who just refuse to take in that this is a factor, and we have the current myth.  If your valves all change tone as they age and parameters drift, which everyone accepts, then where does this leave us with defining a "characteristic tone" for that particular manufacturer?  It's a marketing man's dream as we can see!

Anyway...  Your TSC readings are spot on as you know and you have a decent set of valves in there.  Just enjoy playing them and monitor them over time for any untoward effects.  I think you can pass on worrying over this one.

By the way, here is a link to the post I read where someone else saw JJ tubes in their new GMD40 and a reply from H&K stating JJ tubes ship from the factory now.

https://www.facebook.com/Embdor/posts/i-got-my-hughes-kettner-grandmeister-deluxe-equipped-with-jj-tubes-some-months-a/3461931533826190/

...but the amps’ prices will remain the same.

You don't hear that too often nowadays!

bordonbert wrote:

The modern obsession with valve rolling is something I have always spoken out against as you probably have seen here.  It is a myth that different makes of the same valve have marked differences in tone.  That goes against everything that the electronic industry has set up as a way to keep performance the same across all makes of a valve type.  If it is claimed to be an EL84 then it MUST meet the specifications laid down to be labelled an EL84.  What is forgotten is that there are permitted differences in any parameter within a spread of tolerance allowed on a production line, I believe it is about 20% for valves, but that is nothing to do with differences between manufacturers.  People put in a JJ with an amplification factor towards the top end of the tolerance range, (that's mu, not "gain"), and say a Sovtek at the lower end of the tolerance range and hear a slight difference.  This can't be due to the natural production spread difference in amplification factor, it MUST mean that JJ and Sovtek valves sound different by their nature.  That is non-engineering thinking.  This has been magnified by the many average uninformed (and overly gullible) guitarists who just refuse to take in that this is a factor, and we have the current myth.  If your valves all change tone as they age and parameters drift, which everyone accepts, then where does this leave us with defining a "characteristic tone" for that particular manufacturer?  It's a marketing man's dream as we can see!

As an electronic technician and with a side business in electronics/tube amp repair, I can tell you after many re-tubes on many amps that tubes make a difference.  Not as drastic as a speaker swap, but one can hear the difference,  Some are subtle and others more pronounced.  For example, some years ago I did a tube swap in a Vox AC15 with JJ EL84 vs. Genalex Gold Lion... the JJs were more smooth (or muffled) and the Genalex were louder and with a richer texture.  This was unexpected as I normally recommend JJ tubes but have found the Gold Lions to be the superior tube in the EL84 realm.  I have also had great experiences with the Tung-Sol EL84 as well as their preamp tubes so I now stock those as well as the Genalex.

That's an interesting observation ToneGuy.  My answer to that would be how many of each did you try?  The normally acceptable production tolerance for valve parameters is 20%.  Do you think that amount of variation makes a significant difference?  I think that valves at the opposite end of that spread may well have slight differences in their responses in certain circuits.  However, I always point out that good circuit design includes making sure that device parameters are not the B all and end of all of how the circuit responds.  Forgive me for moving into a different field to illustrate this point but, there are definitely circuits out there which will respond to different transistor types having different gains for example.  If this effect is very marked these are generally poor designs!  It is easy (usually) to design circuitry in any field to be tolerant of and to not have its action changed significantly by varying parameters within the chosen device production spread.

Our range of valves in the guitar world were chosen from simply the most easily and cheaply sourced "radio receiving" types.  Their bandwidth is up in the MHz region despite some internet "common knowledge", (I have seen ridiculous claims of as low as 10kHz for a particular make of 12AX7).  How much difference would a parameter like that have to display between makes, and that would have to be consistently, to make a repeatable audible difference in one of our ultra simple stages?  And valve gain factor variations are also minimised as soon as local feedback is applied.  Yes, I am aware that many stages are designed to bypass it above DC and rely on the valve characteristics.  That is a deliberate acceptance of an aspect of poor design from the direction we are discussing.  It means that any valve of any make will respond differently if it is at the lowest end or the highest end of the scale, not a valve of one make and a valve of another make.  The idea of different makes sounding different means that some are consistently out of kilter with others in terms of their parameters.  It is always a valves' parameters which dictate its action in any circuit not some mystical quality one make possesses over another.

When you say that some differences between makes are subtle I would certainly agree.  If you were to say some are massively different I would disagree.  I have never had demonstrated to me any pair of valves of differing makes which have a massively different effect on tone.  The parameters which are the major contributors to that are well laid down in the valve's type spec and any lessening of tightness to that would be unacceptable in a new valve labelled with that number. And I should also point out it is meaningless to compare the sounds of an old set from one make which you have decided now need replacement with a brand new out of the box set from another make. Would you expect them to sound the same being one in need of replacement and one brand new? What about worn out and new sets of the same make, should they then sound the same? If so why are you changing them?

So one simple question would seem to be how can you be sure you are hearing the difference between the valve makes and not just between different examples of the same type, or perhaps even the effects of differing signal source material, (even playing the same phrase through the same guitar is different every time)?  How many have you tested against each other and of course, how did you appraise the differences between them?  You have to come up with a "non-golden eared" solution to that last point too.  That means using objective non-placebo effect test equipment in a single test bed circuit.  Time delay effects between comparisons, signal source differences with playing into them on a guitar, effects like that make listening comparisons unacceptable in the real engineering world.  They force the test to become subjective and remove objectivity.  I also pooh pooh the concept of the magic "tone capacitor" too based on a knowledge of the components involved.  Gibson also feel as I do on that score but bow to "customer preferences" in a way I don't have to!  I have conducted double blind listening tests with firm believers in the orange drop over the simple cheap ceramic.  With a switchable pair of caps in a single guitar, switched according to a random process, and with the believer playing for as long as he needs to make his mind up which is in circuit, it always shows that when he knows which is in circuit it is obvious that the orange drop is superior as he can just hear it but when he doesn't know which is in circuit a high number of tests show he gets it right pretty much 50% of the time.  Is he just guessing or does he maybe have only one golden ear?  

Now, the more advanced questions would be, what difference in parameters between valve makes can you suggest which would explain such differences assuming they exist.  Troubleshooting, fixing and correcting is one important area, but so is analysing, explaining and improving.  Can you suggest what technical differences could be causing the aural differences you say exist between makes of the same valve spec?

Of course I don't think my little campaign for skepticism and awareness will change many guitarists minds from the belief that this, like a few others, is a major life changing black and white red hot issue when it can be clearly displayed on the test bench that it isn't.  (I am placing my tongue a little in the cheek here, I certainly acknowledge you are not making claims of that sort but there are many uninformed people who do.)  You might be interested in some info in The Tube Swapping Thread - Page 5 in the GM36 forum starting around Post #124.  VoodooJeff was a member here who was a pro musician using GM36s.  He had a deal with Wathen Valves and had some very exotic and expensive valve sets from them.  I have always found the reality of the situation even with 100$+ valves is very much as he describes it in one of his posts, "subtle at best".  He is a very honest and straightforward guy and he comes unencumbered by my own "engineering leanings".

Incidentally I have just looked at the Wathen Cryogenics site again.  They give a little info re their special treatment procedure.

Wathen wrote:Why do you need Deep Cryo?  When an object is cast or formed for the first time, its microstructure is randomly set while cooling. Resetting the jumbled molecular structure to become more uniform and evenly dispersed leads to a wide range of benefits for the part.

No problem with the explanation, just the vagueness of "a wide range of benefits" so let's look at what they say about that.

Wathen wrote:Some of the benefits of cryogenic treatment include:  Improve wear resistance on steels;  Remove residual stresses;  Stabilization and improved material properties;  Improve corrosion resistance;  Longer part life;  Less failure due to cracking;  Easier more consistent machining;  Improved thermal properties and transconductance;  Better electrical properties, less electrical resistance, cleaner signal conductivity

Only the last claim seems relevant to the sound of a valve to me, (reliability and longevity are different of course).  That is pretty much only of concern in areas where ultra low resistance is the key.  Our circuitry is not like that.  We run at sensible impedances where the loss of fractional amounts of resistance in the small metallic parts inside a valve have no effect.  What area of value do your anode load resistors and cathode resistors normally sit in?

Interesting but - relevant?  I'm still unconvinced.