Vacuum Tubes Today and NOS

Vacuum tubes continue to be used today in high end audiophile equipment in the recording industry, home and in music instrument amplification.  Guitar amplification utilizes the highest percentage of current vacuum tube production.

I have been publishing test report results on the subject of vacuum tubes since 2002.  The news is not good news.  The trend is not a good trend.

If you click on the above image it will enlarge to a more readable size.  I will refer to this image later in this piece.

For more just click on "read more" below

Dan Boul is the President and founder of 65 Amps.  They build high end guitar and bass amplifiers.  I cannot recall the number of times I have heard Dan remark that if somebody made a good 12AX7 for $50 he would buy as many as he could.

There are many types of vacuum tubes.  The quality of all of them is in a sad state to say the least.  In this piece I will focus on the most common of the tubes used today, the 12AX7 for the majority of this piece.  I will also tend to focus on the use of tubes in guitar and bass amplifiers as well as studio recording equipment.

Many of today's amplifier manufacturers base their designs around the substandard performance and reliability of today's production tubes.  Unlike solid state components, vacuum tubes are hand assembled and profit is the driving force behind their manufacture.  Quality control is limited at best and the quality of the materials is held to a standard where cost is the primary consideration not performance or reliability.

I will try to keep this piece from becoming too technical.  Looking at the chart at the beginning of this piece I will simply state that there are many aspects to the performance of a vacuum tube and all of these factors must be held within a balance.  Today's tubes do not maintain this balance very well to say the least.  Plate resistance is typically low.  This allows the two sides of the tube to crosstalk or interact with each other when that is not the desired trait.  Low plate resistance will make the gain of a tube low and the plate current will be too high.  High plate current causes heat and there are other bad side effects.  Bottom line is that today's tubes stray from the first chart at the top of this piece.

In the chart above you can see what a proper constant current curve should resemble.  In a perfect tube it would be smooth and linear.  In the case of a guitar player who has talent of an articulate touch smooth curves would allow them to play with a palette of many colors and shades of colors to use an artist analogy.  Today's tubes are not very smooth or linear.  The artist paints in little more than the three primary colors in some cases.  

Many of today's short plate tubes compress quickly.  They do not have a wide dynamic range.

Some things to consider when we talk about NOS (new old stock) tubes.

1.  Amp makers need a steady reliable supply of vacuum tubes.  NOS tubes are limited in supply and costly.

2.  It is easier and cheaper to build an amp that will perform adequately from a business sense if designed around current substandard tubes.  In many cases if high quality tubes are placed in these amps the short comings of the amp design and cheaper component utilization in these amps will be uncovered.

3.  It is much more costly to design and build an amp that performs very well on current production tubes (as long as the tubes meet standards based on a somewhat rational scale) but will increase in performance when high quality tubes are utilized.  To make this happen one needs to use components such as transformers and capacitors which are much more costly than the commonly used components in many amplifiers.

4.  NOS tubes will sound better, have more linear curves and much longer life as a general rule.  I have pulled tubes out of amplifiers from the 1950s that still test better than 90% of current production tubes.

5.  When comparing two different new production tubes, or comparing any two tubes, compare apples to apples.  If you do not have the test data you cannot compare them as if one tube is stronger than the other it will generally be favored by most human ears.  A weak Telefunken may not fare well against a current production tube that is not as weak.

6.  Just because something is advertised as NOS it does not mean it is a good tube.  There is a lot of junk out there.  The bottom of the barrel is within our sight.  There are pulls and there are pure counterfeit tubes out there with fake boxes and fake silkscreen.  Bottom line here is trust your tube vendor.

7.  Proper NOS tubes = tone, reliability, consistency and value.  If a $75 12AX7 will last decades and sound great isn't that a better value than a $12.95 tube that sounds mediocre and fails or becomes microphonic after a few months or even a year?

8.  Beware of specs from today's manufacturers.  In most cases they do little more than reprint spec sheets from the past even though they do not meet the data.

9. Many new tubes will test as wonderful on typical transconductance testers such as Hickock and the like.  The plate resistance low so all bets are off as some factors that are measured look great but other factors which are not show are terrible.  If you have a tester such as an RCA WT-100 or something that measures plate resistance and plate current these tubes will show their true colors.

Tubes today will not trace like this chart above.  

Typical trace of an NOS tube looking at both sides of the tube at the same time.

The two traces above are the best ones I have on record for a production tube that was good when it came out but degraded during production.  

Explaining curves would be very technical and difficult in this piece.  Perhaps I will do that in another piece in the future.  If there are folks that would like to see this in real time and talk about it talk to Dan Boul over at www.65amps.com and ask him if he would put on a Saturday seminar sometime.  I would be happy to talk about these things.

If you want to read a bit more on the technical aspects of tubes feel free to read this

http://www.guitaramplifierblueprinting.com/Tubeprimerandselection320-2.pdf

There are about a dozen pieces of info on my L.A. blog that can be viewed via this link which will display them all.
http://la-economy.blogspot.com/search/label/music

When I published my last two reports, the JJ803S and Tung Sol 12AX7 reissue, I was asked about NOS.  Below are some NOS results.

There are some non-12AX7 tubes in this group.  The number that you would want to look at is in the last column on the right.  Numbers far below 100 are 5751, 12AY7 or 12AU7.  Numbers of around 100 are for 12AX7.  A number like 104/100 indicates that design spec is 100 and the tube actually performed better than design spec at 104 (this is true/actual gain in circuit)

Once again, if you click on the image it should become more readable.  

Today's tubes generally do not fare as well.  

JJ 803S

Tung Sol Reissue

Thanks to Dan Boul and the crew at 65 Amps for putting up with me on a continual basis for almost a decade at this point.

Update:  Today on the weekly 65 Amps webcast Dan and I spoke on the subject of tubes and some of the points in this post.  Below is the recorded broadcast from today, 3/7/12

I am posting a bit of dialogue from the Dr. Z Amp forum where somebody asked me a few questions:


Interesting read.

How far above and below specification would you consider using a  tube? For example a 12AX7 in a tone driver position, and in a phase inverter position.

You mention a true gain reading of 104 as performing better. Is there point on the true gain scale where a higher reading becomes a problem?

Is there a low reading where you would not use a tube? The new JJ true gain reading goes down to about 73 while the new Tung Sol goes down to about 85.

Please advise, thanks.

My reply:  All these factors depend on a particular amp design and circuit.  Some amps tolerate higher gain, some don't like it.  If an amp has high quality parts it loves great tubes.  Cheap parts or bad design and you better stick with low performance tubes that are not very articulate.

I will not use a tube if it is more than +/- 20% out of spec on some measurements but will on other measurements.  This applies to preamp tubes.  In the case of output tubes they must be within 10% of spec or I will never use them.  Forget all that you have heard about early and late distortion or rating numbers.  In the case of a Groove Tubes rating if it is not a 4,5,6 toss it.  Those tubes had construction problems and faults and are prone to early failure.  The reason for the rating spread was, from my own point of view, to allow the sale of tubes rather than sending them to the trash can where they belonged.

In the case of plate resistance which for a 12AX7 is expected to be 62.5k ohms, I will not use a tube unless it is 56k ohms or more.  I do not care about maximum plate resistance.  Telefunkens are typically over 80k ohms and have low plate current and low transconductance yet their true gain is the highest of 12AX7s.  This is one reason they can be physically microphonic (lots of gain) but when you have a good one they are great.

Sometimes I used tubes with very low plate current, 0.6mA or even less in phase inverter positions when I want the PI to break faster/sooner to get earlier breakup in the power section.  Sometimes I go the opposite route at 1.3-1.4mA.  Depends on the application.

There are no hard rules or guidelines other than a desire to have the tube generally do what it was designed to do which unfortunately is not often the case.

Side note on JJ stuff ... they used to be the highest gain production tubes in the last few years.  Now they are about average and at times they were also low.  Every batch is a spin of the wheel.  Wheel of mis-fortune.

Below is one more reason that tubes were better in the past.

Here is the pile of rejects from the 1920's & 30's before the Tube Smashing machine was built. 

This little chart from the folks over at Tubeworld is their point of view on some 12AX7
types.  It might be a very nice guideline for folks.

I don't think it would be fair to omit the work that was done in 2009 on the Blackburn front by TechTube.  If nothing else these folks proved that high quality tubes could be made today.

I have some photos with captions and data at

http://www.flickr.com/photos/myles111us/sets/72157622045828365/with/3868175750/

- the end for the moment -

Below I will copy and paste some of the commentary on facebook in regard to this post and perhaps some of the other vacuum tube posts.

Myles Rose Today only a small percentage of tubes are tested. If they are not flat dead they are shipped. The supplier's warranty is the QA when somebody ends up with garbage IF the end user wants to pack the tube up and ship it back to the tube vendor. That is only one issue. A bigger issue is the question / worry in regard to when the tube will fail. It is not a matter of "if", it is a matter of when. Tubes in the past would fail but it took years and decades to wear them out so it was really not much of a concern or worry. Unfortunately today you have the concern from the day you plug in the new tube.

Dan Boul Can you post this on the tube manufac's FB's? This is like a fantasy . . .

Myles Rose ‎Dan Boul - when I was at the last tube place I worked I would send the reports that I now do over in your shop to the manufacturers for each run. In the case of JJ they would respond very quickly and were usually already aware of the problem. In the case of another company they did not care and at one point even tried to joke with me that they were the only game in town. Another supplier, Ruby Tubes, was always ahead of the curve and quite consistent.

Hopefully consumers will try to push the tube manufacturers harder in the future. Some of them listen (Tom McNeil / Ruby Tubes). He has a strong relationship with the manufacturing side of his operation due to a good personal relationship. One of the folks I was close to was the largest in regard to buying from some manufacturers but had something a poor personal relationship. There were two of the same product from Shuguang where the Ruby was terrific and the other vendors product was not as consistent.

The best solution would be for somebody like Blackburn / Mullard to get some sort of funding again and resurrect their operation. They were on the right track. All it would take is one manufacturer to make a good ECC83/12AX7 that even at $70 a tube, if it sounded nice and lasted more than a few years would raise the bar. What these folks showed to me is that good tubes can be made today.

Myles Rose Here are some photos from the past on some of the Blackburn testing. These tests were done in Dan Boul's shop over at 65 Amps. Part of the testing was comparing these tubes to the original Mullards that were in Dan's 50 watt Marshall. I was impressed but Dan is more of a player than I am and it is his amp. If he wants to comment here he is free to do so.http://www.flickr.com/photos/myles111us/sets/72157622045828365/with/3868175750/ 

Mark Baier Wow..great photo and concept! Quality control! 5 years ago at the NAMM show, I actually had the owner of one of the major tube companies tell me that they had 100% acceptance rates, that EVERYTHING they built was good to go, with no rejects at all. I was astounded by his declaration, so I asked the question again (there was a language barrier of sorts) but he reiterated; "Everything we build is the highest quality, we have no rejects"..I've never used one of his tube since. Not surprisingly, Myles, your recent tests proved his tube to be the worst of the tested samples. 

Scott Evans · Friends with Mark Baier and 1 other

I'm constantly amazed at the tubes that are presented to me (a consumer) as being 'tested' and 'approved' but yet arrive at me either DOA or microphonic... after a few bad run ins I tend to stick with JJ/Ruby

Myles Rose ‎Scott Evans - The folks at Magic Parts / Ruby Tubes are first rate. 

Michael Didier http://www.kcanostubes.com/ they handle great selection NOS 

Myles Rose ‎Mike Kropotkin at KCA is a great NOS resource and a great technical resource as well. I have a few folks that I personally use that I can count on one hand. He is one of them. 

Leonard Rodriguez Tube smashing machine? 

Myles Rose In the past when things did not meet design spec they were destroyed to assure the well deserved reputation of the company. Today, a very high percentage do not meet design spec but the tubes are sold anyway. So, in the end, today's tube manufacturers also deserve their reputation as well. 

Michael Didier Yes Myles however with the internet a companie's reputation can go sour rather quickly based on users experiences along with word of mouth. Better to keep them coming back rather than a one time sale.  

Myles Rose Unfortunately, there are only a few tube makers and all of them produce products that could be improved. There is really not one that is any better than another over time as the inconsistency shows one tube to be "better" than another one week and worse than most the next week. Tube amps and equipment are still being produced and the makers have to alter designs to work with products (tubes) that are of suspect quality, performance and consistency. In the end the amp maker has to charge more for a product due to increased labor in tube selection to make the product work in the first place, possibly incur damage to their own reputation if warranty issues are not handled to customer satisfaction and bear the burden of the first weak link in the chain ... the tubes supplied by manufacturers who take little if any pride in their products. 

None of the current tube makers have anything to be proud about from my own observations over more than the last decade. It is almost as if there is a secret society of tube manufacturers that meet in the dark shadows and agree to raise prices and reduce quality to raise profit substantially above what is considered fair in many other industries. As one tube is discovered to be an over priced item that has performance that does not support it's retail price the maker just slaps on a new name and label. Sometimes they will change a cheap part that does not cost them any more, a part that does nothing to change the performance of the tube but is visible to the naked eye. Perhaps gold pins will even be employed to allow them to sell the tube as TWO new models with a higher price.

Johnny Martin · 5 mutual friends

That's some good stuff, thanks for posting. Sounds like it is just a crap shoot for us when buying today's tubes, and even then, they are not too great..bummer, now I really don't know what to buy, haha. 


Myles Rose It is a crap shoot for sure. You can push the odds more in your favor by having a good/trusted vendor who tests properly. That is half the battle where you still have some control. Where you have no control is the reliability of the tube. What works today will probably fail during the next year.

John Clark 

Yeah Myles, you'd think that at least one current tube manufacturer would try to make a more consistant quality tube & then corner the market with that fact. I'd gladly pay a few bucks more if I knew I was getting a decent product. I was very interested in what the guys at the old Blackburn factory were doing on the preamp tube front until their financing when kaput. Their 12AX7 offering seemed like the first radical design reinvention of preamp tubes that actually performed & sounded good. I'm still keeping my fingers crossed that those guys will get some funding again or someone else will take up where they left off.

Myles Rose ‎52,000 reads as of 3/18/12 at 2:47pm PST. Not too shabby.

Hopefully one of the folks at New Sensor, JJ, Shuguang or one of the other tube makers were one of the readers. Tip for any of them: Make a 12AX7 (that is a good first step) that retails for $50.00. It would meet design spec for a 12AX7 that was published from 1950-1970, the same spec that you folks say you follow today. The tolerance allowed would be +/- 10% on transconductance, plate current and plate resistance. The tube would carry a life guarantee of 10,000 hours against failure or physical microhonics.

This type of performance was commonplace in the 1950s, 1960s, 1970s.

Any takers? 

Scott Lerner I don't have a single new production 12AX7 that sounds better than a good NOS. I have a ton of both. I always wind up with NOS when I roll tubes!! Thanks Myles...

Myles Rose Scott - that is pretty much right in line with what I see on a very very very regular basis. Since 2002 or so things have been on the decline each year. Prices go up, quality goes down but the number new new names on the same tubes increases the number of the same choices as we had before with reduced quality and declining expectations on the part of the amp user/builder/owner toward the tube manufacturer.

I was asked about plate resistance.  Sometimes mathematics is the only easy way to try to explain:

Vacuum tube Coefficients:

The three most important tube coefficients used for determining the performance of vacuum tubes are, the plate resistance, trans-conductance (or mutual conductance) and amplification factor. These tube coefficients can be determined from the three families of characteristic curves, described above, by noting the slopes of the curves.

Plate resistance is defined as the ratio of change in plate voltage to the change in plate current, grid voltage remaining constant. It is denoted by rp and measured in K

Mathematically rp = , ∂Ep /∂Ip Eg being constant

The value of plate resistance indicates how the plate voltage influences the plate current at constant grid voltage. The coefficient rp is often called as dynamic or incremental plate resistance. The coefficient rp is also obtained by taking reciprocal of the slope of the plate characteristic curve. Typically by gm and measured in micro-mhos.

Mathematically gm = ∂Ip/∂Ep , Ep being constant

The transconductance is the most important tube coefficient as it reveals the effectiveness of the control grid in securing changes in the plate current and, hence, in the signal output of the tube. Its value can be easily determined from the slope of mutual characteristic of the tube. Typically it varies from approximately 1,000µΩ

Amplification factor is a measure of the relative effectiveness of the control grid in overcoming the electrostatic field produced by the plate and is defined as the ratio of small change in plate voltage to the corresponding change in grid voltage, plate current remaining unchanged. It is denoted by Greek letter

Mathematically µ = δEp/δEg -, Ip being constant

The –ve sign indicates that the changes in two voltages are oppositely directed.

Amplification factor is nothing but the slope of the constant plate current characteristic. Typically it varies from 15 to 100.

The three coefficients are inter-related, in accordance with their definitions, as is easily shown by multiplying the relations defining the plate resistance and transconductance.

Rp X gm = ∂Ep/∂Ip ×∂Ip/∂Eg = ∂Ep / ∂Eg = μ

Side note:  On the right side of this blog page I have many links for musicians and guitarists.  There are some links of tube suppliers/vendors that I personally use and recommend to my GAB clients.  There are a number of good tube vendors out there but there are poor ones as well.  The folks I list I trust.

Some people think that a 6CA7 is the same tube as an EL34. They are not. They are interchangeable but are NOT the same tube. New tubes called 6CA7 tubes are little more than big bottle EL34 tubes. Not the same. This is a real example of a real 6CA7.

6L6GC RCA black plate NOS same date codes around 1970, original CJ date codes. 

A little more tube porn before I head out for the day ... a collection of some EL34 tubes. EL34 tubes are of lighter construction than a 6CA7 and the vacuum is much softer. This is also the case of EL34 compared to something like the 6L6. American tubes had a higher vacuum and were of heavier construction. On average, an EL34 running at the same duty cycle as a 6L6 had about 1/2 the life.

OK ... what the heck ... one more tube porn shot. These go for about $300 each today. As I said in my NOS article, folks that have what they call an NOS tube collection actually have a financial investment portfolio. I talked about this the other day with Ritchie Fliegler. During the economic dive in 2008 gold top Les Pauls even lost some of their value. NOS tube costs continued to rise.

A bit of an update.  I posted the below on facebook and copied it over here.

Are you kidding me? This is a pretty decent tube, the Tung Sol Reissue, but .... 

.... but does the reissue directly above look similar in any way to the original? 

The two shots above, some of the dialogue:

Scott Frye If the OLD Tungsol was so kick-ass why didn't the make the re-issue just like it??

Myles Rose Because that would have cost a lot of money and would take time and effort. It is a lot cheaper to take an existing tube, the 12AX7EH and just silk screen on a new name that you purchased the right to use.

Producing a true reissue is very costly. Groove Tubes re-created the Mullard 12AX7. The initial runs were great but each revision had more and more problems surface. In the end the tube was garbage. Their GE 6L6 was a fantastic reissue right down to the 5 clad plate material and micas made in the same factory that made them for GE. Still my favorite 6L6 tube. Their Mullard XF2 EL34 reissue was great in some versions and a bit less terrific in other versions. All in all I liked the tube a lot. Their 6CA7 GE reissue, again... an attempt at an exact copy, was plagued with problems. Right as GT was being sold to Fender it was discovered that the factory had not followed the design drawings properly and there was a problem in the relationship between grid number one and two.

Aspen Pittman, as much as I can complain about some of his antics at times, did some really great things. When he produced a tube reissue he paid for the design, the tooling and hired the expertise such as the folks that worked in the original factories. He made a good strong effort, no expense spared, at making a true reproduction of the original tube right down to the materials whenever possible. Others in this industry just slap a grand old name on a current production tube that has nothing in common with the original other than having a glass bulb and in most case, the proper number of pins sticking out of the base.The original tube designs of GT such as the 6L6CHP (black plate high performance) and the 5881 short bottle were both terrific tubes. Aspen and Steve Sanett of Penta Labs had a great relationship with Tesla and later JJ. The KT88SV was a 50 watt tube with heat sinks on the plates. The E34LS was a 30 watt EL34 that also had a different plate assembly than the JJ which was a 25 watt tube as were other EL34 tubes. These were GT exclusive. The KT66HP was a grand tube. It compared with the real Genelex KT66s of the past. It was made under contract in the Reflektor Factory in Moscow before Mike Matthews took over the entire place. The tooling was GT developed and owned. Around the time that Sovtek was taken over by New Sensor the tooling mysteriously vanished and that was the end of the grand KT66HP. 

Aspen Pittman was directly responsible for some terrific tubes, both accurate reproductions and some great original designs.  He is the only person at this point that you can point to and make that statement.

Myles Rose Economics lesson update for Scott on his original question .... the total cost to turn an existing 12AX7EH into a 12AX7 Tung Sol Reissue would be less than $15 for a new silk screen. Perhaps less, that is what one cost over at Groove Tubes to make a new silk screen from scratch. If New Sensor did not mind tossing out the old 12AX7EH boxes or timed it right when you ordered new boxes that say Tung Sol on them, the cost of new artwork would have to be added. Same for the Mullard reissue preamp tube New Sensor sells or the "Mullard" EL-34. 

My guess based on a bit of experience would be under $100 with enough change back to buy a pretty nice dinner for two at a chain like Black Angus. 

As a side note, the cost to re-produce the Groove Tubes 6L6GE was in the ballpark of $400,000. Like I said, when Aspen Pittman wanted to make a recreation he went full bore in the right direction. Side note ... Aspen has an extensive microphone collection. One type uses a tube that is very hard to procure and very costly. There would be a market for these tubes, may a few hundred at the most. He was seriously considering remaking that tube.

Here are a few links folks may find interesting:

http://cds.cern.ch/record/983744/files/p117.pdf

http://www.john-a-harper.com/tubes201/

http://www.wikihow.com/Find-out-if-an-Electronic-Tube-Is-Good

http://www.w8ji.com/vacuum_tubes_and_vaccum_tube_failures.htm