Overview of some current 12AX7 tubes

[Myles]

My amp repair guy says ARS is total garbage. On another note, Doug's tubes says they have preamp tubes that test out to 110 all day long. Why do I keep reading that everybody else's test out at 85??

110 on what kind of equipment?

There are not too many testers that will read true gain as it is a combination of factors ... transconductance AND plate resistance AT 250 plate volts and 2 volt bias WITH 62.5k plate resistance or 100v 1v bias 80k plate resistance etc.

If he is using something like a VTV tester that has a scaled marked "gain" this is NOT gain and in fact the scale might as well be labelled "frommer" or "blinkies" as it has no relationship to gain at all.

If you look in the RCA receiving tube manual or on GE, Mullard, or anybody elses spec sheets the info on gain is quite detailed and specific.

A Hickock, TV7, or even RCA WT-100 will not measure true gain but if you can measure TC and plate resistance it can be calculated.

As an example, here is a test sheet on a Sovtek 12AX7WC:



... and here is one on an Ei long smooth plate from the past that would not look too good on a common TC tester but due to their high plate resistance they had great gain.



I would love to know where this gain of 110 you talk about came from via what tester at what voltage etc.

[Steve Drojo]

^^

[Myles]

I wonder what will be a good tube for the Phase inventer? Will this tube in the PI affect the tone of the amp?, for example if i choose a dark tone tube , will the amp also have a darker tone?

 

 

Some folks say the phase inverter does not effect tone. I do not agree. In many amps how the phase inverter breaks down and distorts with what it passes on to the output tubes is what you experience long before you hear the output section distort.

 

I personally prefer long plate tubes for PI use as they seem to break down more smoothly and do not cave in on themselves as quickly. There are many factors here but this is the basic thought I have on the PI issue.

[Myles]

Regarding the 12AX7R3.... Any theories as to why it sounds so good as the Tung Sol version, and so bad (imo) as the EH version? I mean, it's really night and day. The Tung sol is fat and warm. The EH is thin and harsh on the top end.

Are you sure you are comparing apples to apples? You need to have one of each tube that measure the same or else all bets are off when trying to compare. You may have one tube near spec and another way off spec.

[Myles]

Myles, on the regular old tube testers looking at 12AX7 tubes, what is being read by the tester when the meter points to a certain number? I know you're talking about a gain factor of 100 or 85 or whatever for the 12ax7s, but I'm guessing the old testers are just pointing to a relative number in relation to the voltage or current coming out of the tube? From what I've gathered you actually need something along the lines of one of the VTV testers to read gain, correct?

 

 

It depends on the tester but most read transconductance which is related to gain but NOT gain. You can have very high TC and NO gain if the plate resistance is low.

 

The VTV gain scale IS NOT A GAIN SCALE AT ALL. That scale could have been labelled ANYTHING. Check with Charlie at VTV or Eric and ask what that scale really means. It is a good tester to test balance and some basic things but gain is not one of them.

[Doug's Tubes]

It depends on the tester but most read transconductance which is related to gain but NOT gain. You can have very high TC and NO gain if the plate resistance is low. The VTV gain scale IS NOT A GAIN SCALE AT ALL. That scale could have been labelled ANYTHING. Check with Charlie at VTV or Eric and ask what that scale really means. It is a good tester to test balance and some basic things but gain is not one of them.

 

 

Hey all. Actually, I checked with Eric Barbour the designer of the VTV tester who should pretty much know all there is to know about this excellent piece of gear which Myles first turned me on to back in 2001 and encouraged me to purchase. When a similar question arose back in April 2007 in a thread right here on HCAF, I emailed Charlie and Eric. Here is Eric's verbatim response to basically the same comment above about gain. I've left off the last part of the email which contains extremely vulgar inflammatory personal comments irrelevant to this discussion so as to stick to the facts regarding the tester:

 

This is bull{censored}. The testers' meter shows the actual AC voltage

output of the tube. It is a dynamic measurement, taken from the

output of a dead-typical RC coupled gain stage. Very similar to any

mutual-conductance tester of the past--only taken with a low-distortion

sine wave oscillator and a precision AC voltage detector.

It just uses an arbitrary scale, because it's impractical to

measure actual mu in this way. So did all other previous tube testers.

 

If that isn't "gain" then the word has no meaning.

 

Mu is only accurately measured by taking plate curves.

In a typical circuit, the voltage gain will always be LESS than

the mu. And mu varies (a lot!) with plate voltage and loading.

That's why one needs an arbitrary meter scale.

Old mutual-conductance tube testers were never very accurate

in any case. And what good is knowing the mu? Does the tube have a lot

of gain, or not?

 

[Myles]

Hey all. Actually, I checked with Eric Barbour the designer of the VTV tester who should pretty much know all there is to know about this excellent piece of gear which Myles first turned me on to back in 2001 and encouraged me to purchase. When a similar question arose back in April 2007 in a thread right here on HCAF, I emailed Charlie and Eric. Here is Eric's verbatim response to basically the same comment above about gain. I've left off the last part of the email which contains extremely vulgar inflammatory personal comments irrelevant to this discussion so as to stick to the facts regarding the tester: This is bull{censored}. The testers' meter shows the actual AC voltage output of the tube. It is a dynamic measurement, taken from the output of a dead-typical RC coupled gain stage. Very similar to any mutual-conductance tester of the past--only taken with a low-distortion sine wave oscillator and a precision AC voltage detector. It just uses an arbitrary scale, because it's impractical to measure actual mu in this way. So did all other previous tube testers. If that isn't "gain" then the word has no meaning. Mu is only accurately measured by taking plate curves. In a typical circuit, the voltage gain will always be LESS than the mu. And mu varies (a lot!) with plate voltage and loading. That's why one needs an arbitrary meter scale. Old mutual-conductance tube testers were never very accurate in any case. And what good is knowing the mu? Does the tube have a lot of gain, or not?

 

 

 

Doug ....

 

I would be happy to have Charlie and Eric come down again. There are many ways to define gain but the method I use is the same method used by the major tube companies of the past.

 

In essence ..... I would be happy to have Eric and Charlie bring down tubes that measure 100-120 on the VTV tester and see what readings are found on different equipment. I will guarantee the numbers will not match.

 

I am not saying the VTV tester is not a great device. I am saying that if you read a number on its gain scale you probably will not come up with a similar number on devices that cost many thousands of dollars rather than around $800.

 

I would be happy to work with either or both of them.

 

I am sure this is a matter of symantics for the most part and that is why I stick by RCA etc., book spec and test methods.

 

As far as the VTV tester ... I have one myself and it is in fact serial number 0001. It is a great tester for many things but does NOT indicate actual gain as per tube sheet specs at 250 plate volts, 2 volts of bias at 62.5 of plate resistance. I am not talking about mu or TC here. Gain is a factor that takes TC into account with other factors and as I said earlier, the testers that measure things like TC are NOT measuring gain either. They are measuring another type of output. Sort of like torque vs horsepower in a way.

 

I am not here to argue testing methods. I just wrote this post initially to give a few basic thoughts toward preamp tube characteristics.

 

Doug... since you order stuff from Rick, next time you need anything tell him to ask me to grab a preamp tube or two from your order and I will mark them with some sort of ID number and send full test reports back with them. You will see the true actual gain as well as many other measurements. Then put them on your tester and try to make any of the numbers match the number on the VTV tester. When you discover this is not possible send the tubes and sheets to Eric Barbour and ask him for an explanation. You will find out some interesting stuff I am sure. The VTV tester is a very cool unit and has it's uses but it is not a piece of tube development lab equipment or even a high end tester such as an RCA WT100. It offers a lot of bang for the buck.

 

The tools we use are for tube engineering and development and not too many folks have those sorts of things at their disposal. General tube testers are great for a lot of things but when one is trying to find out as many aspects of a given tube as possible for development the tools will differ. A cool side benefit though is to find out how each tube makers production changed and point out these changes so we can report back to the factory when something is amiss.

[echodeluxe]

wtf are tubes?

[Lambros]

Myles, thanks for the detailed explanation. Basically my hunch was right that this new tube is fairly tame as far as long plates go. I've tried and owned a lot of the classic long plates (Mullard, Amperex, RCA, to a lesser extent Brimar) The one that came closest to the Telefunken vibe was the Ei when it worked for the short life these tubes are infamous for. Like you said you find the one that works...your beside yourself on how good the thing sounds...then they go really bad in a matter of weeks, days in some cases.
I figured 8-10 tubes was enough, anyway.
My favorite now is the long plate Mullard with the halo getter, it had more umph than the Amperex bugle boy and it was definitely designed to be less microphonic...incredible...such a nice compliment in the V2 spot to the short plate ecc83 in V1. It tried in the V1 too it worked fine it just didn't have the punch of the short plate Mullard 60s vintage (copper supports and thicker mica). Anyway Myles, what it boils down to is that I was hoping that the new JJ was as good as the 83s. I hope they make improvements with it.

Here is something I just wrote to the same sort of question on TGP on the ECC83S. The long plate ECC803S does not have the gain that the shorter plate ECC83S has. The JJ long plate runs in the center area of today's current tubes on average. I like the 803S and I think I have one in V2 of my own 65amps London. Anyway ... here is what I said in the other forum: Quote: Originally Posted by frank4001 Thanks Myles....Can you tell us who makes the the 12ax7 that is closest to original 12ax7 specifications? I read where new ones fall short in some area like gain ect. Frank, This changes from run to run and batch to batch. What is generally true, and true at the moment, is the JJ ECC83S is the highest gain 12AX7 that is currently in production. Most 12AX7s today have a true gain of 80 or less on average at 250 plate volts with a 2 volt bias. The JJ has many over 85 and I have seen a lot over 90 and have hit 105+ on some in some batches. One of the lowest gain tubes is the Sovtek 12AX7WA. The Sovtek 12AX7WC is not in the same class. It is average in specs with most of the other folks and is a good tube if you like the tone. The long smooth plate Ei was the highest gain of anything out there. High gain and sloppy construction makes the tube physically microphonic a lot of the time but when you find a good one they are great. The GT 12AX7M originally had great gain in the first tooling run or two. For the next 8 tooling changes it had the lowest gain of anything out there. The tooling change that was done from October to December finally got things to look right. Now the tube has gain on par slightly above the average production 12AX7, maybe like a good 12AX7EH as an example but darker in tone than the EH or Tung Sol reissue. The JJ is still, at the moment, the most proper spec tube with the highest gain. That is the tube I use for V1 and V2 positions in the SAG-MHG kits I build for my clients and that GT sells as it is the only tube out today that can produce a gain of above 85 in decent enough quantity to make selection possible for a decent amount of tubes in 100. In 100 JJs I can get about 40-50 over 85 in gain. On the others it would be less than 20%.

 

[Osram]

Are you sure you are comparing apples to apples? You need to have one of each tube that measure the same or else all bets are off when trying to compare. You may have one tube near spec and another way off spec.

 

 

Well this is based on a sample of 8 TS's and 3 EHX's.

 

All TS's sounded the same, and all EHX's sounded the same.

Whatever fairy dust Matthews sprinkled on the TS version must be some powerful stuff:eek:

[the_bleeding]

How would one tell the difference between the different "generations" of chinese 12AX7's?

 

 

9th gen's have octagonal plates, where previous generations have smooth round ones for the most part. I run a marshall and i always make sure my chinese tubes have the octagonal plates.

[1big1]

myles, i was wondering if you've had any experience swapping pre amp tubes in the line6 spider valve and if so what effect it had on the amps tone?

[Doug's Tubes]

Doug .... I would be happy to have Charlie and Eric come down again. There are many ways to define gain but the method I use is the same method used by the major tube companies of the past. In essence ..... I would be happy to have Eric and Charlie bring down tubes that measure 100-120 on the VTV tester and see what readings are found on different equipment. I will guarantee the numbers will not match. I am not saying the VTV tester is not a great device. I am saying that if you read a number on its gain scale you probably will not come up with a similar number on devices that cost many thousands of dollars rather than around $800. I would be happy to work with either or both of them. I am sure this is a matter of symantics for the most part and that is why I stick by RCA etc., book spec and test methods. As far as the VTV tester ... I have one myself and it is in fact serial number 0001. It is a great tester for many things but does NOT indicate actual gain as per tube sheet specs at 250 plate volts, 2 volts of bias at 62.5 of plate resistance. I am not talking about mu or TC here. Gain is a factor that takes TC into account with other factors and as I said earlier, the testers that measure things like TC are NOT measuring gain either. They are measuring another type of output. Sort of like torque vs horsepower in a way. I am not here to argue testing methods. I just wrote this post initially to give a few basic thoughts toward preamp tube characteristics. Doug... since you order stuff from Rick, next time you need anything tell him to ask me to grab a preamp tube or two from your order and I will mark them with some sort of ID number and send full test reports back with them. You will see the true actual gain as well as many other measurements. Then put them on your tester and try to make any of the numbers match the number on the VTV tester. When you discover this is not possible send the tubes and sheets to Eric Barbour and ask him for an explanation. You will find out some interesting stuff I am sure. The VTV tester is a very cool unit and has it's uses but it is not a piece of tube development lab equipment or even a high end tester such as an RCA WT100. It offers a lot of bang for the buck. The tools we use are for tube engineering and development and not too many folks have those sorts of things at their disposal. General tube testers are great for a lot of things but when one is trying to find out as many aspects of a given tube as possible for development the tools will differ. A cool side benefit though is to find out how each tube makers production changed and point out these changes so we can report back to the factory when something is amiss.

 

 

 

Like Eric said, who cares what the mu is, or how current production tubes dont stack up against the tubes from the good old days like we already know they dont!! Thats irrelevant. I'm merely using AC voltage in a consistant test environment on quality testers engineered by a degreed electrical engineer with 25+ years of electrical engineering experience for comparing results among a given group or batch of tubes to measure their relative gain to each other. You want a lower gain 12AX7, or a higher gain 12AX7 from the batch? The VTV tester provides me with an obective way of determining that in a consistent manner.

[outtahear]

Don't bother being pragmatic in this thread-

 

Iffin' you ain't bloo-printin'........

[Kelly]

So what kind of tester does Groove Tubes use that makes tube prices double??? I could use one in the cabinet business.

[outtahear]

It measures upsidaisium levels in cubits-like Moses and RCA intended.

[knucklefux]

i think with GT, you end up paying for the tube you got, and a tube or two that was thrown away on your behalf:thu:

[outtahear]

i think with GT, you end up paying for the tube you got, and a tube or two that was thrown away on your behalf:thu:

And yet, as a professional amp tech for 25 years, GT's been dead last in Q.C. IME, especially in the last 10 years or so.

Could it be other overhead?? (Lotta advertising/promo-type clowns..)

Lord Valve's been best for new stock, Mike @ KCA for any NOS I don't have, and need ASAP. I've had customers who've had great experiences w.Doug O' Dougs Tubes, as well as BoB @ Eurotubes. Non-huge types, who seem to get it right the first time, more often.

A great exchange policy is bull{censored} in liu of bad QC. It costs me time, ergo $$$$$. to send back (on my coin???) bad product to the same people who let bad product slip the first time.
Especially when the same glass is available for less $$, and that works correctly a better percent of the time.

That's no {censored}in' bargain, in my book.

Free advice, neato graphs, and balanced inverters are great, I suppose but I've never worked any "name" gigs where it's been a prefered, let alone requested set up. (Even after the "Lets give it a shot" earball.)

Also would like to point out what might be just the ticket for a Tele/Dr z blasting country, tends to neuter a Marshall-type circuit geared for a less, oh "polite" style of musical expression.


IME, YMMV, JFAC, FDIC.

[Steve Drojo]

If they all light up, it's a good day for me!

[outtahear]

Kinda like fans......

[JerryP]

i think with GT, you end up paying for the tube you got, and a tube or two that was thrown away on your behalf:thu:

 

 

I believe many that don't live up the their standard find their way to Ebay. They aren't just thrown away unless their totally dead or shorted would be my guess.

Jerry

[Myles]

9th gen's have octagonal plates, where previous generations have smooth round ones for the most part. I run a marshall and i always make sure my chinese tubes have the octagonal plates.

 

 

Generation 8 has squarish micas (you are talking micas here and not plates by the way). The squarish ones have four points of bottle contact where Gen 9 have 8 points.

 

Gen 7 has round micas for both of the upper mica pair.

 

Gen 4-6 have a different getter arrangement with the lower round mica.

[Myles]

myles, i was wondering if you've had any experience swapping pre amp tubes in the line6 spider valve and if so what effect it had on the amps tone?

 

 

I have not done that.

 

Sorry.

[Myles]

Like Eric said, who cares what the mu is, or how current production tubes dont stack up against the tubes from the good old days like we already know they dont!! Thats irrelevant. I'm merely using AC voltage in a consistant test environment on quality testers engineered by a degreed electrical engineer with 25+ years of electrical engineering experience for comparing results among a given group or batch of tubes to measure their relative gain to each other. You want a lower gain 12AX7, or a higher gain 12AX7 from the batch? The VTV tester provides me with an obective way of determining that in a consistent manner.

 

 

 

Doug ... if what you are after is a "weaker" tube or "stronger" tube then the VTV tester is fine in many cases but I have found in some cases it can be misleading, especially if you try to use the phrase "gain". If you want to take two tubes that show the same "gain" on the VTV and send them to Rick I will be happy to run full reports on them and have the tubes and reports sent back to you so you can see that even though they may read close on the VTV in the one area it tests they might read very differently in other factors including gain, plate current, screen current, transconductance and plate resistance.

[Myles]

So what kind of tester does Groove Tubes use that makes tube prices double??? I could use one in the cabinet business.

 

 

 

On GT pricing you'd have to talk to Rick Benson or Aspen. The tester I use for development and QA is not the same tester as is used in production as it takes about 15 minutes to test one tube.