Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com
Asher Guitars WD Music Products Amplified Parts Mod Kits DIY Nordstarnd Pickups Warmoth.com

Real tech info on 5751 needed

Discussion in 'Glowing Bottle Tube Amp Forum' started by ballynally2, Nov 29, 2017.

  1. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    Even order distortion, like this assymetrical output, tends to sound fuller and warmer long before it really becomes noticeable as crunch. Coupled with the lower gain, looks like the net result is a lower gain preamp with some nice subtle coloration using a 5751.

    If I get some free time later, I'll see what an idealized 5751 circuit would look like. That Fender 12AX7 preamp stage, dating back to the fifties, uses the "golden ratio" where the plate load is twice the anode resistance and bias is set to achieve a plate voltage of 2/3 the B+ supply.
     
    Old Tele man likes this.

  2. ballynally2

    ballynally2 Tele-Meister

    Age:
    53
    387
    Jun 7, 2009
    donegal ireland
    yep, that would be my guess
     

  3. Commodore 64

    Commodore 64 Friend of Leo's Ad Free Member

    Age:
    42
    Mar 1, 2010
    Kent, OH
    Other ways include:
    1. Attenuation resistor in series with the signal.
    2. Splitting the plate resistor into 47k/47k and tapping the signal off the junction (you can modify the split plate load ratios, too as needed)

    I don't concern myself too much about what a 12AY7/5751/12AX7 are doing in the circuit, they are ubiquitously accepted as circuit compatible. If it sounds good, it's good. If it changes the amp for the better, I don't care how the bias shift affects the symmetry.

    Also, both of the examples used with 2V input would be clipped prior to the max voltage swing on the graphs, so I'm not sure how much asymmetry we'd actually hear. My scoping, I've seen a max ~90V peak to peak on a 12AX7 gain stage with 100k resistors.
     
    Last edited: Nov 30, 2017

  4. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    I agree a lot with that answer. If it sounds good, it probably is good.

    And back to my little art exercise, it looks like just swapping the plate load from a 100k to an 80k makes a big difference in linearity, assuming a 300V B+ supply. The gain is ever so slightly lessened, but the output is much more symmetrical. But who's to say if that change is one we actually like, right? :)
     
    ballynally2 likes this.

  5. ballynally2

    ballynally2 Tele-Meister

    Age:
    53
    387
    Jun 7, 2009
    donegal ireland
    good points.i concur..
     

  6. ballynally2

    ballynally2 Tele-Meister

    Age:
    53
    387
    Jun 7, 2009
    donegal ireland
    yes, changing parameters to make things look better on a graph while liking the originals when listening makes little sense..
     

  7. Silverface

    Silverface Poster Extraordinaire Platinum Supporter

    Age:
    65
    Mar 2, 2003
    Lawndale CA
    I'm curious if anyone has actually checked the reality numbers for newly manufactured 12AX7's & 5751's and verified that they are in line with the specs? Are the specs being used the ones for those individual tube brands - or the RCA tube manual?
     
    telemnemonics likes this.

  8. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    For my part, I was using the JJ datasheet for both tubes. Hopefully their tubes are at least somewhat on the ballpark of that data.
     

  9. telemnemonics

    telemnemonics Poster Extraordinaire

    Age:
    58
    Mar 2, 2010
    Maine
    I have a bunch of 5751 and will swap them into a 12ax7 slot routinely when getting to know an amp, and they stay 1/3 to 1/2 the time.

    But one thing I noticed is that buying them from vendors who show test results on all their tubes is that for example RCA and IIRC GE black plate 5751 test way lower than GE early gray plate 5751.
    You see the same with any other 12ax7 brands tested by the same seller at the same time using the same calibrated tester. The number may not mean much but the numbers being consistently much higher for a batch of one brand NOS compared to a batch of another brand NOS does seem to me to mean something.

    There is/ was a test list by a new tube seller where IIRC actual measured gain numbers were shown for various brands and types, and again IIRC there are some 12ax7s with lower gain than some 5751s.

    I have to assume that all the design standard published specs can vary between brand and era.

    WRT the 5751 providing more distortion: What does this mean?
    And WRT the 5751 providing an imbalanced signal: What does that mean?

    Sorry for my ignorance, I have not read either of those statements before.
     

  10. ballynally2

    ballynally2 Tele-Meister

    Age:
    53
    387
    Jun 7, 2009
    donegal ireland
    if you're using nos valves like i do you should go w the specs of the rca manual fr the 1950s.
    Individual valves' specs will vary of course.

    [QUOTE="telemnemonics, post: 8017430, member: 44059"

    There is/ was a test list by a new tube seller where IIRC actual measured gain numbers were shown for various brands and types, and again IIRC there are some 12ax7s with lower gain than some 5751s.
    .[/QUOTE]

    With differences in variables it's hard to assess its impact other than listening. I maintain that a lower output v1 with a low (audible) distortion rate leads to a better balanced tone through the gain stages.The higher less clean signal in the first stage the more chance of distortion that the rest of the gain stages have to work with.
    That might actually be a desired effect hence the od/boost pedals for blackface amps.
     
    Last edited: Dec 1, 2017

  11. Commodore 64

    Commodore 64 Friend of Leo's Ad Free Member

    Age:
    42
    Mar 1, 2010
    Kent, OH
    Just watned to add appreciation for clintj's efforts on the load lines and plots. Thanks for doing that!
     
    clintj likes this.

  12. Commodore 64

    Commodore 64 Friend of Leo's Ad Free Member

    Age:
    42
    Mar 1, 2010
    Kent, OH
    Because the tubes are slightly different, plate resistance, transconductance, and other factors: the resulting voltage at the cathode with a fixed plate resistor (100k) and B+... is slightly different (1.3 vs 1.5 volts given clintj's test values), since the cathode is biased with a fixed resistance (1.5k). The cathode voltage, then, is developed across that resistance based on the standing current in the tube. That's what sets the bias in a cathode bias system. Standing current in a 5751 vs a 12AX7 is slightly different -- That's what the data sheets tell us.**

    So, on the 5751, there's a slightly different cathode voltage, which shifts the bias on the 5751 a little colder. This induces asymmetry in the output wave form. In this case, according to clintj, the amplitude of negative side of the sine wave is smaller than the positive side. This causes harmonic distortion, about 4.5% according to clintj's numbers/analysis.

    Marshall induces this type of imbalance for preamp distortion in the JCM800 by using a topology thus: A center-biased triode-->Cold Clipping Stage (10k cathode resistor) triode (inducing harmonic distortion with big asymmetry)--->warm-biased triode (820 ohm cathode resistor). This gives the thick "Marshall" tone we know and love. This is taking the principles of asymmetry due to preamp biasing to a much greater degree than we are discussing here.

    So in a circuit center biased for a 12AX7, a 5751 is going to provide a little less gain and a little more harmonic distortion all other things being equal. A little more harmonic distortion would generally be perceived as less sterile, or warm sounding to our ears (Unless you are Albert King). Not earth shattering, but this is a nice little exercise in understanding tube behavior in a guitar preamp. Good thread!

    **We assume tubes are behaving like the data sheets say.
     
    Last edited: Dec 1, 2017
    telemnemonics likes this.

  13. telemnemonics

    telemnemonics Poster Extraordinaire

    Age:
    58
    Mar 2, 2010
    Maine
    Interesting about the cold bias asymmetry result, thanks @Commodore 64

    Is this bias change with tube type change limited to a 5751 difference?
    Or is this also happening when we sub a 12ax7 for a 12ay7 in a Tweed?
    If it happens with all the typical pre tube subs, is there a typical result that follows?
    Is the colder bias going to a 5751 in a 12ax7 slot due to going to a lower gain tube, meaning that going to a higher gain tube will result in a warmer/ hotter bias?

    Right now I have a Brimar 12at7 driver in my non master 50w Marshall.
    Helps make the volume knob more usable and I get a great clean sound at slightly above conversation volume.
    Would there be a similar colder bias (on the 12at7)/ asymmetry/ distortion thing happening? I realize that 4.5% THD could/ would still sound clean.
     

  14. Commodore 64

    Commodore 64 Friend of Leo's Ad Free Member

    Age:
    42
    Mar 1, 2010
    Kent, OH
    The best way for us to know would be for you to take voltages on that tube and note the resistor values. Then we (err. clintj ;)) could plot the load lines and tell you exactly how the signal is behaving in that stage. A 12AT7 is more of a current driver than a voltage driver, so I would expect higher standing current, so higher cathode voltage (cooler bias) when compared to a 5751 or 12AX7 (as long as the cathode resistor is the same).**

    Yes, it's also happening with a 12AY7.

    Main point is, it's not SIMPLY a lower gain multiplier on the signal. Gain factor is derived from the spacing, size, and perhaps doping of the components within a tube.

    I highly second the suggestion you look into Merlin Blencowe's book: http://www.valvewizard.co.uk/. I have it. And I still don't understand most of it, but I reread it all the time in conjunction with threads here that remind me how little I understand. And each time, I get a little more elucidation (I hope).

    **For reasons I don't yet understand, the distortion on a 12AT7 (and other current driving triodes [12AU7]) is considered to be "less musical". One time, I read something in Merlin, and it made sense. I remember an AHA! moment, but I can't remember it right now, (dammit).

    Edit: By the way, some of the Dumble magic is the tweaking of bias and harmonic distortions in the preamp, some people say. So this is a fun thing you can do to your amps with little changes like cathode and plate resistors. A mass produced amp isn't going to be tweaked to that level of minutia.

    Swapping in tubes with different characteristics is also a way to do this. But it's kinda like throwing a dart. The 9-pin tube socket is the dartboard. The dart's gonna stick, but it might not be a bullseye.
     
    Last edited: Dec 1, 2017
    telemnemonics, clintj and JD0x0 like this.

  15. Bendyha

    Bendyha Tele-Afflicted Ad Free Member

    Mar 26, 2014
    Northern Germany
    telemnemonics, this is results of a 12AX7 to 12AT7 swap.

    250V supply 100K plate, 820R cathode: at Q. approx.
    12AX7 : Bias - 1.3V, Current - 0.85mA, Plate - 162.5V
    12AT7 : Bias - 2V, Current - 1.4mA, Plate - 115V

    Both are about equally central enough biased for maximum clean headroom
    Approx. maximum symetrical swing before clipping or cut-off;

    12AX7 : 2.6V bias swing, swings output at plate from 80 - 225V = 145V
    12AT7 : 4V bias swing, swings output at plate from 30 - 185V = 155V

    The circuit parameters for the 12AX7 are almost optimal.
    The 12AT7 will work in the same circuit, but would be much better suited to a higher supply voltage with a smaller plate resistor, somthing like 350V and 56K.
     
    Last edited: Dec 1, 2017
    telemnemonics likes this.

  16. JD0x0

    JD0x0 Friend of Leo's

    Age:
    27
    Feb 22, 2009
    New York
    My field testing disagrees with this. I realize AT7's are not 'audio tubes' so what? We're not looking for linear audio response like a HiFi amp would need.. I think with the right values a 12AT7 can overdrive nicely.

    I will say in my EHX English muffin which is a 2x tube Overdrive pedal running high voltage (280VDC+ B+) A 12AT7 in the V2 slot sounds really good. The cathode follower has a lower impedance with the 12AT7 and to me makes it sound more open an clear than other tubes. (BTW, stock tubes in that pedal are 12AU7)

    In my monsterized twin the 12AT7 PI gets some really nice distortion, IMO. Plate resistors are mismatched @ 47.5k and 52.5K with a 10K balance pot. The tail resistors are 470/10k/100 and NFB resistor was raised to 2.2k up from 820. (Stock values being 47K plates, 470/33k/100 w/ 820 NFB resistor)
    I typically hate noticeable amounts of PI clipping but this amp just sings when the PI starts break up.
     
    telemnemonics likes this.

  17. clintj

    clintj Friend of Leo's

    Apr 4, 2015
    Idaho
    If I had a spare 5751, I could easily plug one in and scope the stage. Unfortunately, my last one was included in a repair job and left with the customer. Such is life.

    You could say that changing the input stage on a tweed is simply a gain issue. It's only getting shown a signal of a couple hundred mV, and the distortion factor needs a larger input signal to become noticeable. Those amps with a cascaded preamp, where a larger signal hits the grid, would be more prone to the cold bias making itself known - like the HR series amps, where V1a feeds V1b for example.
     

  18. ballynally2

    ballynally2 Tele-Meister

    Age:
    53
    387
    Jun 7, 2009
    donegal ireland
    that depends on what slot the 12at7 is in.we are talking valve amps here.pedals are a different cattle of fish.in amp a 12at7 seems more useful in pi than v1/v2.in v1 the gain might be too low as such despite other specs.12ay and 5751 closer to the required 12ax7.
     

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.