Forgive me if this has been discussed elsewhere.
I’ve been searching for the reason why some tubes “sound bright” whereas some others have rolled off highs, etc. Now, I know that tubes per se don’t “sound” at all, so humor me for a moment. When I read that X tube sounds bright in a reviewer’s opinion, and then try it in my amp, I get (usually) the same result, comparatively speaking.
Is it the mechanical/distances of the tube? I thought that was pretty much dialed into the tube type.
Plate size, shape, or configuration?
Is it the cathode coating? If so, what is it about the cathode coating makes one brand’s “house sound” different from the next? For example, new production Svetlana EL34s “sound brighter” than the JJ versions I have, in conformity with others’ reviews. Supposedly new production Tung Sol have a “darker” sound compared to the Svetlanas—I don’t know, I never tried them.
Is the brighter sound actually intermodulation distortion that my ear picks up as a brightness?
Moreover, what is it about an EHX or JJ 6ca7 that makes the bottom end sound fuller/bigger than a EL34 (at least the ones I’ve tried). Is it the beam-forming plates? Why?
Thanks.
I’ve been searching for the reason why some tubes “sound bright” whereas some others have rolled off highs, etc. Now, I know that tubes per se don’t “sound” at all, so humor me for a moment. When I read that X tube sounds bright in a reviewer’s opinion, and then try it in my amp, I get (usually) the same result, comparatively speaking.
Is it the mechanical/distances of the tube? I thought that was pretty much dialed into the tube type.
Plate size, shape, or configuration?
Is it the cathode coating? If so, what is it about the cathode coating makes one brand’s “house sound” different from the next? For example, new production Svetlana EL34s “sound brighter” than the JJ versions I have, in conformity with others’ reviews. Supposedly new production Tung Sol have a “darker” sound compared to the Svetlanas—I don’t know, I never tried them.
Is the brighter sound actually intermodulation distortion that my ear picks up as a brightness?
Moreover, what is it about an EHX or JJ 6ca7 that makes the bottom end sound fuller/bigger than a EL34 (at least the ones I’ve tried). Is it the beam-forming plates? Why?
Thanks.
A lot of that jargon that reviewers use is pure nonsense to make people nuts.
Bright/dark/soundstage/ etc etc... endless boutique words.
The way an amplifier "sounds" with various tubes (tube-rolling?) is mainly the result of the circuitry itself.
To make it simple - a particular amplifier was designed with specific tubes in mind, to perform to the designer's specs.
Changing to different tubes will naturally change its performance.
An EL84 of one brand will behave differently under different biasing, loading, and HV conditions in different amps.
Bright/dark/soundstage/ etc etc... endless boutique words.
The way an amplifier "sounds" with various tubes (tube-rolling?) is mainly the result of the circuitry itself.
To make it simple - a particular amplifier was designed with specific tubes in mind, to perform to the designer's specs.
Changing to different tubes will naturally change its performance.
An EL84 of one brand will behave differently under different biasing, loading, and HV conditions in different amps.
It is called synesthesia when predicates from one modality are used to attribute something to different sensors. Like sharp sound, meaty music, and so on. There is no agreement among audiophiles about them. And of course tubes do not bear such properties, especially when related to different frequencies. Tubes do not care what frequencies how to distort, until wavelength is getting closer to sizes of electrodes, way-way beyond audio band.
Most laughable advertisings are about sound of mica layers, color and shape of a getter.
What makes the bottom end sound fuller/bigger, properly 6CA7 that you tried in your amp somehow caused lower output impedance of your particular amp, may be it had higher transconductance than your El34, so it damped resonances of your woofer better. Or may be vice verse, depending on what you personally call fuller/bigger, allowing the woofer to resonate.
Most laughable advertisings are about sound of mica layers, color and shape of a getter.
What makes the bottom end sound fuller/bigger, properly 6CA7 that you tried in your amp somehow caused lower output impedance of your particular amp, may be it had higher transconductance than your El34, so it damped resonances of your woofer better. Or may be vice verse, depending on what you personally call fuller/bigger, allowing the woofer to resonate.
Luckily I'm early to the conversation!
WiseOldTech is right — but quantifying his comparative phrasing —
The way an amplifier “sounds” having various pin-compatible-but-still-perhaps-quite-different tubes exchanged is mainly the result of the circuitry and the valves transconductance itself.
For instance, let's say I've (as the designer) penned into my amplifier-of-the-mind a very high gain 'naked triode' as the first FAS stage; a stout highly-linear pair of triodes as the middle gain-plus-inversion stage, a pair of even more stout small-signal pentodes as cathode followers of both phases, driving a 4-plex of breathtakingly inexpensive TV vertical deflection beam pentodes as the finals. Gain is high enough that –10 dB of GNFB is also part of the plan. That's the basic design…
• Laudable in scope, gain, linearity
• Laudable in economy, ease of finding parts
• Laudable in terms of power (probably 50 W clean, out)
• Laudable in underlying design.
This amplifier — let us go on to say — has a S₁ (stage 1) utilizing both halves of a 12AX7 as a cascode loaded conventional triode amp, with capacitor bypassed cathode bias on the 'bottom section'. Running at oh, maybe 2 mA, with a B+ at the top at 350 V, but with the cascode 'upper section' delivering maybe 150 V to the lower triode. The upper cascode load has a 33 kΩ anode load. The anode is at 150 + 75 = 225 V, quiescent. The bottom triode's cathode is at 1.7 V, having a 820 Ω resistor, bypassed by a 47 µF, 16 V electrolytic capacitor.
Now it turns out that you can plug in one HECK of a lot of pin-compatible tubes.
12BZ7, 12AY7, 12AU7, 12AT7, 5751, 12DT7, 12AV7…
But they all have substantially different “sweet spot” operating points. Swap any one of those into the circuit as described, and in this case, most likely, the effect will not be a glaring difference, short of overall gain. The use of cascode topology nearly obviates the Miller plate-grid capacitance effect; the use of a cap-bypassed cathode resistor to set the operating point is pretty resilient to tube rolling. THIS design wouldn't much care (or reward) the experimenter making changes.
However tho' true, its also false.
The use of a cathode resistor isn't really a tight-setter-of-a-particular-quiescent-current (and thus whatever voltage makes that happen!). Rather it is a complex interactor-with-the-valve … allowing more current to flow at a higher (more negative) grid bias, if the value allows it. And doing that could seriously compromise the upper section's cascode topology anode load. At the very least, if the operating current changes significantly from 2 mA, the dynamic range of the front end radically changes.
Anyway, that's just an example.
DEEPER … would be to consider the usually-unavailable-to-the-roller (non-)linearity curves of the various substitution valves, run at near 2 mA. A lot of triodes are designe to prefer 5 mA to 15 mA. It is only the 12AX7 that 'likes' operating between 0.5 and 2.5 mA. Changes things quite a bit. Factor of 10×
Anyway… Just saying,
GoatGuy ✓
WiseOldTech is right — but quantifying his comparative phrasing —
A lot (over 75%) of reviewers use pure nonsense (75% useless words) to make people nuts (to mostly agree with the conventions-of-the-moment known as Smart-In-Group Status Quo)
The way an amplifier “sounds” having various pin-compatible-but-still-perhaps-quite-different tubes exchanged is mainly the result of the circuitry and the valves transconductance itself.
For instance, let's say I've (as the designer) penned into my amplifier-of-the-mind a very high gain 'naked triode' as the first FAS stage; a stout highly-linear pair of triodes as the middle gain-plus-inversion stage, a pair of even more stout small-signal pentodes as cathode followers of both phases, driving a 4-plex of breathtakingly inexpensive TV vertical deflection beam pentodes as the finals. Gain is high enough that –10 dB of GNFB is also part of the plan. That's the basic design…
• Laudable in scope, gain, linearity
• Laudable in economy, ease of finding parts
• Laudable in terms of power (probably 50 W clean, out)
• Laudable in underlying design.
This amplifier — let us go on to say — has a S₁ (stage 1) utilizing both halves of a 12AX7 as a cascode loaded conventional triode amp, with capacitor bypassed cathode bias on the 'bottom section'. Running at oh, maybe 2 mA, with a B+ at the top at 350 V, but with the cascode 'upper section' delivering maybe 150 V to the lower triode. The upper cascode load has a 33 kΩ anode load. The anode is at 150 + 75 = 225 V, quiescent. The bottom triode's cathode is at 1.7 V, having a 820 Ω resistor, bypassed by a 47 µF, 16 V electrolytic capacitor.
Now it turns out that you can plug in one HECK of a lot of pin-compatible tubes.
12BZ7, 12AY7, 12AU7, 12AT7, 5751, 12DT7, 12AV7…
But they all have substantially different “sweet spot” operating points. Swap any one of those into the circuit as described, and in this case, most likely, the effect will not be a glaring difference, short of overall gain. The use of cascode topology nearly obviates the Miller plate-grid capacitance effect; the use of a cap-bypassed cathode resistor to set the operating point is pretty resilient to tube rolling. THIS design wouldn't much care (or reward) the experimenter making changes.
However tho' true, its also false.
The use of a cathode resistor isn't really a tight-setter-of-a-particular-quiescent-current (and thus whatever voltage makes that happen!). Rather it is a complex interactor-with-the-valve … allowing more current to flow at a higher (more negative) grid bias, if the value allows it. And doing that could seriously compromise the upper section's cascode topology anode load. At the very least, if the operating current changes significantly from 2 mA, the dynamic range of the front end radically changes.
Anyway, that's just an example.
DEEPER … would be to consider the usually-unavailable-to-the-roller (non-)linearity curves of the various substitution valves, run at near 2 mA. A lot of triodes are designe to prefer 5 mA to 15 mA. It is only the 12AX7 that 'likes' operating between 0.5 and 2.5 mA. Changes things quite a bit. Factor of 10×
Anyway… Just saying,
GoatGuy ✓
Writing a “PS” to my own comment…
The above observation that changing (“rolling”) the S1 valve for another one might well significantly change the operating BIAS point, makes a very strong argument for employing robust CCS regulator devices — instead of C-bypassed resistors — given their simplistic willingness to adjust operating current with bias voltage … definitely not necessarily in a direction that the original designer might have stipulated.
So… CCS 'diodes', like the 1N5299 … or any in the whole family, are definitely decent candidates for the job. One can also quite successfully deploy the LM337 device, or if we look around, another three or four dozen others in various novel (and surprisingly easy to actually wire up) topologies.
What they (CCSs) would do to our lovely circuit would be to have the S1 cascode-loading triode anode resistor also operating as planned at about ¾ of applied B+. Now, that would be sweet. And SAFE. At least the whole front end dynamic range could be preserved.
So…
The old wise designer could work similar magic with the second section … the differential (LTP ← Long Tailed Pair) gain-and-phase-inverter section. Same idea. CCS in the common cathode. But no capacitor bypass!
Per discussion with fellow …BMX… poster, there's a great opportunity to deploy a smallish powdered iron core thousand turn choke in series with the constant-current regulator. Totally puts the quabash on any noise it emits.
In turn, one could roll any from the whole list of the above … into the socket. And without having to worry about whether the upper LTP load-and-VAS-gain anode resistors are not at their design-sweet-spot. Cool.
So…
Following these with yet another dual-triode (or as is my preference, a pair of small-signal pentodes) as simple cathode followers to sturdy up the now-amplified-and-carefully-phase-inverted signal … doesn't need CCS's to do the job right. Its not like you're going to find many pentodes to roll in. But, say that you do … they're nearly impossible to 'fake out' from an operating point, in the cathode follower circuit. Nothing to fix, nothing to tweak.
Lastly though, you got the output sections.
You know… just reading the above, I am lead to consider the utility of cobbling together rather-higher-power CCS cathode-bias inline circuits, which set the nominal quiescent operating current of each output beam pentode. Let 'em float from each other! Just get their nominal … say 40 mA … set independently.
Then “valve rolling” with other beam pentodes on the output becomes a game of “whether-or-not-we-blow-them-up by red-plating-or-worse”. At least they'll all be driven at the nominal zero-signal current.
Anyway. There you are.
2¢ worth.
GoatGuy ✓
The above observation that changing (“rolling”) the S1 valve for another one might well significantly change the operating BIAS point, makes a very strong argument for employing robust CCS regulator devices — instead of C-bypassed resistors — given their simplistic willingness to adjust operating current with bias voltage … definitely not necessarily in a direction that the original designer might have stipulated.
So… CCS 'diodes', like the 1N5299 … or any in the whole family, are definitely decent candidates for the job. One can also quite successfully deploy the LM337 device, or if we look around, another three or four dozen others in various novel (and surprisingly easy to actually wire up) topologies.
What they (CCSs) would do to our lovely circuit would be to have the S1 cascode-loading triode anode resistor also operating as planned at about ¾ of applied B+. Now, that would be sweet. And SAFE. At least the whole front end dynamic range could be preserved.
So…
The old wise designer could work similar magic with the second section … the differential (LTP ← Long Tailed Pair) gain-and-phase-inverter section. Same idea. CCS in the common cathode. But no capacitor bypass!
Per discussion with fellow …BMX… poster, there's a great opportunity to deploy a smallish powdered iron core thousand turn choke in series with the constant-current regulator. Totally puts the quabash on any noise it emits.
In turn, one could roll any from the whole list of the above … into the socket. And without having to worry about whether the upper LTP load-and-VAS-gain anode resistors are not at their design-sweet-spot. Cool.
So…
Following these with yet another dual-triode (or as is my preference, a pair of small-signal pentodes) as simple cathode followers to sturdy up the now-amplified-and-carefully-phase-inverted signal … doesn't need CCS's to do the job right. Its not like you're going to find many pentodes to roll in. But, say that you do … they're nearly impossible to 'fake out' from an operating point, in the cathode follower circuit. Nothing to fix, nothing to tweak.
Lastly though, you got the output sections.
You know… just reading the above, I am lead to consider the utility of cobbling together rather-higher-power CCS cathode-bias inline circuits, which set the nominal quiescent operating current of each output beam pentode. Let 'em float from each other! Just get their nominal … say 40 mA … set independently.
Then “valve rolling” with other beam pentodes on the output becomes a game of “whether-or-not-we-blow-them-up by red-plating-or-worse”. At least they'll all be driven at the nominal zero-signal current.
Anyway. There you are.
2¢ worth.
GoatGuy ✓
It is called synesthesia when predicates from one modality are used to attribute something to different sensors. Like sharp sound, meaty music, and so on. There is no agreement among audiophiles about them. And of course tubes do not bear such properties, especially when related to different frequencies. Tubes do not care what frequencies how to distort, until wavelength is getting closer to sizes of electrodes, way-way beyond audio band.
Most laughable advertisings are about sound of mica layers, color and shape of a getter.
What makes the bottom end sound fuller/bigger, properly 6CA7 that you tried in your amp somehow caused lower output impedance of your particular amp, may be it had higher transconductance than your El34, so it damped resonances of your woofer better. Or may be vice verse, depending on what you personally call fuller/bigger, allowing the woofer to resonate.
As we usually agree … I find myself … ahem … agreeing. Wholeheartedly. I talk too much, and you hit it right on the head. Thanks! GoatGuy ✓
bright, dull, rolled, extended, strong bass, is just relative to the system the tubes are into...
What someone is describing as 'sounds smooth and extended' is going to be described as narrow and irritating for another.
As for 'tube characters etc' well, listen to as many amps as possible in the same system, you will quickly realize that each tube sounds different, including, nos, and same brands, different pairs.
It is logical, they have different distortion, bandwidth, current limits, and behave all in a certain way with what is called 'music'.
As for comparing different brands of tubes, of course some tubes are better, which ones? it is hard to tell without a very detail THD analysis.
Don't believe the comparisons, read the reviews but choose for yourself...
I personally don't like at all electroharmonix, why? I don't like how they sound period. I don't like chinese tube or russian vintage tube for the most part, I find them mellow, veiled. I crave for N.O.S. tubes from haltron, GE (black and gray plate, very complex) etc, but it could be just me. JAN, Japan gold pin tubes NOS, and many others.
To me each tube species is very different, but to be honest, just buy a good tube, bargains do not exist, like in wine, price means something.
(ps. I concur with Wavebourn saying lot of merchants up the prices with triple mica, black plate , black mesh plate etc, but they are not = to quality, I found some gray plate as good as black plate, black plate is not a synonym of quality, gold pins is not going to sound better, you have to consult lists of reviews of nos tubes, then buy a few, then measure them, then confirm by listening, just do not get trapped into siemens gold pins, there are other choices.)
A good example is the 12 series, vs the 6.3V heaters, you can get extraordinary good tubes for a decent price. An NOS doesnt mean always more $, look how much a good 6sl7 from tung sol cost! it is 45$ for a low noise matched...
What someone is describing as 'sounds smooth and extended' is going to be described as narrow and irritating for another.
As for 'tube characters etc' well, listen to as many amps as possible in the same system, you will quickly realize that each tube sounds different, including, nos, and same brands, different pairs.
It is logical, they have different distortion, bandwidth, current limits, and behave all in a certain way with what is called 'music'.
As for comparing different brands of tubes, of course some tubes are better, which ones? it is hard to tell without a very detail THD analysis.
Don't believe the comparisons, read the reviews but choose for yourself...
I personally don't like at all electroharmonix, why? I don't like how they sound period. I don't like chinese tube or russian vintage tube for the most part, I find them mellow, veiled. I crave for N.O.S. tubes from haltron, GE (black and gray plate, very complex
) etc, but it could be just me. JAN, Japan gold pin tubes NOS, and many others. To me each tube species is very different, but to be honest, just buy a good tube, bargains do not exist, like in wine, price means something.
(ps. I concur with Wavebourn saying lot of merchants up the prices with triple mica, black plate , black mesh plate etc, but they are not = to quality, I found some gray plate as good as black plate, black plate is not a synonym of quality, gold pins is not going to sound better, you have to consult lists of reviews of nos tubes, then buy a few, then measure them, then confirm by listening, just do not get trapped into siemens gold pins, there are other choices.)
A good example is the 12 series, vs the 6.3V heaters, you can get extraordinary good tubes for a decent price. An NOS doesnt mean always more $, look how much a good 6sl7 from tung sol cost! it is 45$ for a low noise matched...
Last edited:
Moved to the Lounge.Perhaps I should elaborate. There’s a definite difference between the pair of Ehx 6922 that I have and the JJ 6922 (Ecc88) pair that I own. I can tell when one pair is in the amp vs the other. So can my wife, without looking. I also tried the two in a different amp that I own with the same result.
Is this just a an example of ever-so-slightly different electrical properties between these two pairs—one that might change in the next pair that I order? You see, I’ve gone through multiple quads of Svetlana EL34s over the years and I tell you they all sound the same to me. As soon as I put in a quad of JJs, I could tell an immediate difference. I won’t get into which one I liked better, etc.
By the way, this isn’t some backdoor into a discussion of the merits of NOS tubes, etc. I decided long ago to go with new production only; not just for money savings, but for consistency purposes and availability also. Save for the OD3 voltage regulators, everything else is new production: JJ 5ar4, JJ ECC88, EHX 6ca7.
Edit: sorry, I didn’t know that this should be in the Lounge! Sorry mods!
Is this just a an example of ever-so-slightly different electrical properties between these two pairs—one that might change in the next pair that I order? You see, I’ve gone through multiple quads of Svetlana EL34s over the years and I tell you they all sound the same to me. As soon as I put in a quad of JJs, I could tell an immediate difference. I won’t get into which one I liked better, etc.
By the way, this isn’t some backdoor into a discussion of the merits of NOS tubes, etc. I decided long ago to go with new production only; not just for money savings, but for consistency purposes and availability also. Save for the OD3 voltage regulators, everything else is new production: JJ 5ar4, JJ ECC88, EHX 6ca7.
Edit: sorry, I didn’t know that this should be in the Lounge! Sorry mods!
Last edited:
As a general rule, the more you can 'hear' the difference between different valves of the same nominal type the worse the circuit and the worse the valves. Good valves in a good circuit should all sound almost identical. There may be minor differences in gain and noise.
However, in a poor circuit or with poor valves you can get
- different bias points so different distortion
- different HF rolloff due to Miller effect (perhaps combined with too high source impedance or too high grid stoppers)
- different LF rolloff due to anode impedance
Such problems are more likely with modern valves, as they don't stick to the classic datasheet spec as closely as good NOS valves.
However, in a poor circuit or with poor valves you can get
- different bias points so different distortion
- different HF rolloff due to Miller effect (perhaps combined with too high source impedance or too high grid stoppers)
- different LF rolloff due to anode impedance
Such problems are more likely with modern valves, as they don't stick to the classic datasheet spec as closely as good NOS valves.
In a nutshell that is basically what I explained earlier - the tubes depend on their "under-chassis" components for operation. People tend to "tube roll" and play around if they aren't satisfied with a certain type of performance from a tube amp, when the real controlling culprits are under the chassis.
You know, or are you just saying it to pacify the engineering types here?
How can tubes not have sound when everything else does? Perhaps you should rather ask these questions about passive components which for the lack of transconductance are much simpler to examine.
Why does a short piece of silver wire sound different to copper or gold? Apart from handwaving i have never seen any sensible suggestion of why this may be. The engineering types jump straight into denialism as the easiest way out.
Point taken on the sound of tubes.
Perhaps I’m focusing on them since, at least for me, I can’t unplug and plug in a different resistor as easily as I can a tube.
As far as silver wire, it seems lost to me on my setup. Unless I’m mistaken, isn’t there tens—if not hundreds—of feet of copper wire in the voice coil of my Klipsch woofers? While my amp’s circuitry is indeed comprised of silver wire, after it goes to the terminals, everything past them is copper. I will admit, however, that the best speaker wire I’ve tried is single conductor copper wire I got over at Home Depot that’s used for doorbells. I think 50’ cost like $8.00. (My doorbell sounds great by the way.) I was at one point considering silver speaker wire, but thought it foolish to spend the money on something that went straight to a long run of copper voice coils.
Perhaps I’m focusing on them since, at least for me, I can’t unplug and plug in a different resistor as easily as I can a tube.
As far as silver wire, it seems lost to me on my setup. Unless I’m mistaken, isn’t there tens—if not hundreds—of feet of copper wire in the voice coil of my Klipsch woofers? While my amp’s circuitry is indeed comprised of silver wire, after it goes to the terminals, everything past them is copper. I will admit, however, that the best speaker wire I’ve tried is single conductor copper wire I got over at Home Depot that’s used for doorbells. I think 50’ cost like $8.00. (My doorbell sounds great by the way.) I was at one point considering silver speaker wire, but thought it foolish to spend the money on something that went straight to a long run of copper voice coils.
some of us use tube to have 0 feedback designs. I do like almost to 0 feedback designs.
The less feedback, the more you will hear the tube differences.
I would not called the circuit flawed just because the feedback is low... there are clear advantages to 0 feedback circuits.
Do you prefer the JJ... i like JJ over harmonix.
Why keeping the door shut to old stock tubes? some used tube are too super good, tested.
I don't think you will find new stock of ecc88 which sound as good as holland amperex gold pins... as soon as I used it in my phono I knew it could not get better than this. As low THD as the best feedback transistor phonos, with the quality of tube sound.
Same thing with 6sn7, the JAN which are not very costly are as good as other super costly tubes and the new production from the best (except psavane treasures, not tried yet) are far from good as the nos.
There are exceptions, the ecc99 is a new tube developed by JJ and it is very good, a true bargain, if you want to ditch all the nos to the garbage, use ecc99 all they way.
The less feedback, the more you will hear the tube differences.
I would not called the circuit flawed just because the feedback is low... there are clear advantages to 0 feedback circuits.
Do you prefer the JJ... i like JJ over harmonix.
Why keeping the door shut to old stock tubes? some used tube are too super good, tested.
I don't think you will find new stock of ecc88 which sound as good as holland amperex gold pins... as soon as I used it in my phono I knew it could not get better than this. As low THD as the best feedback transistor phonos, with the quality of tube sound.
Same thing with 6sn7, the JAN which are not very costly are as good as other super costly tubes and the new production from the best (except psavane treasures, not tried yet) are far from good as the nos.
There are exceptions, the ecc99 is a new tube developed by JJ and it is very good, a true bargain, if you want to ditch all the nos to the garbage, use ecc99 all they way.
Supposedly my amps have 0 global NFB (Decware Torii Jr. and SE84CKC). I thought I preferred the EHX 6922s over the JJs at first, but found something much more involving about the JJs that made the EHX or Gold Lions sort of clinical. On the other hand, JJ 6922s into JJ 6ca7s or EL34s seemed too veiled, and the EHX 6ca7s brought back some of that transparency.
I’m forgoing the NOS due to a fear of loving some combination and then not being able to replicate it years down the road without spending a lot of money. I fell for the Phillips 5ar4 metal base thing and realized after trying the modern day JJ version, I couldn’t tell the difference. I’m sure it’s just the voltage drop, after having tried brown base versions, Matsushita built RCAs, etc. and seeing the difference when rebiasing.
Which brings me back to my original question: what is it between two different pairs of 6922s that makes such a difference? Transconductance? What governs transconductance? Certainly not the micas, or support rods, getter shape, etc.
I’m forgoing the NOS due to a fear of loving some combination and then not being able to replicate it years down the road without spending a lot of money. I fell for the Phillips 5ar4 metal base thing and realized after trying the modern day JJ version, I couldn’t tell the difference. I’m sure it’s just the voltage drop, after having tried brown base versions, Matsushita built RCAs, etc. and seeing the difference when rebiasing.
Which brings me back to my original question: what is it between two different pairs of 6922s that makes such a difference? Transconductance? What governs transconductance? Certainly not the micas, or support rods, getter shape, etc.
Could be related to the solder joints. It might be interesting to compare
soldered terminations with clamped ones.
samsdad, thanks for the reply, you explained well, providing the amp type,
the differences are way beyond my understanding.
i will send you a link to how tubes are made. you might find explanations but not enough to choose the good tubes by visual inspection i may fear
when i return from swimming in the lake (I do this everyday, it helps my audio system to sound good)
the differences are way beyond my understanding.
i will send you a link to how tubes are made. you might find explanations but not enough to choose the good tubes by visual inspection i may fear
when i return from swimming in the lake (I do this everyday, it helps my audio system to sound good)
Last edited:
differences between tubes and even triode sections inside of the envelope are what made tube rolling work....
i doubt circuit designers wanted that....
different tube mechanical builds, cathode material constructions. made sure rolling worked...
a look at tube data sheets and plate curves, you will see how tube characteristics vary with plate voltage and cathode currents and grid voltages...and the curves are not linear nor a straight line, they are bent all over the place...
i want my tube amp builds sound great at get go, hitting the sweet spot and not to be helped along by tube rolling....
tube rolling to me is depressing thing....
i doubt circuit designers wanted that....
different tube mechanical builds, cathode material constructions. made sure rolling worked...
a look at tube data sheets and plate curves, you will see how tube characteristics vary with plate voltage and cathode currents and grid voltages...and the curves are not linear nor a straight line, they are bent all over the place...
i want my tube amp builds sound great at get go, hitting the sweet spot and not to be helped along by tube rolling....
tube rolling to me is depressing thing....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.