1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

For 6AS7G tube rollers here .....

257 258 259 260 261 262 263 264 265 266
268 269 270 271 272 273 274 275 276 277
  1. L0rdGwyn
    Here we go. Will post some impressions after some quality head time. TBH, I'm hoping the differences between the 421a and 5998 are not huge, otherwise my hoarder mind will require me to get a second as backup :L3000:

    Last edited: Mar 1, 2018
  2. dobigstuff

    You will love that Western Electric 421A. See below my combo in my Darkvoice that I use with my HD700's
    Western Electric 421 A.jpg
    L0rdGwyn likes this.
  3. L0rdGwyn
    Loving the 421A so far, it is very detailed, nuanced, great soundstage, more bass refinement than the 5998.

    I've made an interesting observation, curious to see if anyone else has had this experience. I have owned five different Tung Sol 5998's. I currently own three, sold two others not too long ago. Of those five, I have noticed one particular 5998 sounds significantly better than the rest. As I am sitting here comparing it to the 421A, many of the characteristics of this (I'm going to call it) super 5998 that make it better than the other two that I still own (as well as the two I sold) are present in the 421A, but in larger quantities.

    I had never looked before, but I compared the construction of my super 5998 to the other two, and there is one very noticeable difference: the two rods on either side of the heater filament are a much thicker, copper-colored metal on the super 5998, whereas the same rods on the non-super 5998's are shorter, thinner, and a nickel-colored metal. Interesting. Also, it is unrelated to year of construction; the super tube and one other are both manufactured in 1962, the third in 1954. I am no expert on tube construction/anatomy, I can post pictures if anyone is interested in any of this blabbering.

    I'll have to do some more comparisons between the super 5998 and 421A. I am listening through the Sennheiser HD 660S which has a 150 Ohm impedance, so slight advantage to the 421A given that it has a higher transconductance than the 5998 (18,000 vs 14,000, respectively) and thus will further lower the Crack's output impedance, which should favor using a lower impedance headphone like the 660S. I'll compare with a higher impedance headphone too to even the playing field, but I doubt the damping factor alone can explain the improved performance; it truly is a step up from the 5998, even my super tube (but the gap is much smaller) :)
    Last edited: Mar 1, 2018
    Monsterzero likes this.
  4. attmci
    Have you measured those 5998?

    Moreover, here are some old discussion about real super 5998s:

  5. L0rdGwyn
    In terms of emittance or THD? I don't own the equipment to do either LOL but all of the tubes were bought in NOS condition and TV7 tested (according to the sellers I bought from, of course). I personally have not put enough hours on them for low emittance to be a concern.

    Thanks for the link! I'll check out the discussion.
  6. attmci
  7. mordy
    If you want info on old guy radiola, check out the internet......

    Here is an offering of a 421A from a different seller:
    At least this one looks more authentic, but it is supposed to be just a regular 5998
    This Canadian seller has a couple of what looks like TS5998 labeled 6520. They don't measure NOS, but the prices so far are in the reasonable range.
    Here is a link to these 5998 offerings:

    After clicking on the above link, got to see other items to find the other 5998 offerings.
    Last edited: Mar 5, 2018
  8. mordy
  9. thecrow
    And this guy has 100% rating on ebay - f’n ridiculous

    Mu blood is starting to simmer
  10. Scutey
    I saw this on eBay as well, was tempted briefly, then decided something didn't ring true.
  11. attmci
    I know the story and had never interested in OGR.

    I have enough 5998/6AS7G tubes. You should pick those up. :) But re-branded tubes always sell at a discount.
  12. mordy
    Hi attmci,

    I have one pair of TS5998. Missed a rebranded pair for $35 because I did not know what it was... I am keeping on looking - part of the mystique of tubes is to look for bargains, at least for me lol.
  13. attmci
    Gm vs mu

    Mike McCarty provided some detailed description of Gm vs mu.

    When you buy a NOS, NIB, or used tube, always look at the Gm (i.e. a new Monconi spec: 7.5mA/V or 7500 µmhos) which indicates the health of the tube. Dying tube has low Gm, thus low mu.


    Gm vs mu
    by Mike McCarty
    Most tube testers just check emission. Some of them also check for some sort of gain; often a
    "transconductance" is measured. Most of them do *not* measure transconductance or Gm, but
    rather tube gain or mu. I've noticed some confusion here and there on the net, so I thought I'd
    explain the meanings of mu and Gm, and their relationship. Throughout this message, I use
    "voltage" as a synonym for "EMF".
    A tube is modeled as a voltage-controlled current source. That is, at any given plate voltage, the
    current in the plate (and cathode) is proportional to the voltage on the grid. There are several
    things wrong with this model, as with any mathematical model. It's simplified, and not terribly
    accurate, but nonetheless it is very useful due in part to its simplicity.
    Tubes are, of course, not linear, so exact proportionality is a fiction. But within bounds, this model
    gives a reasonable first approximation to what happens. A short enough piece of any smooth
    curve looks like a straight-line segment. It turns out that the curves associated with
    tubes are "straight enough" over about 2/3 of their reasonable operating conditions for straightline
    approximations to be useful there.
    So, what are the exact definitions? First, they do not involve DC or static conditions. These are for
    small value AC signals, or to put it another way, small modifications to the static bias
    conditions. Second, the exact mathematical definitions are in terms of things developed in
    calculus courses, called partial derivatives. If you haven't had calculus, don't worry. I'll give
    information, which doesn't involve it.
    Gm is the ratio between the plate current change and the grid to cathode voltage change with
    plate to cathode voltage held constant. If we call Ip the plate current, and Vg the grid voltage,
    Gm = dIp / dVg
    where you may read the "d" as "small change in". Another way to read it is as "The small AC
    plate current resulting from a small AC grid voltage, with the plate voltage held constant, or short
    circuited for AC" (like by bypassing the tube with a huge capacitor).
    If you know calculus, then it means the partial derivative of plate current with respect to grid
    voltage. Its unit is that of current divided by voltage, or the mho or Siemens, and it is a sort of
    Since this "conductance" is from the input to the output circuits of the tube, and goes "across" the
    tube, it is called "transconductance" or "mutual conductance". Since "G" is the symbol
    used in electronics for conductance, this explains the use of the symbol "Gm". It is not a real
    conductance in any sense of the term.
    In the fictional mathematical model, this is a single constant. With a real tube it depends on plate
    voltage, plate current, tube temperature, frequency of the signal, mood of the operator, etc.
    Now mu is defined similarly, but it is a ratio of voltages
    mu = dVp / dVg
    This is the AC voltage gain for small signals, being the AC voltage in the plate divided by the AC
    voltage on the grid, with the plate current held constant. It has no unit, being a pure number. In
    the fictional model, this also is a single fixed constant for any given tube. For the mathematically
    inclined, it is the partial derivative of the plate voltage with respect to the grid voltage.
    Now, also in the fictional model, there is an effective plate resistance associated with the tube,
    usually denoted by Rp. By definition, Rp = dVp / dIp, in other words the AC voltage on the plate
    divided by the AC current through the plate. Thus we have that
    mu = Gm x Rp
    That's why, in an earlier message, I said that Gm and mu are effectively the same. They are
    essentially proportional to each other. Well, not exactly of course. Gm is a conductance, and mu
    is a pure number. But in the fictional model Rp is *also* a constant (for any given tube, that is), so
    there you are.[1]
    Now, mu is easier to measure, at least approximately, so that is what "transconductance testers"
    usually measure. A small AC voltage is placed on the grid of an otherwise appropriately
    biased tube, and the AC voltage on the plate is displayed, perhaps divided by the (fixed) AC grid
    voltage, giving an approximation to mu.
    I have with me some data sheets for various tubes, which include the parameters for the fictional
    models, portions of which I reproduce here.
    I give Gm in umhos, Rp in K ohms. The values of Gm, Rp, and mu are from the data sheets,
    while I compute Gm x Rp myself. These are all twin triodes, and the second element of the tube
    is indicated with an asterisk ("*") after it.
    Tube Gm Rp mu Gm x Rp
    ---- -- -- -- -------
    12AX7 1250 80 100 100
    12AX7* 1600 62.5 100 100
    12AT7 4000 15 60 60
    12AT7* 5500 10.9 60 59.95
    12AU7A 3100 6.25 19.5 19.375
    12AU7A* 2200 7.7 17 16.94
    Actually, the data sheets simply divide mu by Gm to get Rp, and round off to a couple of figures.
    [1] The mathematically adept will note that the Gm and mu definitions require different static
    conditions on the other circuit parameters, and hence it cannot be the case that mu = Gm x Rp.
    But this whole paper is "to first order".
    Mike McCarty, VRPS member.
    Last edited: Mar 18, 2018
    Scutey likes this.
  14. zeroduke
    Last edited: Apr 2, 2018
  15. thecrow
257 258 259 260 261 262 263 264 265 266
268 269 270 271 272 273 274 275 276 277

Share This Page