Read Tubeprimerandselection.pdf text version

WHAT TUBES SHOULD I USE IN MY GUITAR OR BASS AMPLIFIER? A tube primer by Guitar Amplifier Blueprinting

Myles S. Rose

What are the differences between tubes? What is a phase inverter? What about a matched phase inverter? What is matching? Static versus dynamic matching? Aren't all tubes really the same? Is 100 watts twice as loud as 50 watts? How much power do I need? These questions and more are covered in this document


Version 1.31 6/10/03 ­ biasing information added

What Tubes Should I Use In My Amplifier? - A Modern Day Tube Selection Primer

Myles S. Rose ­ Guitar Amplifier Blueprinting - 2003 & Groove Tubes Special Applications Group Today in 2003, there is a lot of misconception on where tubes come from, how they actually perform, and what are their characteristics. Information that was valid only a year or two ago, is not always valid today. "Chinese tubes are less powerful than the Russian tubes". "Chinese tubes do not hold up to high voltages". "Matching is matching", or "there is no need to match tubes". This information gets passed along, and becomes something of gospel, when most of it is just plain rubbish. In the case of the Chinese tube issue, a few years back, this generally was the case. Today, the typical Chinese power tube is 15% stronger in many cases than it's Russian counterpart. The Chinese have great tooling, and money. They are putting a lot of resources into tube manufacture. The Chinese 12AX7 as one example, is the most consistent tube in quality and specifications when compared to a lab spec 12AX7. As your proceed with this paper, examples and characteristics of many tubes will be given. These are not bits of data collected from magazine clippings, Internet spatter, or hear say. This data comes from my personal time sitting with thousands of tubes, one at a time, with very high end test equipment ­ not simple tube testers. Amps come in all types, sizes, styles, etc. You cannot turn a Marshall into a Fender, or visa versa, no matter how hard you try. If you "need " both of these sounds, the bottom line is, you need one of each of these amps. Some folks like Paul Rivera, have two different voicings in the same amp. This is pretty unique, as most folks just change the gain structure on the various channels, with the same voicing.

A 50 watt Sovereign amp from Victoria Amplifier ­ very versatile with it's variable front end which uses an EF86 in the front end, a 6BM8 in the reverb circuit, a cross between a Plexi era Marshall and Tweed Bassman. This is not a master volume amp. You can turn the gain all the way down and it's pure Plexi era. Turn the gain up, and it moves through JCM 800, 900, 2000, and beyond Mesa Rectifier territory. This is one example of an amp that is very unique, and built, and very unique. I have a full write-up on this amp someplace on my personal website. If you would like a copy of the review sent to you via email, just drop me a note. You can also learn more at for Victoria in general. A review of the Sovereign is at


Some of the classic s ......... Fender Black Face Deluxe Reverb and the Tweed Deluxe (in this case a Victoria 20112T) Perhaps to some folks thinking, the best small club and blues amps around.

The Dr. Z Carmen Ghia. Sure, the good Doctor makes lots of amps, but this is one of the most amazing pieces of gear around. This is a studio amp masterpiece and also a super club amp, Dr. Z is one of the folks out there that copies nobody. His amps are all very unique and built to the very highest standards of construction. His Route 66 has won awards that were not political, they were from real players. The Z28, MAZ 18 and 38, and other models each do their target aspects VERY WELL. For much more on Dr. Z amps, internal photos, comments, pricing and more, head to


If you are one of those folks that thinks that wattage = manhood ... this older amp may be your cup of tea. The top section is two 60 watt EL34 based amp sections, and the bottom slave is 8 GE 6550 tubes, for a total of about a ½ a kilowatt of tube power. This amps runs with four 2x12 cabs and four 4x12 cabs. For more on Rivera, and their amps (which are a lot more current that this particular example, head to There are amps for every taste and need out there! Some of my other personal favorite folks out there can be seen in the amps / guitars area of my website at

Preamp Tubes

Please note that the tests on tubes was done on raw factory samples. Some tube vendors test and/or select or grade, and some do not. Many rely on the end user's reluctance to return a single tube such as a preamp tube for problems, so just send their tubes out with no testing or minimal testing. A preamp tube that is not microphonic when first installed, can become microphonic in short order after a few heat up and cool down cycles with expansion and contraction. This is one reason to know and trust your vendor. Some companies such as Groove Tubes have a very long warranty on tubes, in the case of preamp tubes, six months. Some reasons for the cost of GT premium tubes are due to the labor intensive testing process (each tube, one at a time, in place in high gain amps and tested for various attributes), and their cost for the tubes initially when buying from the major tube factories. In many cases, the GT reject rate is in excess of 60%. The data below, are for the raw factory samples, not GT tested examples. This is the way they come from the factory, and the sort of performance you can expect if your vendor does not test for gain, output, noise and microphonics. In the case of power tubes, gas leakage, grid leakage, and low vacuum are also test parameters that require expensive and complex test equipment which very few vendors have in house. The complete list of Groove Tubes preamp tubes can be seen at:


The primary and most common family of preamp tubes is the 12AX7 / 7025 / ECC83 family. These are directly interchangeable in most guitar or bass amplifiers. Contrary to popular opinion, these are not the same sounding tubes in many ways. They each have their own sonic signature and sound characteristics. Below one can see the difference between a commonly used Sovtek 12AX7WA (the green line) and a long smooth plate 7025 (the blue line). The 7025 is a brighter tube, with more treble and high mid response. The 7025 was used in the Black Face Fender era, and was one aspect of the bright clear sound of these amps.

There are many factors in a tubes character. In the case of preamp tubes, aspects such as gain (Mu), output, transconductance, and other factors, are all part of a bigger picture. Many people use the terms gain and output interchangeably. This is NOT correct. A tube can have high gain, but be very low in output current capability. This is a very common mistake today. In a clean front end preamp circuit, such as a tweed or blackface era amplifier, low output is less of a factor than in a modern channel switching amp with a complex circuit topology design. These newer amps perform their best if a high current drive tube is utilized. Today's most common factory tube, the Sovtek 12AX7WA, is picked for many reasons. It is not expensive, it is sturdy (so the amp makers have less warranty return), and it is generally low on gain and output. The latter characteristic makes the amps more quiet with less background noise, and this helps cover up other manufacturing or design shortcuts. Today's tubes are VERY inconsistent. You may re-tube your amp, just to find that is sounds worse than before the replacement. Why? As one example, a "spec" 12AX7 should have a gain of 100 for a "typical" tube based on specs set many decades ago. It should have a transconductance of 1600, and an output of 1.2 milliamps. If your V1 (first gain stage tube) had an output of say 1.0 milliamps before it was replaced, and the new tube had 5

an output of 0.8 milliamps, right there you just lost 20% in your front end. Remember, this signal gets amplified now down the rest of the amplifier chain. 1.2 milliamps is the "spec" number we look for in a 12AX7 / ECC83 / 7025. Unfortunately today, the vast majority of the tubes we see made newly are on average, much less, at times 0.7 milliamps on average. This is almost a loss of 50% of our expected output. Poor plate materials that are inconsistent, fast work, incomplete vacuum pump down procedures, cathode coatings with high percentages of impurities, and more. These all contribute to today's inconsistency. These are some of the areas where the Chinese are improving faster than in other geographic areas. The most important tube(s) in the preamp section are in the first stages, the tone and gain stages. Changing V1 in many amps will yield the most results. Tubes used for current drivers as their primary function, such as reverb drivers or effects loop drivers, do little in most amps to change the tone. This is generally not part of the tone or gain stage. The exception to this is the phase inverter or driver, of the power tubes. There will be more on this subject later in this paper. Moving on to the tubes. What we are looking for are specific factors in testing, and other factors in sonic aspects. Rather than list all the technical data for each tube, where I have this data, I will provide additional information for gain, output, and consistency. The consistency factor I will call QA for quality. For this "spec", the lower number, the better. A perfect tube factory run would have 0% inconsistency. A factory run where the tubes were +/- 20% off target factory expected specifications, would have a number of "20". The standard specs that are the target numbers for gain is 100, and output is 1.2 milliamps, in the case of the 12AX7/ECC83/7025 family. Output will be shown as a percentage of factory expected target. If a tube has 1.2 milliamps of output, it's Output would be 100%. The most important factor in all of the above, is know and trust your tube vendor. Tubes need to be tested for microphonics, low output, and noise. With today's material inconsistency, a tube that is quiet today may become microphonic after a few hours of use, so a vendor with a good warranty is one that one may wish to have in your corner. NOTE: This test data is from bulk factory batches of large quantities of tubes, at times 1000 tubes of each type from each factory run. This illustrates how random these tubes are, and shows what a chance you may be taking with untested tubes in general. These tubes, after testing using good methods, will narrow this inconsistency considerably. In one tube companies case, Groove Tubes, their reject rate on factory tubes can exceed 50%. You can sometimes easily tell how well a tube has been tested easily ... price. When there is a lot of test time and labor, and half of your stock hits the trash bin, the end retail cost has to be higher. The data below was collected at one part of late 2002 and early 2003. For the most current data, see the updated information elsewhere in this document. 12AX7-C ­ Chinese 9th generation tooling Gain: 93 Output: 92% QA: 17% This is possibly the best of the current 12AX7 family. This is a warm and linear tube. It is suited to rock, blues, and jazz. It is perhaps the most versatile of the 12AX7/ ECC83/7025 family. There are many Chinese variants, and many tube vendors sell older tooling versions.


12AX7-R ­ Sovtek 12AX7WA Gain: 86 Output: 77% QA: 42% This tube comes from the Reflektor factory in Russia. It is perhaps the most commonly used tube by many amp makers. It is reliable, and quiet. The main reason it is quiet, is it's generally lower gain and output.

12AX7-R2 ­ Sovtek 12AX7LPS Gain: 83 Output: 83% QA: 42% This tube is made in the Reflektor factory and sold under the Sovtek name. It is also known as the Groove Tubes 12AX7R2. It is a long plate structure, that is brighter than a 12AX7WA, 12AX7R3 (EH), or ECC83. It is not as bright as a 7025. It's long plate structure can be more prone to microphonics is some combo amplifiers.

12AX7-R3 ­ Electro Harmonix 12AX7EH Gain: 87 Output: 83% QA: 17% This tube is made in the Reflektor factory on newer tooling and also sold as the Electro Harmonix 12AX7EH. It is a shorter plate structure, a linear tube. It is not quite as warm or linear as the 12AX7C, but is just about as versatile. It is a quiet tube, and works nicely in many amplifiers. It is a touch brighter than the 12AX7C, but not as bright as a 12AX7R2 (LPS) or 7025. This photo shows a factory tube on a Kaye Audio Labs tester, before it goes through any GT processing. This is one aspect of the factory testing I perform at Guitar Amplifier Blueprinting. From noise data here, these tubes move on to other test equipment such as curve tracers. I go though thousands of samples each month from all the tube factories collecting the raw data to see how close or consistent to specs they come.



Above - This is typically how factory bulk tubes come to the tube vendors. This happens to be 200 Sovtek 12AX7LPS tubes (2 layers of 100 each).

On average, the reject rate on most new tubes can run 50% and higher if the tube seller has tight specs. Many vendors just ship whatever they buy, and hope for the best from the end user as far as them not complaining. After looking at a LOT of tubes, my advice is; trust your vendor.

ECC83-S - from the JJ factory Gain: 85 Output: 112% QA: 58% This is the sound of Marshall and the British era amps. A stronger mid range response, with a bit of roll off on the high end compared to many other tubes (but not the 12AX7WA). The BLUE line shows the typical response curve of the ECC83 in an average tested sample compared to the reference 12AX7WA (as the WA is a common tube in so many amps today). This tube may be brighter or darker depending on amplifier. This is due to the higher output of this tube and it's ability to "push" signal with it's high current through complex front end circuits where other tubes show their limitations.

ECC83's need to be purchased from a trusted tube vendor that tests tubes pretty extensively, as they can be very inconsistent. A good one is a find for sure. These are the masters of current drive, and in a modern amp such as a Bogner, Diezel, Rivera, or Mesa Recto series, these are the tubes that will push signal through those complex front ends. Also limited SAG in High Gain Kits:


7025-Y from the Ei factory Gain: 90 Output: 46% QA: 25% - 60% Long smooth plates, the most bright of the current tubes. The graph on this tube is at the start of this paper. These are the characteristic sound of the Fender Tolex years. These are very articulate. In V1 and V2 of a Fender Tolex era amp, these give the original sound signature. These tend to be too bright in Marshall Plexi era amps for some tastes when used with a Strat or Tele, but if you tend to load these amps down with pedals in the front end, they can help resolve the loss of treble from the pedals. Long plates can tend to be microphonic in combo amps. The Ei factory was the OEM for many "premium" tubes folks in the past, such as Siemens and Telefunken at times, so if the plate structure looks familiar, this may be a clue.

5751 New Manufacture Expected Gain 70 Expected Output 1.2 milliamps Used in V1 by some folks like SRV to change the ratio of preamp distortion to power tube distortion. Current manufacture from Russia is so far outside of NOS sample specs on curve traces, that this tube is not at all like NOS 5751 versions. Perhaps down the road things will change, but in 2002-2003 this is the case. If you use these new 5751's and like them, that is great, but they are not a vintage 5751. When there is a good new 5751 available, it will be noted here in an update.

12AT7-Y Ei / 12AT7-C / 12AT7-A USA NOS JAN spec Expected gain 60-70 Expected Output 10.0 milliamps These are commonly used as phase inverters and reverb drivers. In the first gain stage of an amplifier, they will drop the gain a bit, make some amps a bit quieter in background noise, and yield more clean headroom. This is also one way to change the percentage of output distortion to preamp distortion in non master volume amps when used in the first gain stage. One of the differences between the later Fender amps and the Marshall JTM-45 / Fender tweed era amps, was the use of the 12AT7 in the Fender tolex era amps in the phase inverter position, rather than the original 12AX7. The 6201M is a selected low noise tube for use in some microphones. It is an NOS JAN selected tube. As of May 2003, GT will generally be offering the 12AT7A NOS as it's 12AT7 if marked as 12AT7 without the Y or C designator. 12AT7-A 12AT7-Y (currently not in stock ­ replaced by 12AT7A) 12AT7-C (currently not in stock ­ replaced by 12AT7A) 6201M

12AY7 ­ various makers / 6072M NOS JAN Spec selected low noise Expected Gain 44 Expected Output 3.0 milliamps This tube was the first gain stage of many Fender tweed era amps for many years. It has about ½ the gain of a 12AX7, but much greater current output. In modern Marshall type amps such as a DSL or TSL series, this tube in V2 will bring the touch, feel, and sonic qualities, closer to Plexi era Marshall amps if this is an objective. This tube is also known as a 6072 or 6072M, a selected low noise version used in studio equipment and some microphones. These are the common V1 in Fender Tweed era amps in many models. 12AY7 6072M 10

12AU7-Y ­ Ei Expected Gain 16-18 Expected Output 10-11 milliamps These are commonly used as phase inverters in McIntosh Hi-Fi amps, some Ampeg amps, and will yield the most clean headroom in many amps. In the latest Fender Pro Series amps such as the Pro Reverb, Concert, and Twin, this tube in V2 (or a 12AY7), will drop the gain of the gain channel where some feel the tone is of a less intense or "buzzy" character. This will then allow a much wider range in the sweep of the volume and gain controls of this channel.

EF86 ­ 6267 Expected Gain 5000 Expected Output 3.0 milliamps QA: 50% + A 9 pin pentode, used in the first gain stage of some newer amps such as the Matchless, Bad Cat, Dr. Z, and others. The current versions of this tube from the Russian factory, are very inconsistent. Their output is way above spec, sometimes as high as 50% over what is to be expected. They have transconductance that is not close to the original tubes 80% of the time in factory sample tests. These factors make finding a low noise version of this tube very difficult. These need to be closely screened and tested by a tube supplier that has very sophisticated test gear that is generally out of reach for most Internet tube vendors or smaller vendors.

Preamp Tube overall chart of some NOS and many current tubes ( updated 5/21/03)

Tube Tolerance Ave output % of mA expected Ave TC (gm) TC % Average gp Average Gain

Guitar Amplifier Blueprinting

1600 Reference small signal tubes - in some cases, single tube tests. Items marked * were three to five tube test averages

GE 6072A 1970s Date Code * GE 6072A Black Plate 1963 * GE 12AX7WA (1980s) * GE 5751 (1950s) JAN/Philips 12AT7WC 1980s * RCA 12AX7 NOS 1950s-60s RCA 5751 Black Plate (1950s) * RCA 7025/12AX7 NOS 1960s * Telefunken ECC83/12AX7 - diamond base * RCA 12AU7 NOS Cleartop *

2.6% 1.9% 8.7% 2.5% 11.6% 8.8% 3.7% 6.6% 11.6% 2.1%

96% 100% 92% 98% 105% 99% 96% 97% 102% 96%

1710 1750 1520 1190 5780 1620 1251 1580 1680 2180

1750 1750 1600 1200 5500 1600 1200 1600 1600 2200

42/44 44/44 98/100 69/70 72/70 100/100 69/70 97/100 102/100 18/18 11

12AX7 Telefunken smooth plate new Dynaco 12AX7/ECC83 Telefunken Diamond Bottom Ribbed Plates * 12AX7 MINIWATT / SUPER RADIOTRON AUSTRALIA 1960s 12AX7 Amperex Bugle Boy Holland LONG PLATE 1950s * (10) 12AX7 AMPEREX BUGLE BOY HOLLAND 1967 * 7025 Amperex Holland orange Globe Logo 1971 CV4004 Brimar 1961 CV4035 / 12AX7 Brimar NOS flying leads 7025 GE ribbed plates 1950s * 7025 GE ribbed plates 1960s M8137/CV4004/12AX7 Mullard Box Plates * 12AX7 Raytheon black ribbed plates square getter halo 1950s * JRC-12AX7 RCA Black Plates 1954 * 12AX7A RCA gray ribbed plates 1960's-1970's * ECC83/12AX7 Siemens long plates early 1960s * 12AX7 Sylvania Gray Plate square getter halo 1958, 1959 *

12.6% 2.7% 0.8% 4.4% 6.1% 7.7% 4.5% 9.8% 2.2% 3.7% 11.2% 5.4% 4.9% 12.6% 9.9% 8.2%

102% 94% 100% 112% 109% 94% 92% 108% 100% 100% 92% 94% 101% 93% 95% 102%

1610 1680 1606 1640 1580 1690 1570 1740 1610 1590 1640 1520 1590 1660 1620 1710

1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600

98/100 105/100 100/100 105/100 101/100 92/100 91/100 112/100 99/100 102/100 92/100 100/100 100/100 96/100 104/100 101/100

12AX7R3 Electro Harmonix (10/7/02) 12AX7R3 Electro Harmonix (10/29/02) 12AX7R3 Electro Harmonix (12/12/02) 12AX7C - Chinese (10/28/02) 12AX7C - Chinese (12/13/02) ECC83 JJ (10/07/02) ECC83 JJ (11/07/02) ECC83 JJ (12/13/02) 12AX7R - Sovtek 12AX7WA 10/7/02 12AX7R - Sovtek 12AX7WA - (10/28/02 batch) (note a) 12AX7R - Sovtek 12AX7WA - (12/13/02 batch) 12AX7R ­ Sovtek 12AX7WA (5/16/03) - This batch was stronger than the prior groups, but was more inconsistent. There was generally a very large spread between the A and B sides of most of these dual triodes so these are NOT recommended for phase inverter use. Average output ­ 1.1 mA (closer to spec of 1.2 and better than the past) 91% of expected output. Good by today's new tube

25.0% 16.7% 16.7% 25.0% 16.7% 33.0% 66.7% 58.3% 41.7% 41.7% 41.7%

95% 85% 83% 83% 92% 93% 113% 112% 78% 88% 78%

1507 1382 1403 1461 1588 1467 1664 1604 1190 1293 1180

94.19% 86.38% 87.69% 91.31% 99.25% 91.69% 104.00% 100.25% 74.38% 80.81% 73.75%

0.0162 0.0161 0.0160 0.0170

#DIV/0! 85.31 87.14 91.31 93.41 #DIV/0! 85.33 84.87 #DIV/0! 87.96 86.13

0.0195 0.0189

0.0147 0.0137


production from all vendors, and better than their past production of this tube. Output ranged from 67% of what was expected as spec to 125% of what is expected. Very inconsistent current output. Average gain ­ 82.99 (spec 100) Good by today's production standards in new made tubes. Average TC ­ 1278 (1600 spec) ­ slow rise times, not all that great for high articulation players such as metal folks or folks that play harmonics. Lack of harmonic content and detail. Tolerance range ­ 58% Still not good at all. There is a lack of consistency from one tube to the next, near the bottom of the batch of vendors for the most part. More noisy than past production ­ due to more proper output and gain. 50.0% 12AX7R2 Sovtek LPS (10/7 batch) 12AX7R2 Sovtek LPS (10/23 batch) 12AX7R2 Sovtek LPS (10/31 batch) 12AX7R2 Sovtek LPS (12/12 batch) 7025 ­ Ei (10/31 batch) 7025 ­ Ei (12/12 batch) 7025 ­ Ei (4/03 batch) Output = 1.1 / TC = 1490 / Gain = 93.1 / QV = 33% - Nice gain this run, current way up too, great batch. Svetlana 12AX7 4/8/03 JJ 12AT7 / ECC81 4/7/03 - output = 8.9/ 10.0 gain = 58.8 / 60.5 67.0% 75.0% 42.0% 33.3% 25.0% 83% 70% 83% 97% 87.50% 46.67% 1469 1418 1505 1557 1419 1064 91.81% 88.63% 94.06% 97.31% 88.69% 66.50% 0.0182 0.0186 0.0155 0.0118 #DIV/0! #DIV/0! 82.69 83.71 91.55 90.17






TC = 4711 / 5500


Additional Small Signal Tube Tests and Reference:

Tube Tolerance Ave output % of mA expected Ave TC (gm) TC % Average gp Average Gain

Guitar Amplifier Blueprinting

1600 Reference small signal tubes - in some cases, single tube tests. Items marked * were three to five tube test averages

GE 6072A 1970s Date Code * GE 6072A Black Plate 1963 * GE 12AX7WA (1980s) * GE 5751 (1950s) JAN/Philips 12AT7WC 1980s * RCA 12AX7 NOS 1950s-60s RCA 5751 Black Plate (1950s) * RCA 7025/12AX7 NOS 1960s * Telefunken ECC83/12AX7 - diamond base * RCA 12AU7 NOS Cleartop * 12AX7 Telefunken smooth plate new Dynaco 12AX7/ECC83 Telefunken Diamond Bottom Ribbed Plates * 12AX7 MINIWATT / SUPER RADIOTRON AUSTRALIA 1960s 12AX7 Amperex Bugle Boy Holland LONG PLATE 1950s * (10) 12AX7 AMPEREX BUGLE BOY HOLLAND 1967 * 7025 Amperex Holland orange Globe Logo 1971 CV4004 Brimar 1961 CV4035 / 12AX7 Brimar NOS flying leads 7025 GE ribbed plates 1950s * 7025 GE ribbed plates 1960s M8137/CV4004/12AX7 Mullard Box Plates * 12AX7 Raytheon black ribbed plates square getter halo 1950s * JRC-12AX7 RCA Black Plates 1954 * 12AX7A RCA gray ribbed plates 1960's-1970's * ECC83/12AX7 Siemens long plates early 1960s * 12AX7 Sylvania Gray Plate square getter halo 1958, 1959 * Tesla ECC802S Tesla ECC82

2.6% 1.9% 8.7% 2.5% 11.6% 8.8% 3.7% 6.6% 11.6% 2.1% 12.6% 2.7% 0.8% 4.4% 6.1% 7.7% 4.5% 9.8% 2.2% 3.7% 11.2% 5.4% 4.9% 12.6% 9.9% 8.2%

96% 100% 92% 98% 105% 99% 96% 97% 102% 96% 102% 94% 100% 112% 109% 94% 92% 108% 100% 100% 92% 94% 101% 93% 95% 102% 10.9 of 10.6 11.1 of 10.6

1710 1750 1520 1190 5780 1620 1251 1580 1680 2180 1610 1680 1606 1640 1580 1690 1570 1740 1610 1590 1640 1520 1590 1660 1620 1710 2153 2601

1750 1750 1600 1200 5500 1600 1200 1600 1600 2200 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 1600 2200 2200

42/44 44/44 98/100 69/70 72/70 100/100 69/70 97/100 102/100 18/18 98/100 105/100 100/100 105/100 101/100 92/100 91/100 112/100 99/100 102/100 92/100 100/100 100/100 96/100 104/100 101/100 16 / 17 18.8 / 17 14

12AX7WA General Electric JAN NOS Green Printing 12/85 batch. 50 samples tested 5/21/03:

All very quiet. Some popping on turn on, but after the tube warms up this stops. More gain than current production 12AX7 tubes. A touch more output than industry spec, far above current production 12AX7s from any current factory. A very nice tube, in a high percentage of samples, very close matches on two sides. Test Specs: <- Output average compared to spec tube Average Output/TC 1.2 1434 102.50% <- Range above of spec tube highest tube High mA % 1.4 116.67% <- Range below spec tube lowest tube Low mA % 1.1 91.67% Average gp 0.0154 Average gain 93.12 <- Lower number is best Tolerance Range 25.0% 25% As of June 2003, this will be the new format that will be published in this document on tube tests from vendors.


Biasing output tubes is critical to great tone and long tube life. Some folks feel that setting the bias so the tubes run very hot, will make the amp sound better or be more powerful. This is not the case. Think of the bias much like setting the idle on your car. You can set the idle to 4000 rpm, and the car might run fine at high speeds, but mid and low speed drive ability will suffer. You want an amp to sound great at all levels, don't you? Bias may be set by a few methods. One is the crossover notch method. This requires a signal generator, load, and scope. This method is not repeatable, as it is the technician's "take" on the waveform that is being judged. The second and most popular method used today, and by most amp makers, is the current draw method. Here, an adapter is placed between the output tube and it's socket, converting the current to an output voltage in a 1 to 1 ratio. You set a conventional DVM meter to the 200 millivolt scale, and read your current directly ­ one milliamp = one millivolt. In a typical Fender type amp using 6L6 output tubes, this will be around 30 milliamps or so, depending on plate voltage. These tools are sold under many names such as bias tool, bias probe, bias rite, bias king, etc. Some have meters attached, some do not. I personally prefer the tool with no meter attached as it is less complex, less costly, and takes less space in my toolbox. After all, I already have a DVM that I carry anyway, and need one to check other voltages. Adjusting bias on most amps is fast and easy, and these tools quickly pay for themselves the first two or three times used. An example of the Groove Tubes tools can be found, with or without a separate meter on the website page at Many folks do not know what the maximum output of a given tube may be. In fact, some tubes of the same family are very different. An example of this is the EL-34. Most EL-34 tubes have a stated maximum plate dissipation of 25 watts. In the case of the Groove Tubes E34LS tube, this is different. This is a 30 watt tube, quite a bit stronger. This is a proprietary tube design made in the JJ factory on GT tooling. Looking at the JJ version and the GT version side by side, you will see heat sink wings on the GT version ­ more plate mass. Tossing a set of these E34LS tubes into an amp that is no PROPERLY biased, is asking for trouble, and almost guaranteed. When I say PROPERLY biased, I am not only speaking of the current draw, but the method used to set up the amp initially for a tube change. Just plugging the E34LS in and "chasing" bias is a mistake. In the time it takes to get the bias in the proper range, you may blow fuses, ruin the tubes, or take many hours off their life. 15

The proper way to do this, is first install your bias probe on one of the old tubes, and set it's level to the lowest possible milliamp draw. You may also just take your meter on the DC scale (another reason for a separate meter), and set this to the HIGHEST negative voltage. This means that ­52 volts is the more preferred value than ­49 volts as an example. Now, you can remove the old tubes with the amp off, install the new ones, and work up to the proper idle dissipation target. 6L6 and 5881 tubes come in ranges from 19 watts to 30 watts. A KT-66 may have the same rating and be interchangeable with a 6L6, but in the same circuit, it may draw 10 more milliamps at load, so it's bias may be very different than a 6L6. Different 6550 tubes and KT88 tubes may also vary. If you have something like Groove Tubes, you will not have to rebias if you use the SAME TUBE OF THE SAME RATING after properly setting the bias the first time. If the tube type changes though, a rebias is advised. "Same tube" here means just that. If you had a 6L6 Svetlana, and you changed to a 6L6 JJ tube, these are NOT the same, even though both 6L6 tubes. Same tube means the same family, type, and manufacturer. Cathode biased amps (Class A) are thought of as self biasing, and this is accurate to a certain degree. The amp maker is expecting a tube within the original spec range to be used. Today, this varies more widely than ever before, and if a tube outside the "normal range" is used, the amp may run very hot, with tube life short at best, bad and harsh tone, and at worst, tube and output transformer failure. Using a tube with a mid range rating number, such as GT's 4-7 ratings, will be the best choice in most cases. You can also use a bias tool (there are 9 pin adapters for EL-84 tubes) and check your dissipation too. Remember, in Class A amps, the tubes run close to or at 100% idle dissipation. If a tube has a maximum plate dissipation of 12 watts, such as an EL-84, a bit of math will be needed to be done to see what milliamp value should be as the maximum. If this is too high, just go to a lower range tube. It is that easy. As an example, if you have a #7 tube in an amp and it is running too hot (high millamps value), drop to a 5 or 6 and the value will drop into a more proper range. This is also a cool trick on fixed bias class A/B amps such as Hiwatt and Mesa Boogie amps. Mesa amps like mid range tubes in the 4-7 range. Elsewhere in this document is a conversion scale to covert the Mesa and Fender color codes to rating numbers. Mesa amps typically are OVERBIASED, meaning they run cold. This is part of the reason that many feel they sound a bit grainy. On my own Mark 1, the stock idle with Mesa mid color (range) tubes is only 38%. What we are looking for generally is 50-70%. Using a GT #7 tube in the amp brings this into proper range. Cold running (overbiased) amps sound grainy, lack power, and definition. Underbiased amps (hot running) amps sound harsh, do not perform as well at low and medium levels, and have shorter tube life. Maximum plate dissipation on tubes will be found in tube data sheets or in manuals such as the RCA tube manuals. As a quick point of reference on a few common tubes: EL-84 6V6GT 5881 EL-34 / KT77 6L6 E34LS KT-88 6550 12 watts 14 watts 19-30 watts depending on type 25 watts 25-30 watts depending on type 30 watts 35-42 watts depending on type 35-44 watts depending on type


The Output Tubes ­ The Pentodes and Beam Pentodes

A typical pentode trace. Many people believe that a 6L6 is a 6L6, an EL-34 is an EL-34, etc. This is far from the case. Each of these tubes have different sonic features. Many folks ask the difference between an EL-34 and 6L6. In a nutshell, I guess one can say that the EL-34 has more of an articulate mid-high and high end articulate response. If you like a lot of high end harmonic articulation and distortion with a high degree of articulation, than perhaps this would be the preferred tube. In the case of the 6L6, there are stronger lows and mid-lows, and for mass overall output, this may be the taste for you. Some folks love the 6550 and KT-88 tubes in amps if they get most of their tone and distortion from pedals or effects. These tubes yield a lot of clean headroom. Many ask me which is "more powerful" between the EL-34 and 6L6. Some feel the EL-34 is stronger, due to the 100 watt Marshall or Hiwatt type amps of the past. This is not the case in reality. These amps at times had higher plate voltages, so at times produced more wattage for a given tube type or tube set. A decade or so ago, Fender had an amp called the "75" that produced about 75 watts out of a duet of 6L6's, as one example. There will be charts as we progress, that show the same types of tubes used in the same circuits using the same plate voltages and bias voltages. The output is in milliamps. A tube with a higher reading, is more powerful than a tube with a lower reading. It is as simple as that. This does not mean a higher number is "better" ... there are many other aspects regarding tone. This is just a simple power measurement for the folks asking me which tube is stronger. This does not indicate or mean that a higher current tube is better in any way. This is simply a basic power output scale. There are many other qualities that are the aspects of the tone and sonic signature of an output tube.


The complete list of Groove Tubes output tubes can be seen at

The links below on each tube when shown, will be for a duet, a matched pair of two tubes. The link just above has the complete list, where one can find duets, quads, and single tubes for use in single ended class A amplifiers as one example.

6V6 Types

Some of the classic RCA 6V6 tubes 6V6R - a newer Russian 6V6 that is from the same tooling as the Electro Harmonix 6V6. This tube holds up well to plate voltages of 450+ volts, and will do very well in amps such as the Fender Deluxe Reverb.

6V6C - a new 6V6 off new tooling that is holding up to higher plate voltages nicely and has a sound more folks are starting to prefer over the 6V6R as time passes during 2002-2003.


Comparison of new and NOS tube tubes in a mid rating (bear in mind, the "NOS spec" number we are looking for here is 45.0 milliamps below for a 6V6


6V6 ( #6 rating) 6V6A USA NOS RCA Blackplate 6V6GT Tung Sol 6V6 Chinese 6V6R Russian EH

Output in Milliamps

43.7 45.4 47.2 51.9

6L6 / 5881 / KT-66 Type


The shorter bottle (but very strong) 5881 6L6-CB - Chinese, softer vacuum, coke bottle shape. Quicker to distort, with a warm soft tone. A great blues tube for smaller venues when you want to tone it down a bit. In a 50 watt Fender type amp, in a Groove Tubes 1-3 rating, the typical Super Reverb will start to break nicely at "3", with a humbucker guitar at "10". This is the most common supplied 6L6 in most production amps today. This is stock in all of Fender's 6L6 based amps with the exceptions being their Custom Shop amps such as the 2003 made Vibro King and Vibroverb amps. Vibro King amps prior to 2003 were shipped with this tube, New amps are shipped with the 6L6GE.

6L6-B (also known as the R) - Russia, Reflector, the Sovtek 5881WXT, Fender 6L6GC. Sturdy tube, one of the physically most robust. Great for touring stuff that gets thrashed by heavy handed road crews. This is the stock tube in currently manufactured Fender mid and lower priced amps. A very nice general purpose tube.


6L6-S - from the JJ factory but re-based and re-pinned by Groove Tubes if or when they come in with stamped pins with dipped pins that tear up some sockets. The rolled polished pins are also tapered. A bit of an extended midrange that a lot of heavy rockers love. This tube is favored by some Fender folks looking for a bit of an edge over the stock tubes, or in a Blackface Twin Reverb, Bassman, Bandmaster, or Showman, will give a bit more aggressive nature. With ECC83s in V1 and V2 of these amps, they will come a bit closer to a British sort of edge or character.

6L6-R2 - The Svetlana 6L6. Prior to 2002, this was a great tube, something of a Sylvania STR-387 copy. Due to the manufacturing changes, possible lower vacuum and material inconsistencies, these are not the same as the tubes in the past. You need to get these from a vendor that will test for low vacuum, gas leakage, and grid leakage. This is costly and time consuming, so expect vendors that have the equipment to test for these things to charge a premium.

5881-A - NOS, Tung-Sol, JAN Philips 6L6WGB, or other NOS tubes at times, great in Fender Black Face amps for a lot of my Blueprinting clients. The tube used a lot in Black Face Fender amps as original. These amps also used the Sylvania STR-387 and GE. These may not be carried in the future as the new 6L6GE is preferred.

KT-66-C - China, the KT-66 that folks like Valve Art and others sell as the KT-66 at times. Many times the Valve Art folks and others sell the Sino co-op tube which is not as high of quality as the other Chinese tube from the main factory in China. Good to about 450 volts B+ maximum. The KT-66 is a full bodied tube, with stronger and more linear mids than most 6L6 tubes. The GEC (UK) KT-66 was used in the Marshall JTM-45, and the first 100 watt Marshall amps. It's tone can be heard by Eric Clapton on the John Mayall Blues Breakers recordings. This is the tube that is standard issue in the Carr Amplifiers Mercury model

KT-66-HP - Russia, built under contract for GT to GT specs on Groove Tubes tooling. A copy of the GEC tube. Handles higher plate voltages than the KT-66C. Great to 525 volts, and the tube the Dr. Z Route 66 was designed around. The tube to be used in Marshall JTM-45. Also a great tube in Fender Black Face and Silver Face amps. Th "KT" meant "Kinkless Tetrode, and this tube was developed to get rid of the 6L6 kink in the response curve. This was accomplished, and this is why the mids of this tube are smoother and more linear than a 6L6 or 5881. This tube is standard issue in the Dr. Z Route 66 amplifier.


6L6-GE - made in the USA by the Groove Tubes factory with USA parts and labor. This was the tube used in the four Fender Showman heads at the 1968 Hollywood Bowl concert by Jimi Hendrix. (not always a Marshall guy). This tube has received a lot of praise in various reviews. One original set of these from February 2002, I have had running at 105% output, 24 hours a day, seven days a week. They currently have over 6700 hours on them, and they have only dropped 3-4 milliamps during this period. This is due to a very high vacuum, and USA parts, cathode coatings, etc. These are more expensive than most non USA 6L6's due to materials and labor costs being much higher. Their cost is offset with a very long life, and sonic qualities. They typically have at least a 15 degree wider sound stage image in any amplifier that uses 6L6 tubes. The traces on this tube are duplicates of the traces of the original General Electric tubes. This is not surprising, as these are made on the original tooling from the last plant that made these in the USA, and they use the same formula for plate materials and coatings. The mica spacers come from the original source. I coined these tubes "NVM", for New Vintage Manufacture", before I came to Groove Tubes while I was doing the original run testing. This tube is the standard offering in Fender's Custom Shop amps such as the Vibro King and new (2003) Vibroverb.

GE6CA7 ­ A strong beam pentode with an active beam forming element. Some folks think of this as an EL-34, but this is a stronger tube in all aspects. This is a drop in replacement for an EL-34, but a rebias is advised. This tube has a much higher vacuum than an EL-34, and much stronger construction. It is currently being developed for re-release as the Groove Tubes 6L6GE had been released. Look for a possible late summer 2003 release. The initial tests on this tube show very strong results. These will run at over 800 plate volts. There are current numerous "6CA7" tubes being made today, which are little more than the same plates and materials of a makers EL-34 in a larger bottle. In most cases, these do not even have the performance of the same companies EL-34 offerings. These do not trace or perform in any way as the original USA GE 6CA7 of the past.

EL-34-C - Chinese EL-34 (25 watt tube). This latest ChineseEL-34 has closer response curves and performance of the Siemens EL-34's used in the late 60's and early to mid 70's Marshall amps.


EL-34-R - Reflector (Russian) EL-34 (25 watt tube) sometimes sold as the EH tube. Pretty linear and Mullard like in some aspects of its character. Not as close to the Siemens tube in Marshall amps or in the curve traces. The current Chinese offerings are more close to the original Siemens tubes in a Marshall amp.

EL-34-R2 - Svetlana (Russian) 25 watt tube, used a lot as standard in Rivera amps. This is a stronger tube than the R or older C in the pre 2002 factory runs. The 2002 factory runs were down about 15% on power. Most current EL-34 types show better curves and more linear response than the 20022003 versions of this tube due to changes at the factory. If you have a vendor that can test for low vacuum and gas leakage, then this is a super tube.

E34L-S - JJ factory but with a GT developed heat sink assembly extruded onto the plate assembly. Hold up a JJ and a GT and you can see the difference without any sort of equipment (with the naked eye its easy). This is a 30 watt tube and bias should be checked and adjusted when one uses these. Very high output and strong mids. Used by Gibbons, Perry, and Walsh to name a few. This is the stock EL-34 amp tube for Bad Cat and Matchless. The JJ version of this tube is a 25 watt tube. This tube cannot be dropped in to an amp without a rebias. It is a much stronger tube. In a typical Marshall 50 watt amp, seeing 70-80 watts is not uncommon. This tube has much stronger mids and lows than the "normal" EL-34. If you are looking for that gut punch power, than this may be the tube for you. If you want your British amp to have it's original sound and character, then stay with one of the 25 watt EL-34 tubes.


6L6 / KT66 / 5881 / EL-34 Tube output current output comparisons

The "numbers" we are looking for here, for an average NOS sample run on a 6L6, is 72.0 milliamps as an example. In the case of the GE's the factory originals ran from 72-78 milliamps for the average sample tube as per the original GE spec.

If you want to know which tube is the most powerful in a given circuit, this chart below will answer that question. COMMON FAMILY (#5 RATING) 6L6 / EL-34 / KT-66 / 5881 6550 / KT-88 For 6L6 types, the standard expected reference output at 1957 RCA test settings should be 72.0 milliamps. CURRENT IN mA

The EL-34 tubes tested here were tested at 6L6 voltages and bias settings to show their performance in the same circuit. In an circuit set up to test EL-34 tubes specifically, the expected output would be 100 mA, higher due to lower bias voltage, and higher plate voltages.

7591 Russian 6L6GCMSTR Ruby Tubes (older) 6L6C Chinese 6L6B Sovtek (5881 WXT) 6L6CB Chinese Coke Bottle 6L6R2 Svetlana October 2002 6L6GE GT - Original run 6L6R2 Svetlana pre 2002 5881 JAN Philips 5881 Chinese 4/03

58.4 68.1 69.0 71.0 71.6 72.3 72.7 73.5 74.8 75.2 24

5881 NOS Tung Sol 6L6GE GT November 02 6L6GCR Shuguang 5/03 JJ 7027 6L6GCM STR Ruby April 03 KT-66 Genelex GEC 6L6S JJ KT-66 Genelex GEC KT-66HP Russia EL34R2 Svetlana (pre 02 date) EL34R EH / Reflector EL-34 Siemens marked Mesa EL34C China KT-66 China EL-34 Mullard 1970's E34Ls JJ / GT tooling 30W GE NOS Ref 6CA7 7591 Russian 6L6GCMSTR Ruby Tubes 6L6C Chinese 6L6B Sovtek (5881 WXT) 6L6CB Chinese Coke Bottle 6L6R2 Svetlana October 2002 6L6GE GT - Original run 6L6R2 Svetlana pre 2002 5881 JAN Philips 5881 Chinese 4/03 5881 NOS Tung Sol 6L6GE GT November 02 JJ 7027 6L6GCM STR Ruby April 03 KT-66 Genelex GEC 6L6S JJ KT-66 Genelex GEC KT-66HP Russia EL34R2 Svetlana (pre 02 date) EL34R EH / Reflector EL-34 Siemens marked Mesa EL34C China

76.1 76.2 76.6 78.3 79.4 80.4 81.7 83.7 85.3 86.3 89.0 89.8 91.0 91.6 93.1 97.1 108.2 58.4 68.1 69.0 71.0 71.6 72.3 72.7 73.5 74.8 75.2 76.1 76.2 78.3 79.4 80.4 81.7 83.7 85.3 86.3 89.0 89.8 91.0 25

KT-66 China EL-34 Mullard 1970's E34Ls JJ / GT tooling 30W GE NOS Ref 6CA7

91.6 93.1 97.1 108.2

A big KT-90


Philips NOS 6L6GC

RCA 7027A NOS 27

6550 / KT-88 Tubes

6550A - USA NOS GE no longer in stock. The strongest of all the 6550's. This is the longest lasting of any of the 6550 family.

6550-C - Chinese and a fun tube, as you can get these to distort in amps where the 6550A would not. These are in two types, the coke bottle shape (distorts faster), and the straight bottle with more power. Typically used in amps where a lot of power or clean headroom is desired.

6550-R - Russian 6550, takes longer to distort than the Chinese, maybe somewhere between the A and the current straight bottle C.

KT-88-C2 - Chinese KT-88 ­ Nice and strong This tube finds a home in amps such as the Park 75, Marshall Major, and 6550 type amplifiers.

KT-88-SV - JJ factory made off Groove Tubes tooling. Different plate assembly than the JJ version, with a large heat sinks welded to the plate. It may have a different base, pins, at times if necessary. There is also a KT-88 that is a JJ tube that is stronger by a wide margin than the original Gold Lion GEC KT-88. It biases differently. The KT-88SV (think of the "V" as "vintage", where the output of this tube and design was redone to copy the GEC Gold Lion. The SV traces as the original GEC tube traced and is great in McIntosh amps, Park 75's, and Marshall Major amps.

Power comparison chart of current output for the KT-88 / 6550 families

KT88 / 6550 / KT90 family 6550/KT-88 bias raised to CURRENT reduce current. Much additional IN mA power and clean headroom over 6L6 / EL34 if power supply has the capacity. At 6L6/EL34 voltages, typically 120-140mA + #5 rating


KT88 Svetlana 4/8/03 KT88 Svetlana 4/8/03 6550R Svetlana KT-88SV JJ / GT KT-88C China 6550C China KT-90 Ei

85.0 87.0 92.5 97.3 103.6 107.3 114.8

The EL-84 Tubes

EL-84R - Reflector Russia (as the Sovtek). Reliable, but not as articulate as the others. Darker sounding in most amps. The stock tube in most Fender, Vox, and many other EL-84 based amps. Not as bright as the EL-84-S, and not as articulate. If you liked the sound of your amp with these tubes it, then you may want to experiment with some others too, for a different tone which some prefer.

EL84-Y - Ei ­ Brighter than the S or R version. These can be unreliable in some amps if they are not checked for low vacuum or gas issues. Low vacuum will cause short tube life, and flat out failure at times, usually at what may seem the worst time. Many vendors sell these ­ be sure your vendor tests for gas leakage, grid leakage, low vacuum, and has a good warranty. A good tested set of these sound great though, so it may be worth the attempt to experiment.

EL-84S - JJ - The best of the current EL-84's by most folks thinking for higher power, more articulate mids and highs than the EL-84R. These are reliable, powerful and articulate. They are not as long lived as the "R", and not as reliable, but that is usually due to their moderate inconsistency. Buy these from a vendor that does proper testing.


Output power chart for EL-84 family

The "stock" NOS number here is 45-48 for reference. TUBE EL-84 EL-84 Sovtek EL-84S JJ Philips EL-84 General Electric 6BQ5 EL-84Y Ei Current in Milliamps 44.6 45.3 46.5 48.0 53.0

7027 and 7591 Tubes

7027 ­ JJ ­ Used in amps in the past such as Ampeg.

7591 - Reflector, Russia. ­ Used in amps in the past such as Ampeg.


Rectifiers ­ vacuum tube and solid state

Many people have many opinions on rectifiers in vacuum tube guitar amplifiers. Some people feel that solid state rectifiers are more reliable. This is true in one sense, but gives vacuum tube rectifiers an "implied" unreliability. I found vacuum tube rectifiers rarely fail. If they do fail, it is usually a case of physical damage, or the rectifier being bad in the first place, and failing in newly made amplifiers where the rectifier failed due to bad manufacturing. In the case of the Chinese rectifiers, when new, I have seen a 50% failure rate. A bad rectifier will usually show itself in one of two ways. (1) power light comes on, but NO sound at all comes from the amp. (2) when you flip the amp off standby and into play when all seemed fine, your fuse will blow. There are rectifiers with 4 pins and 5 pins, of the same type. The four pin rectifier may be indirectly heated, and not work well in some amp designs, so if a 5 pin rectifier comes out of your amp, put a 5 pin rectifier back in your amp. The rectifier basically converts the AC line current coming into your amplifier to the DC voltages that are needed. A solid state rectifier is easier to fit into an amp design, can provide more power, does not run as hot, and is less expensive than its tube counterparts. The Marshall JTM-45 used a tube rectifier, but when Marshall came out with the 50 watt version of the amp, a solid state rectifier was a change in the design. Most Fender amps over 40 watts use a solid state rectifier. If you have an older amp with a tube rectifier and want to replace it with a solid state replacement, be sure to first check your amp and make sure all the other portions of the amp are in good condition. The solid state rectifier is capable of higher voltages. As you will see in the charts, different rectifiers have different characteristics. In many amps, you can get a different sound and feel by replacing the rectifier with different types. Matchless is one example that lets the user choose various tube rectifiers as part of their features. A solid state rectifier will give very fast rise time and response as the voltages are produced very quickly. A vacuum tube rectifier will yield more to the player's touch dynamics, sound warmer and less harsh by some folks feelings, and give the compression and sustain in a much different way than its solid state brother. When one initially hits a loud note or chord, with a tube rectifier, there is voltage sag, in some cases, a LOT of sag. As the note or chord starts to decay, the voltage then builds, and what you have in essence, is a built in compressor / sustain device. If you look at the charts, you will see how fast the voltage is developed with a solid state rectifier versus a tube rectifier. Using the same circuit, we replace the rectifier section. We used the same voltage input in all cases, but in the case of the 5Y3GT rectifier in the last test, our transformer output of 333v had to be reduced to 330v, as the 5Y3GT would have been pushed just a touch beyond its design limits. This change had negligible results on the final outcome. The tests were measured over a period of 500 milliseconds, or ½ a second.


Using the commonly used configuration of 1N4007 diodes, its easy to see in comparison with the vacuum tube rectifiers, the difference in rise time to get the voltage we are after. It is a bit less apparent, but in this case with the solid state rectifier, we had 449.08 volts available for our B+ voltage, the highest voltage in the group. This is also something to keep in mind when you replace a tube rectifier with a solid state replacement, as your output tube bias will probably need to be checked. In the case of a class A amplifier such as a Vox AC-15, AC-30, a Matchless amp, most Carr amps, or others, the bias is automatically taken care of, but in older class A amps, you may also want to be concerned about the output transformers and capacitors if they are older. The amp will be running at higher voltages, and in some cases, much higher voltages as you will see later. In .01 seconds, with the solid state rectifier, we basically had full power. Our voltage was 449.08 as previously stated. Moving along, we substituted a 5AR4 vacuum tube rectifier. This is one of the strongest vacuum tube units used today. It took about two times as long, .02 seconds, for out maximum voltage to be developed, 413.9 volts. This is a fairly large drop from the earlier solid state rectifier. Going to a 5U4-G, commonly used in some Mesa Rectifier products, we found that it took .40 seconds to develop full voltage. This is a long time compared to the solid state device, and twice as long as the 5AR4. This is almost a half a second, and most people can hear something that is ½ a second long in duration or delay. Our high voltage was 357.06 volts, or almost 100 volts less than the solid state device. The 5R4-GYB, often used in place of the 5U4-G. About the same rise time as the 5U4-G above, but had a maximum voltage of 330.93 volts. This is a nice rectifier in some Fender Deluxe type amps that have been made with everything from the little 5Y3 to a GZ34 in the past. Its a great blues rectifier in these amps. Lastly we tried the 5Y3GT. Rise time was almost a full half a second at .44 seconds. Our maximum voltage was 303 volts. Now you have a little more information on rectifiers, and how the changing of types in a rectifier equipped amplifier can change the sound of the amp, the power of the amp, and the characteristics of touch, feel, and sustain. In most cases, the NOS rectifiers are better than those of today. The Chinese ones still have a long way to go to reach the specs of those of the past. The Russian ones are the best of the new ones, and the JJ factory is currently working on a re-release of the GZ34 which shows promise. Mesa Boogie products currently use mostly the Chinese variety. As a safety note - you can generally go "down" but not always "up". This means you can generally put a 5U4 or 5Y3 in an amp that had a 5AR4, but it is not recommended to put a 5AR4 or 5U4 in an amp that had a stock 5Y3. The increase in voltages may be too much for the other components in the amplifier. Rectifier substations / cross reference: GZ34 = 5AR4 GZ32 = 5U4 = 5V4GA GZ31 = 5U4GB GZ30 = 5Y3GT

Rectifier charts below ­ These are newer charts than those on my website from a later series of tests. The voltages may not correspond with the data in the text above, as the power transformer was changed slightly. In any case, the rise times and maximum voltages reached can be compared between the various rectifier types just as easily. 32

5AR4 / GZ34 above ­ 416 volts max.

5U4G above ­ 364 volts max.


5U4GB above ­ 362 volts max.

5V4G above ­ 380 volts max ­ Great in Dr. Z, Matchless, Bad Cat


5Y3GT above ­ 322 volts max.

6AX4GTB above ­ 387 volts max


The GT Rating system ­ an explanation of "distortion rating" also known at times as "hardness rating"

Basically, a #1 will distort sooner, and a #10 later. If, for example, with a mid range tube, say a #5, makes your amp start to break in the output section at a volume setting on the amp of "4", then with a lower number tube, like a #2, your amp would have a same sort of break into output distortion at say a volume setting of "3". With a higher tube, such as an #8, then you amp would stay clean to about perhaps "6" on the volume. High rating numbers are not more or less powerful, they just distort later. These are preferred by heavy rocker that want maximum clean output, as they get their distortion and tone from effects or pedals. These are not as touch dynamic. Low number tubes are very touch dynamic, and more suited for a lot of folks, for smaller venues and recording. These ratings are especially great in amps used to play blues or other touch sensitive styles. Most folks prefer the 4-7 range tubes, as they are the closest in character and touch to what the amplifier designer had in mind. They are also the most versatile.

Conversion Information

Mesa to GT numbering system:

Mesa vs Groove Tubes scale Red 4 Yellow 4 Green 5 Gray Blue White 5 6 6

Fender to GT numbering system: Fender Blue = GT 1-3 Fender White = GT 4-7 Fender Red = GT 8-10



A vacuum tube is very complex. It looks to be a very simple device, but is anything but simple. In your amp, you may have +/- 20% resistors, and other items of wide tolerances, but vacuum tubes can, and will vary, much more. A typical preamp tube such as the common 12AX7 / ECC83 / 7025 will have many construction variations. There are long plates, smooth plates, box plates, rib plates .... Long and short cathodes, differences in construction. These aspects, and many more, lend to changes in tone and response. Even tubes of the SAME MAKE, SAME TYPE, and SAME MAKER from the same batch, will vary widely. A typical 12AX7R2 as made in the Reflector Factory and distributed under the Sovtek name, will have Transconductance ranging from 900 to 2000, as one example. The target spec is 1600, As you can see from some simple math, this is WAY outside of the +/- 20% range of your resistors or other components. Output in milliamps, expected to be 1.2 milliamps at 250 plate volts and ­1.2 volt bias, in these tubes shows 0.6mA to 1.6mA. The average is about 0.8 (about 30% less than expectation spec by the way). Gain (not the same as output), and plate resistance also vary widely. This is one reason that when you retube your amp with fresh and new tubes, at times you may find, the amp has lost it's life. Why? Your old V1 may have been putting out 0.9 milliamps in current ... your new tube has 0.6 milliamps, 30% less current output than your "old" tube. What can you do about this? Your best option is to use a tube vendor that tests ALL preamp tubes for things like low output, low noise, and lack of microphonics. Many rely on just sending all their tubes out to the customers, relying on a warranty to take care of the testing for them. In most cases, it is not worth the trouble for the customer to send back one single tube, box, ship, and the post office line ...they just forget it. Why are tubes so variable? First off, there is a LOT of hand assembly. This is not like making silicon devices on AMD or INTEL production equipment. There are also a LOT of materials, and the ratio of many of these materials is very critical. Change a plate material or cathode coating, and you may have changes. Move things around in production and you will have changes. There are many aspects that can change tubes, even the weather during production. These little glass bottles are generally not made in a silicon clean room environment. What sorts of materials can change? How can it be complicated ... some metal and glass? Sure, we know that metallurgy is an exact science, and when it comes to airplanes or high tech items, this is paramount. But, in vacuum tubes, it generally does not share the same attention ­ one more variable. The tubes use a number of materials. Some of them are mish metal, nigrosine, porcelain, petroleum jelly, zinc, calcium aluminium flouride, resin (synthetic), ethyl alcohol, lead acetate, malachite green, glycerine, zinc chloride, iron, marble dust, wood fiber, strontium nitrate, lead oxide, zinc oxide, lava, mica, tin, sodium carbonate, sodium nitrate, silver oxide, barium carbonate, calcium carbonate, arsenic trioxide, ammonium chloride, strontium carbonate, potassium carbonate, isolanlite, bakelite, molybdenum, phosphorus, alumina, silicon, borax, shellac, tungsten, barium, copper, titanium, carbon, silica, glass, chromium, magnesia, clacium, platinum, caesium, cobalt, strontium, iridium, magnesium, monel, rosin, nickle, calcium oxide, cobalt oxide, barium nitrate, thorium nitrate. Actually a somewhat complex recipe! After all these ingredients are "mixed", we still have all the variability of assembly and manufacture. Remember when I said that a lot of this is done by hand? Even the machine aspects are pretty tight in the tolerance department, so the machines must be kept within close tolerances, which is not always the case due to wear or lack of attention to maintenance.


How close are these tolerances? Lets look at some of them in a typical output tube where there is more room to work than in a preamp tube, so this is "easy". Grids ­ the diameter of the grid has to be 0.001 inch. Cathode sleeve wall ­ 0.002 inches thick. Air pressure ­ 1/100,000,000th of sea level. The glass bulb ­ inspected under polarized light for stains. The Plate ­ diameter gauged to 0.002 inches. Cathode coating ­ weight variation less than 0.0007 ounce. Grid wire ­ diameter cannot vary more than 0.00009 inch. Heater wire ­ diameter cannot vary more than 00002 inch. So, now we have the basic "parts" from the most easily recognized major aspects of a tube, yet we still are looking at putting it all together, welding or attaching it all somehow, and expecting it all to not only work, but work in an expected manner. It is pretty amazing, that today, these things even work at all!


Matched Phase Inverters and Drivers

An example of an "almost matched" dual triode. In the longest leg of this trace, you will notice at one voltage level range at the top, there is one trace higher than the one attached to it. Many folks match for gain, some for output, some for transconductance, and none that I know of match for rise time. This would be considered a pretty amazing tube. Most ( 45+ our of 50) have the two traces not even close in any single aspect of it's specs, let alone close in all aspects. This takes a LOT of time to match a tube in this way, but a lot of folks can sure hear the results in great amps. In amps with all sort of mismatched aspects, there is a lot of work that needs to be done before the total benefits of a phase inverter even this closely matched can be fully appreciated. Why do two Marshalls or two Fenders, of the same model, and even year, sound different? For one thing, there was a +/- 20% or more variance in components used, but a more common follows.


I originally wrote this with a subject of dead spots in your sound or tone that people felt were due to dead spots in their guitar (or bass) neck. In the last few months while blueprinting amplifiers, I have had to explain over and over about a mis-matched output section and its impact to many. I thought it would be a good idea to get these thoughts down on paper. In most class A/B amplifiers, there is an NFB (negative feedback loop). This is usually labeled as the PRESENCE control). Any disparity between the upper part of the sine wave (produced by half of the output tubes), and the lower part of the sine wave (produced by the other half of the output tubes), is canceled out by the NFB circuit by design. This is the reason some notes "sing" when your amp is pushed in the output section (rather than pushing the input in a master volume amp), and other notes do not have the same magic. Since few tubes are even close to identical, this cancellation is always ongoing. The object is to limit this as much as possible. The most common way people match an output section, is to use good quality matched tubes. The industrial spec for a match can be as high as +/- 20%. A good match by a lot of tube vendors is +/- 10%. I believe that even the untrained ear can hear the difference when a output section matched within +/5% is used. In the amps I set up for the folks that retain me, my spec is less than 2.5%. The most overlooked and misunderstood part of the output section is the 12AX7/ECC83 (Marshall style) or 12AT7 (Fender style in vintage cases) Phase Inverter tube. This is the tube that drives the output tubes. A lot of folks that specialize in making amps sound great don't understand this, but fix this accidentally. They tend to use very good tubes, such as JAN spec 5751's etc., where the match is closer, and closer matched tubes in the output section. They also use tubes that sound good in the first gain stage positions, rather than the common Sovtek WA tubes which most manufacturers use (because they are sturdy, not as expensive, and ship well without developing microphonics). When I scope an amp in the lower frequency region, the vast majority of the time, the upper and lower parts of the sine wave are not equal. This is more "off" than just a slightly mismatched set of output tubes. At this point, I install a matched phase inverter / driver. The problem with phase inverters, is finding a matched tube. You have to remember that a 12AX7 / 12AT7 etc., is NOT a single function tube as an output tube. It is TWO tubes (two triodes), sharing a single bottle. VERY FEW TUBE COMPANIES MATCH THE A AND B SIDES OF PREAMP TUBES. They warrant the tubes to work, and warrant them not to be microphonic, but do not say they are matched. This is not any bad commentary on tube suppliers. This is very expensive, and not deemed as cost effective to some. In the high end hi-fi and audio industry, matched triodes are a basic requirement, even if "low-fi" guitar and bass amp players think this is not necessary. Matching is time consuming and requires specialized equipment. If you can find somebody that has a Tektronix tube curve tracer, and bring them a bunch of tubes, maybe you will be lucky and find a match. Watford Valves in the UK calls these tubes "balanced valves". Groove Tubes sells these as their "MPI" tubes and they are also spec'd for rise time and compression on a Vacuum Tube Curve Tracer in addition to matching output and the rise time. These are what I use for my clients, as it is one thing to hit both side of the output section with the same "push", but to me it is also important that the "push" comes at the same time. Matched phase inverters and output tubes are one of the reasons some amps "sing" and others are pedestrian compared to their brothers and sisters. If you seem to have a lot of dead spots, try a new phase inverter tube. This is usually the preamp tube that is the closest to your output tubes. It is a trial and error process, but you may get lucky. By the way ... THIS IS EXTREMELY CRITICAL WITH DROPPED TUNINGS, 7 STRING GUITARS WHEN IN THE LOWER RANGES, AND EVEN MORE CRITICAL WITH BASS GUITARS WHEN USING TUBE AMPLIFICATION. 40

Preamp Tubes - Gain / Output and Matching ­ more information

I have written other papers on each of these subjects, and questions and comments come up where additional explanation has been helpful for a lot of folks. The first basic point to remember here, is that unlike a power tube such as a 6L6, EL-84, EL-34, 6550, and others, a preamp tube is a "dual triode" in most cases. This includes tubes such as a 12AX7, ECC83, 7025, 12AT7 and others. In regard to matching of two sides of a preamp, many folks feel that in a balanced circuit which in a way, "sums" the two sides of the tube, that balancing is not necessary. The high end audio industry recognizes the need for a balanced phase inverter or drivers, unlike some in the musical instrument industry. This is necessary (the balance of two sides of a triode). I will use an example to try to illustrate. Taking a twin engine airplane, let's say we have a typical light twin with two 300 HP engines designed to cruise a 250 knots. We have each engine at 1800 rpm. One engine is at 24" of manifold pressure, while the other is at 22" of MP. The airplane structure itself, is the "balance", much like the balanced circuit that some amp builders feel will negate the need for a balanced phase inverter. The plane may fly just fine, but we need to add a bit of rudder trim, and our fuel use will be higher, and overall performance and balance of the airplane (or amp) will suffer. Now in the "art" of balancing a dual triode, there are many folks that will perform "matching" for a few dollars of additional cost. In almost all cases, this "match" is a gain match, or a current match, or a transconductance match. Due to very high inconsistencies in today's preamp tubes, matching for any of these factors, is an improvement over an unknown tube. This sort of matching at least makes sure that we do not have a 200 HP engine on one side of our airplane and a 400 HP engine on the other side. These numbers seem like wide examples? Not in the least. A typical new preamp tube these days at a given bias and voltage, in the 12AX7 family, will run from ½ a milliamp to three times that. They typical average is about 70% of what is expected as standard spec by the way. Now, in our airplane example, our horsepower may be way off with two different HP engines in the stock amp, and with the easiest form of balance as above, at least our HP is the same, which is of help. But, we still have the factor that one of the engines is running at a different RPM or Manifold Pressure. This is the "time" component that is missed in most matching. You can balance the circuit all you want, from an amp designers standpoint, to compensate for voltages or current, but you cannot balance the time component, all one can do is average the current or gain factors, not the time factor. A true balanced dual triode, is just like two output tubes. We want their characteristics balanced in many aspects. This is why we do not plug in one EL34 and one 6L6, even though this would work, and make some sound come out of the amp. In fact, in this example, these two totally different tubes would be closer in characteristics than the typical new dual triode of today in many cases. What is required here is to select for output, and TIME. This can only be done on very sophisticated equipment such as a vacuum tube curve tracer. The two curves, as all voltages of operation, and with a signal applied to the tube, are compared and sought to be as close to identical as can be achieved. This is very costly as it is very time consuming. At times, only 1 in 50 tubes will make the grade. Remember, it is THIS little tube that dives your final output stage.


The folks in the high end audio industry know the difference in balanced or matched inverters, and many amp makers do also. There are still a lot of "amp" folks out there that want to fight the points here, mostly because they have not taken the time component into their thinking. If they have tried "matched" phase inverters from simple sources with simple current match methods, this may be one reason that there was not as much difference as they had expected. The Special Applications Group (SAG) at GT was formed specifically to address issues such as vacuum tube development, testing of factory samples and production, and unique products. The SAG-AX7-MPI and SAG-AT7-MPI are just two of their products for the audio industry.

Part 2 - Gain Versus Output in preamp tubes.

Think of your days back in science class, where you built or saw one of those big ball devices, that created a half a million volts, made a great light show, and you could touch it, and have the classes hair stand on end! Gain in a guitar amp is much the same. You can have a 20 watt amp with high gain, and shred all day long, just as "gainy" as a 100 watt monster. It is the wattage, or output, that differ, and what that "power" brings to the stage. Today's 12AX7's as an example, all have about the same gain, which in this case should be about 100. Most today fall below this, in the range of maybe the mid 80's, with some samples going up to maybe 110. The big difference in tubes though is current output. A typical 12AX7 is expected to put out 1.2 milliamps at a given test voltage. Today's tubes in general, put out as little as ½ of that in 80% of the cases. A tube at 0.8 milliamps has a full 30% less output than what is expected. This is like a 50 watt amp putting out 30 or so watts. Some tubes are better than others from various manufacturers. Some examples of this are tubes like the widely used, and perhaps the most popular are the Sovtek 12AX7WA. This is a sturdy tube, generally free of microphonics, with acceptable gain. Part of the reason they are quiet, is they tend to be lower in gain than many other 12AX7's, but are also much lower on output current. It is much the same as with power tubes; where in the same amp, one duet of output tubes will put out 50 watts, while another set only 45 watts. Thus, the Sovtek 12AX7WA, is quiet, due to lack of output and gain in many amp circuits. These tubes are sturdy and inexpensive, and help a lot of amps make it through the warranty period. On the other side of the coin are the JJ ECC83 tube. This is a part of the 12AX7 family, but different construction, plate materials, cathode coatings, and other factors give this tube a bit more gain than most others. This can be 100-120 Mu, rather than the 85 or so of a 12AX7WA Sovtek. The big difference in the ECC83 in general though, is it's current output, as times over 1.5 milliamps, or in some cases, three times MORE than a Sovtek12AX7WA. Like I have said in the past, preamp tubes are a crapshoot. You buy your tubes and take your chances. Some folks like Groove Tubes, screen and test for microphonics, noise, AND low output. The ones that do not pass all tests are rejected, in some cases, over 50% of the factory run. There is still a range of specs even in these more rigidly tested tubes, but the spread is much tighter. In any of these cases, the end user still does not know what the tube is actually doing. The SAG area at GT takes tubes, and runs them through another process, where all the specs are recorded and traces are performed. These traces show rise time, and other factors. Like the SAG-MPI's, there are other "kits" such as the MHG (Marshall High Gain) kits, which can be used in any 12AX7 based amp, not just Marshalls. I guess I should have called this a High Power kit, which would have been more accurate actually. There are Fender High Gain Kits, and Fender Soft Touch Kits, and the SAG generally tailors preamp tubes for specific uses and tastes. In any case, don't confuse gain and output. They are very different qualities in a tube.


Power / wattage ­ how much do you really need?

The vast majority of the time when I first walk into a venue where I will be listening to music for the night, I can generally tell if the performance will be a memorable one. I can generally tell from the equipment setup, and not to brag, I have about a 90% track record. The 10% of the time I am mistaken, it is generally easy to explain. I did not know the performer or group, and what to expect. I was invited as a guest, and the music was already known to me as to not be of my particular taste, or a few other reasons. The big tip off, is amplifier compliment ­ amplifier power that is. There are folks that have blinding fast technique. Speed metal players, fast articulate players, folks with speed as their underlying goal. Frankly, this is not my personal taste. I am generally impressed for ten minutes, but then my attention is lost. I generally ask myself, are these folks practicing, or just looking for the right note? One note played with feeling and that has tone, is worth 100 64th note triplets from my point of taste. But, this is personal preference, and I have a lot of clients that are speed metal masters, that have the ability to play most anything they want, much better than I can play my own preferred style. These folks have unique requirements, and generally a lot of power is wanted, for reserve clean headroom. Most of their sound and tone comes from front end effects and/or pedals. The biggest problem form most cases, is amplifier power. When I see a 100 watt amp on the stage of a 150 seat venue, I know that I am in for trouble ... most of the time. If it is a jazz player looking for a clean sound, then I am safe. If it is a speed player, well, then it is what I expect. If it is a rock or blues band, then I know I am in for a very one dimensional performance most of the time. I know with good prospects, a few other things. The player does not understand amps or tone, perhaps his main rig is broken and this 100+ watt amp was borrowed, or they are into a hi-fi sort of sound with little or no dynamics. A 100 watt amp, or even a 50 watt amp, will not distort in it's output section at rational volume levels. Folks that are known for great tone and to be great players, even in the largest venues, generally stay at around 50 watts or less. They are looking for a particular sound, tone and feel. They let the stadium sound systems do the rest. If you cannot turn your amp, and most amps, to at least 6 or so on the volume, you will never tap the soul of most tube amps. Folks also do not understand "loudness". Many think a 100 watt amp is TWICE as loud as a 50 watt amp. This is not all the case. Double your wattage, and all you gain is 3db. Sure, folks talk about "headroom", and think this is a huge requirement. Folks that actually need headroom are clean players ... rhythm players, jazz players for some styles, and pedal steel players to name a few. A amp with a lot of headroom is a hi-fi amp. It will be clean, and not have the dimension of touch dynamics of a lower powered amp. If you want more loudness out of a 50 watt amp, double your speaker area, or go with a more efficient speaker. Going from a 83db speaker to an 88db speaker, is almost the same gain in volume as going from 50 watts to 200 watts in amp power. Then there are those folks that have the great idea of pulling two tubes out of their 100 watt amp to turn it into a 50 watt amp. To put it bluntly and open myself up to a log of flack - this is a stupid idea. A great amp is made up of many components. Power transformers, output transformers, capacitors, and other parts, make up the design. If one takes a Marshall 100 watt Super Lead, and pulls two of the tubes, and properly sets the impedance selector, turning the amp into a 50 watt amp, what actually happens? Well, we have a 100 watt power supply, that is now even less taxed than before. The "50 watt" Marshall will now have less dynamics, less feel, less touch sensitivity. It will be a nice, clean, hi-fi, 50 watt amp. It's 100 watt power supply will never reach saturation. It's output transformer will never be pushed. It will actually be cleaner than it was as a 100 watt amp. The only distortion you will get is when the output tubes are at their limit, and this will be an unbalanced sound, although some might think this is just to their own tastes.


Modeling amps? In the past, I have written a lot on solid state amps versus tube amps, and modeling amps versus the amp originals they are modeling. Modeling amps have some strong points such as a lot of sounds for the dollar or in a given space, or for recording. Some think in a live venue, modeling amps can have limitations. To my way of thinking, this live aspect is sort of a "yes and no" to me, just as with any amp, tube, SS, or modeling amp. I see many folks with 50 watt amps in small clubs, where the soul of the amp is never tapped. Put an amp like a Line 6 Vetta, Fender Cyber Deluxe or Cyber Twin, or Vox modeling amp out there, and you may be surprised. These amps may be just the ticket in any size venue, and in small venues where maximum overdrive is sought at rational levels, may be a super option. Recently on a five day cruise with my family, there was one of the entertainers who covered just about every music style that was ever done, solo. He had a small rack with a Panasonic laptop. This held his song list, and was on a music stand close at hand. It had a sound card of some sort, and internal drum machine, sequencer, and midi outs to a small synth rack. This was all fed into a stereo SS rig, used for PA speakers, and just there for clean sound. His Guitar amp was a Line 6 Vetta. He was a hit on the ship, and had more folks in one bar area, than most of the "conventional" bands on the ship. He was a fun act to see. His rig was light, portable, and very versatile. In some cases, a modeling amp may be the ONLY proper choice. Reflecting back on what I wrote on that cruise incident when I first produced this document, I have had additional thoughts brought on by questions from many folks. This was my response to one forum post on the subject which I wrote: There are some places where tube amps are totally unacceptable. Recently I was on a ocean cruise with my kids. The crew on these ships work seven days a week for at least a six month tour. Same for the entertainers. The ship gets back on a Friday morning, and leaves late that afternoon. No time to take an amp to a tech as one part of the situation. Add to this, the players may be playing for the show in the big showroom one night, rock in a club the next, and jazz the next night. The amps are moved all over the place, and having an amp that does a number of things rather than excel at one thing, is really a prime consideration. In this case, something like a Vetta, or even the Spyder, would be a killer amp to have. A Super Reverb, Mesa, or Marshall would be about the worst thing one could have. It boils down to the right tool for the job. I heard a few modeling amps on the cruise, and frankly, they were killer. The worst amp I heard was a tweed Fender Bass Breaker, whose tubes were shot, bias was off, and had at least one noisy and microphonic tube in the first gain stage. I was so bad I had to leave that rock lounge called something like China Town on Carnival to Ensenada (over Christmas) if anybody else that reads this was there. Thank the tone gods for a fellow with a smaller Marshall Valvestate and another fellow in a Reggae band with a Line 6 Axis. They sounded great, and the guests loved them. I personally prefer the PROPER wattage tube amp as first choice, but I will take the modeling amp every time over the wrong tube amp. Why? Modeling amps allow a degree of touch dynamics and tonal ranges to be captured at most any level. You have all sorts of controls for this ability. A Fender Tweed Bassman in a small venue will never be able to be cranked to it's level of tone potential for some music styles. A Line 6 Vetta may pull off the "tweed sound" of the virtual Bassman in a much more convincing and pleasing manner, at least to my tastes. I think to end this, all I can suggest is, listen to amps, and play them. See how they react to your touch. If this is not a part of your music and style, such as many folks that start the song at 110db and end it at 100db, then most any amp will work pretty well. As you develop an ear for different tone aspects, and fingers and touch that can give you at least two more playing dimensions, then you will move to the next step of being a better player, and also have a more heightened ability as a listener. 44

........... From Mark Baier at Victoria Amplifier

As an added treat, I will include some additional NOS data for those interested. This was written by Mark Baier of Victoria Amplifier Company ( ). Mark is known as something of a "tone god", and is also on the advisory board of Tone Quest ( ). I have a fairly large Victoria area on my own website that has additional information and photos on many Victoria amps at

Thoughts on 12AX7 type tubes by Mark Baier

Keep in mind that he wrote this back in 1998, and a lot has changed since then on some fronts. Mark is generally pretty darn busy building amps, and I am generally pretty busy testing each new factory batch of tubes that comes out. For the latest 12AX7 information you can click here ... or .... for information on preamp tubes in general and matched phase inverters click here. (Myles comment here) Howdy folks, I'm back with more wisdom for the ages for ya. As I sit here in front of my keyboard trying to conjure up something semi-literate to say, I must confess to having a new appreciation for writers and daily journalists...I suppose it helps to have at least SOME idea of what to write about...How about the vacuum tubes used in guitar amps??? Kind of a no brainier type of subject given what this page is dedicated to...N'est pas?? I'm always a bit surprised by how little the average guitar plinker knows about the glass bottles that power their amps. Sure, much has been written by self proclaimed gurus as to the physics of vacuum tubes, or how to turn a BF Twin into a '51 Super, but how much discussion is ever given to the SONIC differences between a GE and Sovtek 12AX7? I'm gonna invent new and exciting sonic adjectives for the occasion! You'll want to rip your hair out trying to decipher what "chewy and refined with a touch of lacy brilliance" actually means!! When I'm done writing this, I'm gonna compose my monthly article for Wine Spectator... At this point it's probably a good idea to note that I did not conduct an extensive, dedicated listening test using a specific model amp and guitar. These opinions are based on years of playing many guitars and amps of various manufacturers and vintages..The focus of this assessment will be with a '50's Fender Tweed type amp in mind. Aside from being my favorite kind of amp to play tone wise, they make a good platform for evaluating tubes due to their simple circuit topologies. Preamp Tubes: GE 12AX7: Most commonly available as JAN type/mid '80's vintage. Has a soft top end-not edgy or piercing. Nice solid mids and lows. Lots of depth and texture to the tone. Older orange box versions seem to be a bit more brilliant. Can be a noisy tube, this would be a problem for high gain amps. Can sound grainy in some amps. Moderate amount of gain. Good NOS American tone...for years these were standard issue in your Fender amp. If you're using the recent JAN stuff, pack some spares; these were an end run situation, QC was a bit slack. Standard Phase inverter tube in a new Victoria. ECG 12AX7: Like the GE mentioned above, this tube is most commonly found as JAN examples. With the wild west capitalism that is the rule in the tube biz these days, these JAN stocks are the most reliable source of actual NOS US and British tubes. These tubes are visually identical to the '60's and '70's Sylvanias. Like a vintage Sylvania, they display a lot of gain and midrange/top end brilliance. They sound great in a cathode follower stage. Real creamy when driven. A preamp tube that can cut through the mix. Not as textured as an '60's RCA, but possessing a more authoritative body. Can be a


problem tube; prone to microphonic behavior and fizzy static that comes and goes at will.. A nice vocal tube that will color a soundstage with its usage. We use them as a standard tube in our 3X10 & 4X10 amps to give some bite to the darker Mojo 10" Alnico speakers we use. RCA 12AX7/7025: Through the years, RCA produced a few different types of 12AX7's. Typical late 50's types would have a large ribbed plate that is dark gray/black in color. Branded 12AX7 and 12AX7A. This vintage is rich and appealing sounding with a very musical midrange bloom. Round, warm bass with a compelling, lacy top end (don't say I didn't warn you). Very similar in construction and tone to contemporaneous Tung-Sols. Existing examples are increasingly rare and unpredictable in performance. Can be a noisy tube. You know that 'noise' test point on your TV-7/DU?? It was put there because of this tube. Expect to screen this tube before using; average samples will frustrate you with uneven performance. Early-mid '60's-70's examples of the RCA 12AX7 are my favorite in this family of tubes. They have a shorter ribbed plate structure and the actual plate is lighter in color. Very broad, even response. No frequency group is accentuated..everything sounds even and in perfect harmony. Most true NOS examples exhibit goodly amounts of gain though not as forward as Sylvanias. Particularly fine with Fender guitars. Bouncy and expressive. Perhaps the best sounding hi-fi tube around, right up there with Telefukens, expect to pay up for real, tested, NOS examples. I dearly love these in our 80212's. They tickle the ceramic magnet 12's we use beautifully. 7025 versions are screened, tested versions that were specified by many manufacturers because of their low noise, audio intention. Physical characteristics of this tube are virtually identical to the classic Mullard type. I've seen Mullard, Brimar, Amperex and GEC's that appear to be identical to the RCA 7025. Most European examples have a seam at the top of the envelope. Performance is interchangeable as well. I suspect that the lauded Mullard "M" series of tubes are hyper tested examples of this classic RCA structural platform. Telefuken ECC83/12AX7: Known through 2 versions, ribbed and flat plate. Has achieved mythical status, certainly perceived as the benchmark 12AX7 by hi fi nuts, and with good reason. Even scope pulls can test way above minimum and still sound terrific..It's not unreasonable for this tube to last for more than 100,000 hours!! Lots of OOOph when used in a guitar and 3 dimensionally complex. Note decay is very musically textured and fine..Breaks up with an encompassing, balanced presentation. Transparent sounding, giving a broad, even response. Richly toned without sounding muggy. Lots of air and space, a delicate authority is achieved. Very commonly found in old hi-fi sets and industrial test gear. Not as common in American guitar amps like Fender, but seen quite often in off brands like Guild and Silvertone. Often rebranded with the amp manufacturers name. These work great in higher gain amps like Marshall where noise would preclude the use of a more "colorful" tube. EI ECC83/12AX7: A new Eastern European type currently being produced in Serbia. This tube shows lots of promise. Extremely similar to the Telefuken in construction. Can be discerned by the seam on the top of the envelope; the Tele doesn't have this. Sonically, a bit more colorful than the Tele. Can be edgy and harsh sounding. very forward midrange. Terrific tube in higher gain rock amps like a Matchless. Must be carefully selected for noise and microphonics. Aggressive sounding with a definite bite. Not a subtle, refined tube. Expect to reject 50% for noisy operation. Not as musical as a GE or RCA, but still very good. Many are being rebranded by fast-buck hustlers as it. Sovtek 12AX7: Available in a number of versions. Early types are 12AX7WA and WB. Recent examples labeled WXT and WXT+. Higher distortion and gain than RCA and Mullard types. This characteristic endears them to users looking for a distorted, aggressive attack. Not as quiet as the Chinese 12AX7, but possessing a lower noise floor than common US made GE and ECG types. Very reliable long lasting tube. Lacks real voice and character when utilized in old style Fender Tweed and BF era amps. Somewhat lifeless and flat sounding in amps where tube disposition can be discerned. Great tube for amps that derive their tones through circuitry manipulation, i.e. multi-stage cascading gain amps, built in effects amps, "lead" amps with buttons and knobs that say things like 'pull thick', 'drive', 'solo' and 'thrash'.. Not microphonic or particularly noisy. Really a plug it in and forget about it tube..Good for guys and gals who don't want to be bothered by fiddling around endlessly with their tubes... Shuguang 12AX7: Chinese made, current production in limbo. Last batch reputedly made in mid 1997. Very quiet, well made 12AX7. Reputedly made using Mullard or MOV equipment. Rich toned, if a bit lacking in low end and mids. Nice sense of space and air around notes. Fairly high gain, good for modern amps. Low noise floor good for critical high gain applications. Downside to this tube is its longevity. These things wear out quickly, getting smudgy and flat; as if someone threw a blanket over the tone. Easily recognizable by the shiny metal stiffeners between plate structures. Always house 46

branded, common guitar brands include Mesa, GT, Fender, and many others. Designer versions can get pricy. Recent versions of this tube include a shiny plate type dubbed a 7025. Some folks really swear by this variant. If cathode is the same as other Chinese versions, it's gonna have the same longevity problems. More musical than the Sovtek, but not as compelling and textured as a NOS GE or RCA. Very common as recent Fender and Mesa OEM type. GE 5751: Originally a 12AX7 with beefed up internal elements for operation in action man military and industrial applications. Five star versions are highly prized for their careful parameter control. Most examples in stock these days are the JAN stuff from the mid '80's. Very good side to side matching. Very tubey sounding tube. If you're trying soften up a harsh modern amps, this may be the "go to" tube. Don't over use it; too many 5751's in the tone zone will deaden things. Less presence and air to the top end. Nice and clear sounding--kind of a soft brilliance. I like them best as input amplifiers replacing he 12AY7 called for in '50's Fenders. ECG 5751: I don't have a lot of experience with this one..I would expect a greater reject rate than the GE's based on my experience with all the ECG stuff. RCA 5751: A very nice tube--very hard to find. Be careful these days when purchasing NOS examples. I suspect that RCA didn't make a whole lot of these, consequently, you risk buying used stuff unless you know what you're doing. When you do find them, you'll be treated to a rich, full sounding tube that's detailed and 3 dimensional. Very musical. The GE sounds thinner when compared directly..the RCA has more body and depth. This list hardly represents a complete assessment of every preamp tube available, but is intended to address to sonic characters of the most commonly found types. The basic structural platforms are the GE, ECG/Sylvania, RCA (two types), Telefuken, Reflector(Sovtek), and Shuguang. The 12AX7 type tubes available today will no doubt be an example of one of these. I hope my impressions of the differences will be of value to you and enlighten you to the tonal manipulation possible with this benchmark tube type. Myles S. Rose - Guitar Amplifier Blueprinting - 2003



47 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

C:\my files\amp basics\big file basics combines.wpd
Triple XXX om1