Tube swaps and frequency response

[Audiotexan]

FWIW Stan actually has tubes custom made for him in China, which he distributes under his own brand , and he has all the necessary equipment (and then some) and has traced them for his own use (mainly quality/consistency checking) , so he's not trusting old datasheets and hoping they apply, but checking that the OEM supplier sends him what he asked and paid for and not the factory floor sweeps and rejects.

So he's the only person in this Forum who can speak with authority about what modern tubes actually do.

So assuming he does in fact have the clout that he appears to have in the industry...(I have no reason to doubt that he does, just playing devil's advocate here) Why on earth would he *have to* double check their work. His reputation should proceed him, and the OEM fully expecting him to (re)verify, and thus that OEM knowingly triple-check everything that's slated to be sent to him prior to ever leaving the factory floor. Yet he still has to 'trace them for his own use' (quality/consistency)?? Seems like anything they would send him should be to the same spec (within 2-3%), given his industry status as a man who is known to hold a key to so many doors (ears) in such small but critical distribution circles. If HE can't be assured of quality...then what poor schmuck on the street can?

If he (Stan) would more frequently 'acknowledge' the fact that different tube constructions/methodologies sound different, then I'd be closer to giving him the level of esteem that you afford him. But given that we rarely see such (rare exception below), I'll continue to hold him in high regard, but with the personal caveat that I have to assume that he's obviously still running and playing within several of those in the 'close-knit' circles, and refuses to do/say anything that would jeopardize any of his still-working business relationships (even if once removed -since he's stated point blank that he *paraphrasing here* 'gave that to the girls to run'). Which while that makes sense (both business-wise, and moral obligations to those he's doing/done business with), I can't agree with Stan being the only one who can speak about modern tubes with any authority (this forum or otherwise). Because his perspective is jaded/colored/pre-biased by those continued/current and long-standing associations. At the same time, while I have a lot of experience tube rolling, I'm not claiming to be any kind of "authority" as such. I just feel that your outlook is far too limiting, to only give the moniker "authority" only to Stan alone (especially one so obviously-guarded) who's an (ex?)industry insider.
To me, an authority as such, would be folks in the caliber of Enzo (and this forum has several IMO), who frequently backs up his 'opinions' with stated, and specific facts. Not largely generalities which again, with rare exception, which is what I've seen in my interactions with Stan.
While I *love* the experience, and the occasional inside views that Stan so kindly shares, I'm not likely to put him on any kind of 'authority' pedestal near the level you do until such time as he can shed the chains of 'brotherhood' that his past(?) business has bestowed upon him. Regardless of whether that was by choice or by contract (NDA).

Being involved with GT relating to technical aspect causes anyone with a lick of sense to view his writing with suspicion. I have never run across a brand in electronics where the hype far ran the results so spectacularly. GT tubes are least consistent I have seen of any of the rebranders, proven to be less reliable than other labels buying from the same factories. I always assumed it was because testing was not a high priority for the company, the budget went to promotion.

Well, you just summed up Aspen's role as a 'marketing genius' (but still a lowlife IMO). Now why such distaste/disdain for the man responsible (Myles) for having to wade through the BS that Aspen came up with and decided to spew forth on the (then) unsuspecting public? [Kinda reminds me of what I imagine the relationship between Randy and Mike B to be like. Mike batting cleanup.]

Did you once upon a time have a heated difference of opinion with Myles in person that could be coloring things in your current statements/views on his writing? I ask because the manner in which you state your opinions of him/his writing leads one to think that there's definitely more than meets the eye that you're not disclosing. And by default, my personality asks 'why is that'?

I'll readily admit I have a pre-disposed disdain of those in China and Russia (or elsewhere) that would rather sell current manufacture trash that should be in the destruction bin, yet instead choose to make those .10 on the dollar deals, and push large piles of trash to market that should be in the destruction heap. I would hope that does not include Stan, but if it does, then that's his 'cross' to bare. Would I think differently were I in a country with a 'starving economy', quite possibly. But my experience, and outlook is a sum of what I know, or have known, therefor I have the views/outlook I have. That can only be expected. I just long for the standards of 'yester-year'. I hate the current view of "we've got a cornered market, sell all you can!" There's an old saying that I love: "I've upped my standards, now UP YOURS!"
And I hold that as an across-the-board thought, applicable equally to USA, as to those abroad. It's our own fault as a society, that we've allowed standards to have slipped this far. As 'we' continue to accept it. So I don't hold China and Russia totally to blame, but it started with the manufacturer holding themselves to higher standards in years past (albeit those were largely American, German, and British whose tubes we typically hold in the highest of esteems), so IMO, it should start with the manufacturers now, at least if they truly take pride in what they do. Yet select few countries have managed to hold onto it. Germany still comes to mind FWIW, sadly, none of the aforementioned do audio tubes en masse anymore. Anyways...

Knowing the corporate world, I'd be more willing to bet that Myles Rose' thinking was more along the lines of "Fucking lovely. The boss went to meet with Sovtek/Sino and came back with a .10 on the dollar deal too good for him to pass up...*sigh* Now I have to 'create a scheme to get rid of the trash he's bought....%$#*%(#@&*".

Sometimes you do what you have to do out of necessity to keep food on the table, and simply because 'the boss said so'. So that's my take on that. But good or bad, AP had a large part in keeping a failing industry from completely biting the dust, so at least there's a net positive to his credit.

It is funny how everyone needs to create a strawman to argue against, by ignoring the prior posts, and offering a biased interpretation of what was meant in the posts you guys want to criticize.

I don't think it's a "biased interpretation" at all. I think there's a general disconnect (at least there is for me) in what you're trying to convey, or at least, how you're going about conveying it.

My statements that all tubes "sound the same" is not what I wrote.

You seemingly take offence to the above, yet...at the same time: in my book, by default, this one line means that you do understand and more importantly, imply acknowledgement that there are quantifiable differences in the sound being altered according to the changing of a tube. Or in short, the 'sound of a tube'. Why else would you take the time to correct that, and state it in the way you did?

At the most basic level, this is what we mean (or at least what I see trying to be conveyed) when we say that 'this tube sounds....' (in this amp, in this position).

I agree with your position of not wanting to give the false impression that some make the mistake of attributing 'that sound' in amp X to 'is going to have the same sound' in amp Y. As we most all know, it's highly likely that it will not. But if an amp's topology is similar, then chances should be better than average, that some characteristics will make the transition.

Tubes do not have sounds, they are one part of a complex interaction

This is where my particular frustration on this topic arises. You turn around in the very same paragraph, and say they 'don't have sounds' yet take offence to "all tubes "sound the same" is not what I wrote"??? That just boggles my mind why you can't/won't then make the leap to what some of us are trying to say.

Generally: The way the tube interacts is, at it's essence, "how a tube sounds".
Simply really. But even as I write that last sentence, I fully realize the irony for my having a knack for getting lost in the 'obvious' details as well. IIRC, Mark Hammer recently wrote a GREAT op-ed piece about having a conversation over the net (Demons thread?). It's difficult sometimes. For all involved.

between many factors that combine to create a transfer function that describes the operation of the whole circuit.

We get this. No argument whatsoEVER. But if you would just make the internal translation for OUR sake sometimes when you read us saying how a tube sounds, it'd be nice!
If your mind/thought process won't allow such, then I can totally understand/respect that as such. You just process things differently. I do as well, so I can relate. We're both 'dogs with a bone' that we can't let go of to some extent WRT this topic.

The thing I was hoping would "bridge the gap" between the outlooks here, is there IS a difference between how any given brand/batch run of tube interacts with a given circuit, and that is generally correlated by the masses to how the tube 'sounds', when in technician/engineer terms, we understand that the circuit has it's role in that.

But Joe forum-go'er, will never likely have a curve tracer (hell, most TECHS dont) much less a tube tester. So for them the 'how's and why's are unimportant. The only thing that can be gleened at most general sites is: "how's this tube sound in my amp? what's your experience?" And that get's the ball rolling.

I see no harm whatsoever in someone going into 'random forum' and asking for input on what might sound good in model /abc/, made by /xyz/ amps. As chances are, someone there has tube rolled at least a few to get a general idea of what they like, and then the one asking has to go from there, deciding whether that person's outlook/goal on tone is the same as their's is.

The bottom line of it anyways, is those types of things are typically geared towards a beginner asking questions with his first tube amp. (a good thing -more demand for tubes!) As the longer you're exposed to tubes, and the variety therein, the more informed your opinion becomes.

But bear in mind, this all comes from someone who does his own testing, screening in testers first, then in amps, all the while using his own ears, further compounded by having his own personal likes/dislikes, but also has an arsenal of variety (more total tubes with brands combined, than I've ever seen in stock at my local Guitar Center(s)) that allows/enables me to back up my comparing of brands/runs first-hand. And luckiest yet, barring the few in the 'Shuguang/Sovtek sections', almost all being vintage (60s-80s) tubes [vintage in the traditional sense: 25+yrs old]. Most folks aren't afforded that luxury of resources. I count my blessings daily.

So I'm not as informed as Stan on modern tubes. I can live with that. But I'm also not a dealer, nor reseller, so I don't have a dog in anyone's fight (past or present) to color my opinions. They are completely my own, and based on my own opinions/expectations from what I've come to expect from any given manufacturer or vintage (batch run) of tube. I'm simply an avid collector that appreciates those differences, and marvels at what makes so many of them special in their own right.

[octal]

I know we're mostly talking preamp tubes here, but Randall Aiken did some experimenting with comparing the frequency response and harmonic spectrum of various brands of EL84s in a no NFB circuit.

JJ_EH_tube_tests

IMO, the linear frequency response differences he observed were insignificant for a musical instrument application (<1dB) but the differences in the levels of harmonic distortion (16dB stronger 4th harmonic) may well be audible. The relative degree of 2nd, 3rd, 4th harmonics etc. could make a big difference in perceived frequency response i.e. brightness.

[jazbo8]

We can all rationalize things it COULD mean. But we shouldn't have to fill in assumptions for someone else's pitch.

"I see," said the blind man, so he picked up his hammer and saw.

Hammering on... On Myles' page, in the blue box near the rating tables, he wrote "100 is spec at 250 plate volts, 2 volts of bias and 62.5K of plate resistance". Near the bottom of the page under NOS Quality Studies, he wrote "The first figure is the spread in many samples and how far they strayed from the standard of 1.2 milliamps, a gain of 100 in the case of a 12AX7, 70 for a 12AT7, etc., and the spread in transconductance. The second figure is if the current output was above or below the spec of 1.2 milliamps (in percentage above or below this expected 1.2 milliamps). The last figure is the average transconductance for the sample batch. Keep in mind that for TC, the expected is listed elsewhere, such as 1600 for a 12AX7/ECC83/7025, 2200 for a 12AU7, etc."

And on his new blog, there is an updated worksheet for new production 12AX7, at the very top of the sheet, the same reference figures were repeated again, which confirms (at least for me) that the same benchmark has been used for all his/GT's tests. Also consistent with what Stan and Pedro mentioned above, as of 2013, Ruby Tubes seems to be the current leader in consistency.

[Merlinb]

Near the bottom of the page under NOS Quality Studies, he wrote [I]"The first figure is the spread in many samples and how far they strayed from the standard of 1.2 milliamps, a gain of 100 in the case of a 12AX7, and the spread in transconductance.

That's still not very clear. Is he saying that he set the bias and plate voltage to be the same value in every case, then measured the current, mu and gm? And if so, how is he expressing the variance in three parameters as one single number? Just give us the variance in current, mu and gm separately like a normal person!

The second figure is if the current output was above or below the spec of 1.2 milliamps (in percentage above or below this expected 1.2 milliamps).
The last figure is the average transconductance for the sample batch.

OK, at least the last two figures make sense if he's testing them all at the same bias and anode voltage. Which we don't appear to know...

[teemuk]

I know we're mostly talking preamp tubes here, but Randall Aiken did some experimenting with comparing the frequency response and harmonic spectrum of various brands of EL84s in a no NFB circuit.

JJ_EH_tube_tests

"Various brands" is a bit of overstatement. There were only two brands under the test, and just a single tube representing each.

And now the caveats: Since I only had one set of EH EL84 tubes to compare to the JJ/Tesla tubes, this test is not in any way statistically accurate. In order to properly test the tubes and draw accurate conclusions, a larger sample size is required. This test made no allowances for the variation that can occur from tube to tube or in tubes from different production batches.

And those were pretty important sentences considering the overall results...

Aiken's test is a clear indication of what we basically know already: In real life, in real circuits, one can easily measure and verify that slight differences between two tubes inserted to identical circuit exist, both in frequency response (mainly due to different plate impedances and gain) and in overall harmonic spectrum of distortion (likely again largely for same reasons, now leading to different overall bias point and transfer function). We should not neglect different screen current either: It can have effect on overall distortion and the harmonic pattern of it and apparently the screen current draw of EH and JJ tubes was somewhat different.

...But we would need a vastly larger batch of tube samples to determine anything universal about characteristics / performance of different tubes from different manufacturers.

And he DID test them with negative feedback too. The result was not unexpected: Negative feedback makes the circuit less "device dependent" so device variation's effects to overall performance is decreased.

Most power amps are push-pull with feedback so we could assume that such scheme skews the test results into certain direction. Preamp stages in tube amps, on the other hand, are usually open loop and single-ended so this both increases effects of device variation and also skews the results into certain direction.

[mhuss]

I'm a long time 'meter reader' (versus 'golden ears') guy who has been working on tube amps (on and off) since the 80s.

I have a simple-as-possible Champ style amp I use to test 12AX7s (one 12AX7, one 6L6 output tube, oversized SE OT). I am pretty sure I can hear differences between different 12AX7 brands using this simple setup, which is counter-intuitive to my meter-reader brain. A similar tube swap in a more complicated amp is harder for me to hear. In line with the original poster, I'm still wondering why.

Gain/mu seems the most likely suspect. Any stage with a cathode bypass (aka typical guitar preamp) will exercise the gain limits of a given tube. I'd love to see a double-blind test of different tube types in a simple amp like this with tubes from different factories, very tightly matched by transconductance.

Microphony is a secondary suspect, so I would isolate the DUT as completely as possible from the speaker being used.

[eschertron]

I have a simple-as-possible Champ style amp I use to test 12AX7s (one 12AX7, one 6L6 output tube, oversized SE OT). I am pretty sure I can hear differences between different 12AX7 brands using this simple setup

Are you keeping the signals well into the small-signal range of the tubes? Harmonics give more color to a signal than a subtle frequency response variation. If tube A has different ratings than tube B, one can very well be adding harmonics under test conditions where the other doesn't. I don't have 'golden ears' - not by a long shot - but I too can tell, when swapping tubes in my simple-as-possible champ, that a new production sino is splattering HF energy where a tired old-stock tube isn't. Of course, that's running with obvious distortion by the time the signal gets to the output. And thinking of that setup, I'm guessing that whatever differences there are between two 12AX7 tubes (already multiplied by running through both triodes) may well be emphasized by the power section's transfer function.

Gain/mu seems the most likely suspect.

I agree that it doesn't have to do with the frequency response but a lot to do with variations in transconductance and mu generating more or less distortion when the tube is run at the verges of it's design envelope. I think it's the shape of the transfer function (non-linearity, generating harmonics) that is heard as frequency differences between different tubes.

[mhuss]

Are you keeping the signals well into the small-signal range of the tubes?

Of course not! This is for guitar amps, not a hi-fi preamp. I agree the transfer function is probably the key. I also agree with some previous posters that the surrounding components and voltages play a much larger role when out-of-linear-range performance is being considered.

I have a fairly good VT curve tracer, and one of my 'real soon now when I have lots of spare time' projects is to try to better characterize curves to subjective sound. Based on that pre-existing condition, don't hold your breath, lol

Thanks for the feedback.

[guitician]

I have an old set of speakers that I listen to every day that reveal so much in the music it's amazing. It's the same way listening to any amp I use with them, or any source material.

If I use them for guitar... with tubes going into clipping... All tubes sound like crap!

To make tubes sound good you need speakers that MAKE them sound good!

And any really useful double blind test would only be as good as the speakers that they used.

[eschertron]

I have an old set of speakers that I listen to every day that reveal so much in the music it's amazing. It's the same way listening to any amp I use with them, or any source material.

If I use them for guitar... with tubes going into clipping... All tubes sound like crap!

To make tubes sound good you need speakers that MAKE them sound good!

+1. The speakers used for guitar amps, and to some extent the OT, roll off a huge amount of the higher harmonic content. Some amp builders add even more filtering before or after the OT to help cure the "hash" that the tubes are actually producing when clipping like that. Everybody seems to agree that changing any component (speaker, tube, etc) will affect the sound to some extent. What I'm learning from thinking about the engineering principles in these discussions is how much sonic sculpting is going on "behind the scenes" where the effect is not explicit in the schematic, but a result of the physical construction of the component or of circuit interrelationships (impedance matching and so forth).

[tedmich]

I got a styrofoam flat of 36 GE 12AX7s from my local surplus dealer R5-D3 ($3.25 each) They were 3 flavors from Tektronix: 12AX7, 12AX7A and 12AX7/ECC83 all from the same GE factory I believe.

They all tested the same in my crappy tube tester (Precision 660) but sounded remarkably different in my amp (PV Ultra 60 212). Tube tester is one circuit, amp is another. Amp is arguably more important. I kept the 12AX7A's as they were the brightest to my ear and I need the high frequencies, and every 12AX7A sounded brighter than every other tube (I tried them all).

I imagine the intra lot variation in modern tube production makes it quite difficult to assign any dependable sonic characteristics to a specific tube; the bell shaped curve is wider. The tubes could be selected (post production) for sonic characteristics (lower gain, higher gain, frequency emphasis) but does anyone claim to do this? It would be sensitive to the circuit there too.

[freefrog]

Reading quickly this topic, I've remembered that I had something in my archives for a long time.

Below are the frequency response and THD of a tube preamp with various 12AX7's in V1.

Amp : a heavily modified Fender HRD 2.

The signal is a rather loud swept monophonic wave, sent in the guitar input and collected from the "send" output of the FX loop.

Testing rig used in this case : a dedicated/calibrated custom soundcard.

Software involved in this case: Rightmark Audio Analyzer 6.2.3.


Nothing has been changed but the V1 tube: EQ and volumes settings were the same. This test has been done 2 or 3 times in order to double check what the ears noticed. A V1 has then been selected for this amp and hadn't been changed since this day.


I won't comment nor generalize these results: as always, an "holistic" perspective treating the rig + amp as a whole seems necessary. I just know for a fact that these tubes as V1 in this amp tested in the conditions described above consistently give these curves. And the tonal changes noticed matched these data when they have been colected.

FWIW.

[pdf64]

As both HRD V1 sections are in cascade, and have fairly heavy loads that vary with frequency, might the differences be largely due to differing plate resistances?

[g1]

Reading quickly this topic, I've remembered that I had something in my archives for a long time.

Below are the frequency response and THD of a tube preamp with various 12AX7's in V1.

This is quite interesting with the THD being quite different in the 300 to 1Khz regions. What seems to be missing is some numbers on the left side of the graph showing % figures for THD. Any idea what those would have been?

[freefrog]

This is quite interesting with the THD being quite different in the 300 to 1Khz regions. What seems to be missing is some numbers on the left side of the graph showing % figures for THD. Any idea what those would have been?

The db scale in these plots is there for frequency response and for THD. So, THD in these spectrums would fluctuate between 0.04% min and 50% max, if memory serves me.

FWIW, it reflects what I hear(d): the mids are the most transparent with the Amperex - especially since I've modified the tone stack of the amp :-))

I'll try to find another series of screenshots with the same tubes in a miked amp...