Why do NOS valves sound "better" than new production ones?

[worrasf]

There is no doubt in my mind that NOS valves (provided they test "good") sound "better" than new production ones. I put better in parentheses because I seem to remember reading an opinion somewhere that it wasn't actually better just that us vinyl types preferred the "distortion" :scratch:

Anyway, be that as it may, to my ears all the NOS valves I have used in my Crofts produce music that is more enjoyable than the new ones I have tried (JJ, Sovtek, Electro Harmonix for example) but I can't figure out why this should be so. I am presuming that the basic electronics/internals are the same in both types.

Thoughts anyone :hmm:

Steve

[DSJR]

Modern materials are not as good I believe, either for safety reasons or just lack of availability.

Having said that, I'm still stunned how Glenn can get such an incredible performance from his current models with modern "ordinary" valves and I suspect many of these actually "are" better, if not as "charming" as old ones.

[Ali Tait]

It may be that the age of the materials used in the construction has some effect on how it sounds. It's often the case that the older a valve is, the better it sounds. I've heard this for myself in valves like PX4's etc. In fact, one of the very best sounds I ever heard was from a pair of 1930's 300a's, a very rare forerunner to the 300b, though electrically identical. There must be a good reason why Japanese audiophiles will pay thousands for a pair!

[Effem]

Maybe it's where the valves are currently being manufactured that has something to do with it. Most of the NOS ones were made in europe and the USA at a time when they were all fiercely competing against each other to be the best?

[camtwister]

There's a difference between internal construction of some of the new production valves and vintage examples of the same type.
I suggest that the larger manufacturers of new valves of common classes, such as dual triodes, sell most of their products to the musical instrument industry -
particularly the guitar amplifier market. I'd be surprised if tube guitar amps don't significantly outsell esoteric valve amplifiers for eccentric audiophiles.
Anthony TD will know vastly more about this than I, but I'll stick my neck out and state that a power tube in, say, a Marshall amp is chosen for its lack of microphonics
and especially the characteristics of its harmonic distortion when overdriven. These are different qualities to those required in a hi-fi power tube. That is, linear frequency
response, extended frequency response and greater headroom before distortion. During the hey-day of the vacuum tube, radio and TV manufacturers would have been
more concerned with the characteristics required for fidelity and resolution, rather than the qualities required in a good guitar amp.
One approach to reducing microphonics is to have short anode plates, with rigid support. The contemporary JJ Tesla ecc83, for example, has an anode length of about
9mm. A post 1957 Mullard ecc83 will have a plate length of 15mm, shortened from 17mm in earlier production. Perhaps this kind of detail is a factor.
If you squint closely, you'll also notice that the grid construction, interpin shield and getter arrangement is different between a modern tube and a vintage
equivalent, in many cases.

Regarding tube rolling in Stephen's Crofts, and as DSJR alluded, I suspect he noticed a greater difference when rolling in his Series 4, than his 25R.
Glenn's new pre-amps, of which I have two, are superb out of the box. I have experimented with changing the line stage valve, and have happily tuned my system to my
tastes. However, it is just that - "tuning" - a subtle change of emphasis and nothing like the comparatively large transformation of rolling my PrimaLuna (pre and power differences considered).

[worrasf]

Regarding tube rolling in Stephen's Crofts, and as DSJR alluded, I suspect he noticed a greater difference when rolling in his Series 4, than his 25R.


That is true - I was very surprised how much difference changing any of the valves (except the 2 EL519's [TFK] and the 3 voltage regulator valves) in the S4s made to the sound - even the 2 12BH7's (finally settled on a pair of NOS 1960's RCA 12BH7A grey plates - also have a pair of the RCA black plates but the grey's had more detail. Also have a quad of GE 6550A's and a quad philips miniwatt 12AX7's.

Having said that - there was no doubt that swapping out the stock Sovtek ecc83's and ecc81 in the 25R also made a big difference. Now running NOS 1960's RCA 12AX7A's (which did not suit the S4s at all :scratch:) and a NOS GEC A2900 in the line stage which is truly superb :)

Steve

[camtwister]

Cor! An A2900 in the line stage. That's sexy. I can figure how the lower gain compared to the 12AX7 is beneficial, too. Doesn't it draw a higher plate current, though? I presume Glenn has said it's within specification?

[worrasf]

Cor! An A2900 in the line stage. That's sexy. I can figure how the lower gain compared to the 12AX7 is beneficial, too. Doesn't it draw a higher plate current, though? I presume Glenn has said it's within specification?
To be honest I didn't check with Glenn before putting in the A2900 :whistle:
When I ordered the 25R I sent Glenn some valves I was currently using which were a pair of original TESLA E83CC's for the phono stage and a very nice 1959 Ulm TFK ECC801S all of which met with his approval but after a while I found them a tad "harsh" and clinical. I ran a Brimar 6060 Yellow T in the line stage for a few months which I understand has similar characteristics to the A2900 with no problems so hopefully it should be OK :eyebrows:

Interestingly, the A2900 has really quietened down the background hum from the 25R so I guess the previous valves were a bit microphonic.

Regards
Steve

[camtwister]

I ran a Brimar 6060 Yellow T in the line stage for a few months which I understand has similar characteristics to the A2900 with no problems so hopefully it should be OK

Yep, the yellow triangle 6060 and the A2900 are both ecc81...and both very desirable! Good on ya.

[sq225917]

It all has to count doesn't it. The from exact size of parts, alloy specifications of metals used for grid, cathode, anode, the exact mix of the cement in the valve bass, the plating on parts and pins and not forgetting the shape of the glass envelope and the grade of glass used. A valve is one of the few true electromechanical devices, up there with drivers and cartridges.

Everything counts.

in small amounts.

[Ammonite Audio]

I have to concur with the opening statement of this thread, having tried all sorts of 12AU7 and 12AX7 in my Paul Hynes phono amp. New ones have included 'cooking' Shuguang, Shuguang Psvane, Sovtek, EH. I should have saved my money and gone straight for the NOS diamond-base Telefunkens that are hideously expensive nowadays, but are in quite a different league in every musical sense that I can think of. Ultimately, I believe that in the Old Days, manufacturers of tubes just cared more about what they were doing; also perhaps there was just a greater depth of intuition and understanding about what made a better tube?

[sq225917]

I think it might be simpler than that, the Chinese just use any old crap, crap wire, poor glass, the cheapest arts they can find.

[Cliff]

Just put an Amperex Bugleboy into my EAR 834p today, in place of the stock JJ something or other, and spent the rest of the day pulling out record after record. Amazing upgrade.
Just further evidence of the superiority of NOS valves.
Got to replace the other two in there now!:)

[Marco]

As folks have already said, it's all to do with the material and construction quality of NOS valves, as well as the actual manufacturing of the products. I believe that some of the old processes involving poisonous gasses, critical to obtaining optimal performance from vacuum tubes, are now illegal.

There's also the fact that unfortunately we now live in a 'disposable society', where things aren't built to last as they once were, so from that perspective, quality will also suffer, and also performance.

Furthermore, vacuum tubes were originally built for a more discerning and quality conscious customer base, and so certain standards that were adhered to once, no longer apply to the same extent today, and unfortunately that often applies to hi-fi equipment itself, as well as things like valves ;)

Sadly, it’s yet another example of the lowering of standards that calls into question just how much REAL progress has been made in the development of audio equipment in the last 30-odd years, or more.............

Marco.

[AlexM]

I have used numerous 6SN7, 12AX7 and KT88 types, including a range of NOS and current production examples.

I think that modern production power valves can be very good - I love the sound of the Genelex KT88 reissues in my Rogue Stereo 90, they are very closely matched, bias very consistently and seem to be long-lived. I think that New Sensor have invested heavily in QC and test gear, and quality seems to be consistently good. I don't have the same confidence in Chinese product.

Small signal valves seem to vary quite a bit in sound and I have yet to find an acceptable sounding new 12AX7 type, whereas the Mullard M8137s I'm using at the moment are just magic. With the 6SN7s, there are some quite good modern examples that work well in some positions, or in combination with NOS valves. I actually like the EH 6SN7 Gold very much, both as a preamp valve and a driver.

I prefer them to the NOS RCAs, Sylvanias and VT-231 Ken-rads as a preamp valve. For me the best sound is a combination of EH 6SN7 Golds in the preamp, and Ken-Rad VT231s in the power amp driver stage. The sound is very well balanced and I think that system matching is all important here - there are no absolutes.

I think that some manufacturers are taking the mickey with the pricing of their current production (Step forward Shuguang). You could get a couple of sets of GE 6550s for the price of their black treasure KT88s.. I know which I would prefer.

Regards,
Alex

[camtwister]

I think that modern production power valves can be very good - I love the sound of the Genelex KT88 reissues in my Rogue Stereo 90, they are very closely matched, bias very consistently and seem to be long-lived. I think that New Sensor have invested heavily in QC and test gear, and quality seems to be consistently good.



I concur. The New Sensor re-issues are my favourite by a contemporary manufacturer. I have a quad of KT-77s that were matched for me by Watford Valves and they have been a very enjoyable investment. I can't compare them to NOS GEC because my pockets ain't deep enough but I do recommend the Gold Lions. One benefit of current production is that you can get spares and know that the sound and build quality is going to be reasonably consistent.

[Jimbo]

I have just changed a JJ ECC99 in my Croft Pre 25R linestage to a NOS GE 12BH7A and the difference is astounding. Not sure if NOS valves make such a difference in the regulation stage but certainly do in the linestage and phono stage.

Some of the reasons which make NOS valves 'better' sounding than modern equivalents are certainly concerning the materials they were made of and the quality of construction. The Cathode coating material in particular was almost a black art and was made to a formula know only by a few, probably out of materials that would be considered dangerous to health now!

What I have found fundamentaly is that no modern equivalent valves sound as good as NOS and they are therefore certainly a worthwhile investment in improving SQ.

[The Black Adder]

To make valves like they used to would simply not be a viable in regarding sales. People had big hopes for the Mullard resurrection some time ago but it was the name they were after and not the same production values/methods/standards.

[Barry]

As folks have already said, it's all to do with the material and construction quality of NOS valves, as well as the actual manufacturing of the products. I believe that some of the old processes involving poisonous gasses, critical to obtaining optimal performance from vacuum tubes, are now illegal.

There's also the fact that unfortunately we now live in a 'disposable society', where things aren't built to last as they once were, so from that perspective, quality will also suffer, and also performance.

Furthermore, vacuum tubes were originally built for a more discerning and quality conscious customer base, and so certain standards that were adhered to once, no longer apply to the same extent today, and unfortunately that often applies to hi-fi equipment itself, as well as things like valves ;)

Sadly, it’s yet another example of the lowering of standards that calls into question just how much REAL progress has been made in the development of audio equipment in the last 30-odd years, or more.............

Marco.

Good post!

It is always better to chase after NOS samples - they just knew how to make good valves in those days, and they had good quality control!

These days the art has not been lost, but is largely confined to Eastern European manufacturers. Avoid like the plague anything coming out of China. Distributors may screen samples for performance, but the performance won't last - Chinese manufacture is as sloppy as it gets.

Unfortunately marques like Mullard, Philips (largely the same as Mullard), Brimar, Siemens, and Telefunken are well sought after. There are a lot of "cowboys" out there who have pushed up the price (as they have the likes of LS3/5A speakers and British valve electronics on eBay), as well as promoting the Chinese forgeries of Siemens and of Telefunken designs (yes, it not just Rolex watches and Chanel perfume that is counterfeited).

Caveat emptor !

[SPS]

I would also say that many prewar valves sound so much better than post war nos valves, from the 20's to the war were the best years, but the quality went down hill over time.

the telefunken ecc valves are quite something considering their post war manufacture.

Kt66 are a good example, post war straight sided gec still sound very good except when compared to pre war /wartime coke bottle type, vt75 and k66, but the earlier lower powered types (not kt66) have a better sound still.

The trouble with any audio component is that its performance will affected by the rest of the system Hence the differing views on makes of valves

[Jimbo]

Thanks for that Steve, Interesting stuff, I didn't realise valves went back that far!
Trouble nowadays is finding genuine stuff when there are so many fakes out there!

[SPS]

Thanks for that Steve, Interesting stuff, I didn't realise valves went back that far!
Trouble nowadays is finding genuine stuff when there are so many fakes out there!

When buying valves there is always an element of risk. If you know what your intending to buy should look like, ie the fine detail and the price reflects the description and if tested, and you are aware you will make mistakes but you can live withthem, then go for it.

[Balle Clorin]

No wonder different tubes sound different when the also measure very different, at least in my 300B amplifier.Take a look at this plot.

http://www.hifisentralen.no/forumet/attachments/diy-og-utvikling-ha-yttalere-forsterkere-etc/287866d1415397908-driverra-r-eh-6sn7-kan-de-erstattes-av-vt-231-capture.jpg

[awkwardbydesign]

I believe that some of the old processes involving poisonous gasses, critical to obtaining optimal performance from vacuum tubes, are now illegal.


Marco.
In the old days (when I was in the stained glass trade) ruby glass was made with arsenic. Production then moved east, to Poland for example, and then further east. Now I believe it's very hard to get, and selenium red just ain't the same. Valves are probably similar.

[awkwardbydesign]

No wonder different tubes sound different when the also measure very different, at least in my 300B amplifier.Take a look at this plot.

http://www.hifisentralen.no/forumet/attachments/diy-og-utvikling-ha-yttalere-forsterkere-etc/287866d1415397908-driverra-r-eh-6sn7-kan-de-erstattes-av-vt-231-capture.jpg

Well, what ARE the blue ones, then?

[Balle Clorin]

The lower blue curves are ordinary Tung-Sol Reissue, 6SN7 tested and Amplitrex matched . The worst tube is a Shuguang 6sn7 (squares), the second worst is a NOS Raytheon USA 6sn7. YMMV..

The black thick line is Stereophiles measurement of my amp with stock tubes.

[User211]

Well I guess I have been rolling valves for 25+ years. There are 30 of the things in my current system.

NOS isn't always better. But mostly it is. That said there are plenty of NOS valves I haven't liked. It is no guarantee.

The guitar amp argument and new valves being optimised for them? Not so sure. NOS valves sound best in my geetar amp.

Seeing 12BH7A is mentioned I found RCA black plates to be better than the greys.

Using a decent mic I found I could measure system frequency response changes easily between valve rolls. It IS real there really are measurable real world changes going on.

[Balle Clorin]

I did my measurement with a PC and REW program, used a dummy 8 ohm load, and a voltage divider made of high omh resistors+ potmeter in paralell, and connected to my PC mic input.Easily done, and more accurate than measuring in room with a mic.

[User211]

I found up to 3DB difference in varying parts of the frequency range rolling 211s using a Behringer test mic and HolmImpulse.

Varying THD, FR and other parameters all add up to produce the perceived auditory variation. Phase differences must be fairly variable to produce differences in perceived soundstage size maybe? Just guessing.

Plus I'm a variable too. I change my mind/preferences over time.

[Marco]

In the old days (when I was in the stained glass trade) ruby glass was made with arsenic. Production then moved east, to Poland for example, and then further east. Now I believe it's very hard to get, and selenium red just ain't the same. Valves are probably similar.

Indeed, Richard :)

You may also be able to explain why modern stained glass windows simply don’t look the same as their antique counterparts. The colours in the former are definitely brighter and more ‘acid’ in appearance, and they also don’t have the same depth of colour, looking distinctly watered down and more ‘anaemic’, in comparison.

Is this because the pigments used have changed, or what else in your opinion? I ask because I have an interest in antiques and collectables, and live in an old property, where we’d like to introduce the use of antique stained glass windows in certain areas of the house.

Marco.

[AlexM]

That's an interesting set of measurements, and quite a variation in THD, just from a change of driver. It suggests to me that perhaps amplifier designers do tend to design around what is available in production, and that performance is optimised for these valves. The tung - sol 6SN7 I haven't tried, but I presume that it is closely related to the other new sensor valves made in the reflector factory branded as EH and others. I know that the EH 6SN7 gold pin was a high quality tube, and gave plenty of gain with low noise and microphony. It was very consistent too. I
think that the biggest problem with NOS valves is that it is a crap shoot as to what you will get as a consumer buying in the market. My own experience with 6SN7s is that they do sound quite different and 'best' is not a readily comparable between different circuits. I preferred the EH 6SN7 gold pin in my amps to alternatives from Kenrad, RCA, Sylvania and Tung sol NOS.

I think that It is hard to know what the likely impact of varying THD is on sound quality. Generally low distortion ought be preferable, but perhaps we like a bit of added distortion as long as it is even harmonic. I'd be interested in learning a bit more about your test method - can you describe it in a bit of detail?. This isn't a subjectivist forum, but I for one found your post very illuminating.

Cheers,
Alex

[awkwardbydesign]

Indeed, Richard :)

You may also be able to explain why modern stained glass windows simply don’t look the same as their antique counterparts. The colours in the former are definitely brighter and more ‘acid’ in appearance, and they also don’t have the same depth of colour, looking distinctly watered down and more ‘anaemic’, in comparison.

Is this because the pigments used have changed, or what else in your opinion? I ask because I have an interest in antiques and collectables, and live in an old property, where we’d like to introduce the use of antique stained glass windows in certain areas of the house.

Marco.

I've started a thread in Abstract Chat to talk about this, rather than drift this thread too far.

[Balle Clorin]

That's an interesting set of measurements, and quite a variation in THD, just from a change of driver. It suggests to me that perhaps amplifier designers do tend to design around what is available in production, and that performance is optimised for these valves. The tung - sol 6SN7 I haven't tried, but I presume that it is closely related to the other new sensor valves made in the reflector factory branded as EH and others. I know that the EH 6SN7 gold pin was a high quality tube, and gave plenty of gain with low noise and microphony. It was very consistent too. I
think that the biggest problem with NOS valves is that it is a crap shoot as to what you will get as a consumer buying in the market. My own experience with 6SN7s is that they do sound quite different and 'best' is not a readily comparable between different circuits. I preferred the EH 6SN7 gold pin in my amps to alternatives from Kenrad, RCA, Sylvania and Tung sol NOS.

I think that It is hard to know what the likely impact of varying THD is on sound quality. Generally low distortion ought be preferable, but perhaps we like a bit of added distortion as long as it is even harmonic. I'd be interested in learning a bit more about your test method - can you describe it in a bit of detail?. This isn't a subjectivist forum, but I for one found your post very illuminating.

Cheers,
Alex
I use this setup.

1 .Google REW from HomeTheatershack, and install it
2. Hook up according to figure
3. Put on a CD wth a 1000 Hz test tone.(Stereophile has some, or burn a file from the net) and play on your amp
4. Make sure the PC Mic input level is at very low to avoid overloading the input that will distort, Adjust potemeter to get sa usable signal, a low vol setting.
(while measuring adjust PC rec level until just below the point where distorion increases, that is you max level to PC, you may use the CD output for a udistorted signal level check)
5. Start A REW measurement, selct RTA buttoon and click THD button so see % THD
6. Record multimeter VAC , low to high,(I add a oscilloscope to look for clipping)
7. Plot the results in Excel by convering VAC to Watt; VAC*VAC/8ohm=watt
8. I usually go from 8 to 0 VAC by adusting the amplifier volume, leaving other setting constant

NOTE: I only test one channel with a dummy load, the other has speaker connected ,but NO input. A use a dummy load made of 10 82ohm 10 watt resistors tobe safe. Make sure youre amplifer does not overlod the dummy load in watts,

It is kind of hard to notice distorion below 0,5-1%, but 2-5% is easy to hear, at least in pure tones. Harmonic distrion brings a fullness to the sound, just like a chord sounds nicer than a single note. That is why thing may sound nice and measure quite bad on THD, 0.1-0.2% distortion does not matter at all to my ears

http://www.hifisentralen.no/forumet/attachments/diy-og-utvikling-ha-yttalere-forsterkere-etc/292666d1417472559-driverra-r-eh-6sn7-kan-de-erstattes-av-vt-231-rew.jpg