Blackburn/Mullard EL34

Discussion in 'The Workbench' started by StingRay85, Apr 13, 2021.

  1. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    This mythical tube deserves its own topic, so I guess I'll start it here. Since the 60s there are five main types available. The Xf1, that stopped around 62, I will not discuss it further because surely Marshall has never used them. Then you have the Xf2, which can be distinguished by its welded plates, and double ring getter. Around the end of 68, they transitioned this Xf2 model to a single ring getter. The color of the base still changes in the next years from brown to black, and mid 73 the Xf3 model is being introduced, which has stapled plates instead of welded plates. The top mica spacer still has the 10 dents. Then around mid 75, the Xf4 model comes into play, which can be distinguished by only have 4 dents in the mica spacer. So by these differences, even without seeing any codes, you can easily identify the Xf model of the Blackburn EL34. The Xf codes were also used for older Philips tubes produced in the Netherlands, but they can be easily spotted by the light brown base, the double D getters or the presence of X or L codes. Then dating them becomes easy, after the "B" which stands for the Blackburn factory, the first digit represents the year. So you cannot make any mistakes, since there are no overlaps. B0E4 stands for 1970 on an Xf2, while it is 1980 on an Xf4, and also the last code of the batch of EL34 tubes ever produced in the Blackburn factory. The 2nd letter represents the month of production, A being January, L being December. The last number is the week in which it was produced, so ranging between 1 and 5.

    Having collected around 100 of those, 85% of them are used, I have already obtained a decent data set using the Maxi Matcher 2 tester. First of all, how can you see that you have a true NOS/NIB tube? It can be easily recognized by the getter flash, that is shiny and the barium metal deposit has a specific pattern. If you don't see this pattern, I wouldn't trust it's really a new one. Check the pictures in the next post.

    I use a specific system to grade my (used) tubes. First of all, how good does the getter flash look? It's really the best indication of how many hours the tube has burned. If you see it's almost gone, you can be sure it has seen thousands of hours. Also discoloration (black/brown), or rainbow-like colors are never a positive sign. If you see tubes advertised that don't clearly show you the condition of the flash, you better ask for that or assume the seller is hiding something. I've seen many advertised that are shown in a way that hides the true condition of the flash. But enough on that. This is the first sign you need to look for on a good tube.

    Secondly, how good is the tube still looking. Does it still have a good label? If you see brown or black discoloration around the plates, the tube has overheated. Sometimes you can also see how the heat left a mark on the glass at the position of the holes in the plate. Not so good either. How dirty is the base? If it's rough, but the glass looks nice and shiny without any label, you can be sure it has been sitting in an amp for a long time, and it was just cleaned until it shined again. A tube that sat in the original box still has a very smooth base. When the pin guide is missing, you can be sure some idiot pulled it out of an amp with no intention of recovering it, but then somebody found the tube and still thinks its worth to sell it. I have a handful of those, and some do measure ok, but surely this is a bit more risky to install, and value wise you surely shouldn't pay too much for it. Also corrosion on the guide pins is always a very bad sign. A tube that sat all his life in an amp or box typically still has nice and shiny pins, and even the NIB tube has markings below, from testing it back in the Blackburn factory, but also the tube tester of course.

    Third, my tube tester. It measures plate current at 400V -36V bias. The spec according to my tester is 20-50mA under those conditions. From all my NOS/NIB tubes, the lowest measures 20.9 mA (Xf2 from '62), the highest 44.5 mA (Xf4 from '79). I also have 22.2 30.9 25.9 27.6 28.4 42.7 40.3 33.7 37.2 31.2 39.7. This large spread was a surprise for me. From my used tubes that so aren't burned up, the lowest measured 8.3, an Xf2 from '63 which still had a decent getter flash, but clearly visible rainbow markings. Can it still be used by just biasing it correctly in the amp? If I find a tube that matches it in the amp, then I guess so. Next, it measures transconductance in mMhos. The limit for a good tube is 4 according to the tester, but on the first Maxi Matcher it was 5. This is also where I put the limit for a good tube. To obtain this value, it typically requires that the tube pulls at least 26 mA under the 400/-36 condition. The transconductance is depending on the current it pulls, so what I do is sort my best tubes on the amount of mA they pull, and then see what kind of transconductance value they have. If you have tube A and B that have a test result of 30 mA, and one tube measures 5.5, and the other 5.1 mMhos, you can be sure tube A is better than B. In almost all cases, the getter flash will tell you the same story. If the tubes measures below 25 mA, and the flash isn't looking too good, it's probably not a good tube anymore. Also, you can have a visibly worn tube that draws 40 mA with a 5.2 mMhos reading, while a NIB tube measures 27 mA and 5.1 mMhos, the second tube is the one you would rather have. So the tester is not always giving the full story. And this probably applies for any other tester.

    I buy my tubes when they are reasonably priced, without test results. A person who has a tester, typically asks for a lot more (I can't blame them), and is not necessarily honest on its performance. Looking at value, surely if you see welded plates (Xf2), they are worth more than stapled plates (Xf3-Xf4). If you can't see the plate construction, but you see a lot of dents in the top mica, then it's either a Xf2 or a Xf3 and they are worth more than an Xf4. If you see the typical 4 dents, then don't overpay for them. If I see the typical getter flash of a NOS/NIB tube, I'm prepared to pay more for it. Sometimes I get lucky with old amplifiers that still have the original tubes, but then your main focus should be on the condition of the flash. If you see its almost gone, then the tube is also almost gone. But if you see for example a small Geloso tube amp from the mid 60s where the flash on the tubes still look more then decent, it's probably not used much and the tubes (which will be Xf2 double ring) can still be in perfect shape.

    And finally, you have to go and make matched pairs from them. Having them pull the same mA in 400V -36V bias is not enough to match them, if your amp works on 450V plate voltage, the conditions are different from the tester and they can be drawn apart, even when they are reasonably close transconductance matched. Since there is one more important thing to check, how the tube sounds like, it's a good practice to look for the matching tube in the amp starting from the data of the available tubes. This way you can also hear that the tube still sounds the way it should. Like mentioned below, ultimately doing this on a cranked amp is best. For a bass amp this is of course easier to do with 50W than with a guitar amp. An attenuator is probably a good idea.

    So, I have around 50 tubes that meet all the criteria of a good tube (including my NIB tubes); and 50 that don't. The plan is to collect the good ones (e.g. Angus Young or Joe Bonamassa can always contact me), and use & abuse the less good ones (for example when some of the flash is already consumed). When the flash is really almost gone, why not bias it to (or over) it's limits, and use it for a recording amp until it goes out with a bang and a broken fuse.
     
    Last edited: Apr 13, 2021
    AndyD, playloud, Maggot Brain and 3 others like this.
  2. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    20210413_102750_copy_774x1032.jpg 20210413_102731_copy_1032x774.jpg

    So you can see left to right an Xf2 double ring getter, Xf2 single ring, Xf3 and Xf4. Notice the getter flash, and this dark thin transparent band below the shiny part. This is in my opinion the one and only true indication of a NOS/NIB tube. Maybe in the future I'll conduct a test and see how many hours it takes to disappear
     
    Last edited: Apr 13, 2021
    Fernieite, playloud, nickfox and 2 others like this.
  3. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    706
    Likes Received:
    684
    Great info, thanks!
    This thread should be made a sticky.
    I assume that the Maximatcher puts 400V on anode and screen grid?
    Yes, the state of an output valve’s getter flashing is a good indicator of its remaining life.
    I’ve got maybe half a dozen working Mullard pulls that I’ve been meaning to assess one day, and have a 2204 based build on the go to use them, so the above info will be handy.
    It needs to be bourne in mind that an output valve which idles and deals with small signals happily might still put on a spectacular lightning display under load with large signals, most especially when the application is a cranked guitar amp.
     
    Last edited: Apr 13, 2021
    StingRay85 likes this.
  4. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    Indeed, so in current mode, it measures idle current at those 400/-36 conditions. For transconductance, it measures how the tube amplifies an AC signal under 400/-36 conditions. The nice thing about the tester is that is also has a digital fuse to detect shorts/leakages and a tube that is drawing way too much current.
     
  5. shredless

    shredless Well-Known Member

    Joined:
    Sep 15, 2010
    Messages:
    5,546
    Likes Received:
    4,010
    I currently have a pair of xf2 in a 2203x...l dont remember the year. These tubes are my favorite lve ever used. Thanks for the great info!
     
    StingRay85 likes this.
  6. JP2036

    JP2036 Well-Known Member

    Joined:
    Aug 7, 2020
    Messages:
    568
    Likes Received:
    708
    [​IMG]
    [​IMG]
    [​IMG]

    I have a quad of Amperex labeled XF-2's i bought last year from a private seller, no testing data i took a chance. They test "good" on Hickock style testers but there is an anomaly while playing my 100 watt JMP every so often a large loud Buzz sound with come thru the amp and not in a good way.
    I won't run them anymore and im thinking its maybe one tube out of the 4 thats doing it.
     
    StingRay85 likes this.
  7. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    So those are Xf2 double ring getters, with the getter flash dark and almost consumed, and the glass not looking too clean. So heavily used tubes. Testing "good" on any kind of tube tester is really not something that should convince you to buy these kind of tubes.

    I hope you didn't pay too much for them.
     
  8. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    20210413_140932_copy_1238x1651.jpg 20210413_140948_copy_1238x1651.jpg

    What you see (the pictures are not super clear though) is

    1) a almost NOS/NIB flash starting from the left: 34.6/5.15 Xf2 from 66

    2) then one I consider "good": 37.5/5.63 Xf2 from 68

    3) in the middle "reasonably good" as you can see some of the flash is consumed (but they don't go to my "good stash" anymore regardless of how they measure). also notice the tube is not so clean, so surely I pulled this out of an amp at some point. 30.1/4.96 Xf2 from mid 73; so this one fails on two of my criteria although it still looks good

    4) next is showing signs of wear (flash already consumed for a great part, and you can see the presence of a rainbow especially on the first picture). it measured 26.6 mA, and 4.61 mMhos. 21.8 / 4.59 Xf3 from 74. According to the Maxi Matcher spec, this is still acceptable, but I really see no point of paying more than 20€/$ for it, although it probably still sounds better than a JJ.

    5) finally on the right you can see one where the getter flash is almost completely gone. Ironically, this is one of my first mullard tubes that I bought from a seller that mentioned it still measured good 26.6 /4.61 Xf2 from 67. These kind of tubes are one their way out, try to make it a spectacular ending.
     
    Last edited: Apr 13, 2021
    Pete Farrington likes this.
  9. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    706
    Likes Received:
    684
    So there’s some rainbowing around the edges of the getter flashing on at least some of those Philips Miniwatts in post #8, right?
     
  10. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    It is clear that tube nr.4 has it. You might think that 2 and 3 have it too, but that's really not the case but I agree it's hard to judge from the pictures. That's a normal edge around the flash.
     
  11. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    This is what I call a rainbow

    20210413_142957_copy_1238x1651.jpg
    This again the 2nd tube from post 8

    20210413_143050_copy_1238x1651.jpg
    20210413_143106_copy_1238x1651.jpg
    This edge is normal I think, and I wouldnt be surprised that is the transparent layer of flash present on the NOS/NIB tube that is consumed after the first hours or days of operation
     
  12. JP2036

    JP2036 Well-Known Member

    Joined:
    Aug 7, 2020
    Messages:
    568
    Likes Received:
    708
    Brent Jesse's webpage sez black getter flashes on Xf2's are common, they actually sound really good. At some point if i get a 50 watt JMP i may use them. Its hard enough to find decent CP power tubes IDK why i would expect these 55+ yr old EL-34's to be without flaws.
    CP Svetlana EL-34's work great for me anyway.
    I do use Mullard & Amperex in Pre slots exclusively with no issues.
     
  13. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    Well, they might be common because they have been passed on from amp to amp to the point that they are completely used up. Especially in the US where those tubes aren't so abundant. I can show you tens of Xf2 that have the normal flash like any other Xf type tube. A black flash is really not something you should accept as being normal for this type of tube.

    By the way he mentions the base as being dark brown, that's different than the getter flash of course
     
    JP2036 likes this.
  14. Pete Farrington

    Pete Farrington Well-Known Member

    Joined:
    Jan 18, 2021
    Messages:
    706
    Likes Received:
    684
    I strongly suggest to keep any output valves that are suspected to be liable to catastrophically fail well away from any amp with vintage transformers. It’s impossible to know how many prior such incidents they’ve been subjected to. As I see it, output valve shorts are potentially as stressful as a cranked amp used with dodgy speaker cable / impedance selector etc. By the time a fuse blows, the damage can be already done.
     
    Last edited: Apr 13, 2021
  15. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    Good point. But I see no direct correlation between a tube that seems close to being used up judged by the getter flash, and a tube that is at risk of shorting out. I have shorted out tubes that look perfect. In any case, I believe it's unfair that sellers still try to unload these tubes at 100 a piece.
     
    Pete Farrington likes this.
  16. JP2036

    JP2036 Well-Known Member

    Joined:
    Aug 7, 2020
    Messages:
    568
    Likes Received:
    708
    EL34 / 6CA7 TUBES IN STOCK - Brent Jessee's
    http://www.audiotubes.com › ...

    I have no reason to doubt what you're saying but all im saying is that i may be able to salvage a pair out of the quad anyway. If i remember right while biasing there was one tube that was off compared to the other 3. That may be the suspect microphonic EL-34.
    Again i never or hardly ever buy NOS/ANOS tubes w/o testing data. I took a chance and it didn't work out as id hoped although i had the quad in the JMP about a year and didn't have any issues til the 8 month mark i was gigging.
    And the watts output was lower than the Russian tubes.
    But the harmonic overtones were lovely
     
    Last edited: Apr 13, 2021
  17. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    The link doesn't work for me. Maybe I was not very clear. The flash of the tube gives an idea how long it has served already. But doesn't give you any clue on how good it will still sound, or how many hours it has left in it. Only overall speaking, you can claim the tubes with a perfect looking flash have percentage wise more life in them than tubes where the flash is almost gone. I would just not spend my money on the ones where the flash is almost gone, the chance of getting bad ones simply increases too much. I buy so many at a price sufficiently low, that I dont need test results anymore, I just match them myself
     
  18. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    This is the typical brown getter flash of the metal base Netherlands produced EL34. The one with the special foot is actually an EL60 with the octal foot, produced in '54. It measures 35mA / 5.4, so a strong tube, although it has obviously been used and is still a bit dirty (also the pins). The other metal base is from 55, looks new and unused, but only measures 9 mA on the tester.

    20210413_205901_copy_1238x1651.jpg 20210413_205844_copy_1238x1651.jpg
     
    StuC likes this.
  19. StuC

    StuC Active Member

    Joined:
    Feb 12, 2014
    Messages:
    116
    Likes Received:
    153
    I have some boxed xf2s from 1973. I don't think they've seen much use. I think have some metal base ones (used), will have to dig those out for a look.

    [​IMG]
     
    AndyD, Maggot Brain, nickfox and 2 others like this.
  20. StingRay85

    StingRay85 Well-Known Member

    Joined:
    Aug 18, 2019
    Messages:
    1,767
    Likes Received:
    1,361
    Yeah those look new to me. Nice!
     

Share This Page