The Dynaco Tube Audio Forum

Dedicated to the restoration and preservation of all original Dynaco tube audio equipment - Customer support for Dynaco VTA tube amp kits, all Tubes4hifi.com products and all Dynakitparts.com products


    vta70 jj-6550 evaluation and optimal bias

    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Mon Mar 11, 2019 8:30 am

    At 4325h usage of JJ6550 i felt it's time to evaluate the 6550.
    These observations might be of value to other users. Don't
    hesitate to ask for details and or other type of measurements.

    New in my toolkit is a couple of hameg stuff, a 8027 distortion meter,
    a 8037 low-distortion sine generator and more.

    All measurement is done with one channel driven and 230V mains. 8 ohm resistive load
    JJ ecc82 as smalltubes, jj gz34 as rectifier

    Status before any changes : bias at 450mV, jj6550 used for 4325h.

    At 1V output 0.04% dist
    at 17V  ( 36w ) 1.36% dist, this is the highest power below clipping
    The 6550 seems to have lacked some power in spite keeping bias well
    during the 4000h they are used. No power measurment done at start
    of period.

    Some new JJ tubes tested :
    JJ EL34-II :
    1V   0.04%
    ( 1V corresponds to 0.125w power)

    18V 450mV bias ; 0.61%
    18V 500mV bias (optimal) : 0.67%

    ( 18V corresponds to 40.5w power over 8 ohm )

    JJ EL34 (kit 1) :
    1V  0.04%  500 and 600mA
    18V  500mA 0.9%   600mA 0.5%

    JJ El34 ( kit 2) :
    500mV bias
    1V  0.05%  same with 520mV bias
    18V 500mV bias 0,76%
    18V  520mV bias ( optimum value for full power ) 0.64%

    Conclusion seems to be that at full power higher bias reduces distortion.
    At low power bias made no difference , no crossover distortion sighted, in
    spite looking at the distortion remains where the 8027 has an output for
    the remains. What could be observed was noise hum and higher orders of
    distortion components.

    As for the 6550 i have original and current reading taken with a maxi-matcher :


    tube 1 2 3 4
    Bias mA
    new 32 35 34 31
    now 26.6 26.3 26.0 24.5

    Gm ( as per maximatcher )
    new 414 430 428 416
    now 368 345 355 354

    tubes 1 and 2 was left channel ( the measured one)



    Last edited by peterh on Mon Mar 11, 2019 8:43 am; edited 1 time in total (Reason for editing : Added maximatcher readings)
    WLT
    WLT

    Posts : 101
    Join date : 2013-07-13
    Location : Rochester NY

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by WLT on Mon Mar 11, 2019 11:14 am

    Always good to look at data. Thank You.

    It all makes sense and reaffirms long stated comments from this forum and others. Then what can we take away from the results? For me it shows that
    1) Even old tubes can measure well at low levels.
    2) New tubes will measure better than old at higher output
    3) The higher the idle current the lower the THD
    4) The difference in THD is not sufficient to raise the idle current which reduces the longevity of the output tubes (for me).

    In the golden age of hifi many manufactures (Dynaco, McIntosh HK Citation etc.) would run the idle current up to obtain better THD specs. Output tubes were not expensive so buying a set every few years was not a big deal. Today… not the same.

    A few other rants. So even with worn tubes the amp could sound good at low levels. And maybe even at higher levels. You did not state anything about how the new tubes sound compared to the worn set. However the new tubes probably sound better when the amp is putting out a higher wattage.

    Any follow up comments on sound quality?
    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Mon Mar 11, 2019 11:21 am

    WLT wrote:Always good to look at data. Thank You.

    It all makes sense and reaffirms long stated comments from this forum and others. Then what can we take away from the results? For me it shows that
    1) Even old tubes can measure well at low levels.
    2) New tubes will measure better than old at higher output
    3) The higher the idle current the lower the THD
    4) The difference in THD is not sufficient to raise the idle current which reduces the longevity of the output tubes (for me).

    In the golden age of hifi many manufactures (Dynaco, McIntosh HK Citation etc.)  would run the idle current up to obtain better THD specs. Output tubes were not expensive so buying a set every few years was not a big deal. Today… not the same.

    A few other rants. So even with worn tubes the amp could sound good at low levels. And maybe even at higher levels. You did not state anything about how the new tubes sound compared to the worn set. However the new tubes probably sound better when the amp is putting out a higher wattage.

    Any follow up comments on sound quality?
    Sound is the same; good !
    About bias level: these new EL34 measures best at 520mA, thats where they will
    stay. My hours-meter will tell me usage time and time will tell if they will age prematurely.
    The 6550 was used > 4000h, still sound good, but i can see that their electrical
    performance has degraded. I will examine them in a uTracer later on.


    Last edited by peterh on Mon Mar 11, 2019 4:42 pm; edited 1 time in total (Reason for editing : Added future actions with old tubes.)
    WLT
    WLT

    Posts : 101
    Join date : 2013-07-13
    Location : Rochester NY

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by WLT on Mon Mar 11, 2019 7:34 pm

    Just to clarify. In your latest comment when you say 520 mA it is actually 52 mA per tube not 520 mA per tube. Agreed?

    I went back to look at your data and your latest comment. I think you are using a 10 ohm resistor to measure across. The JJ EL34 (kit 1) has the same mV/mA issue. Agreed?

    One other item. At 52mA and almost 480 VDC across the tube the plate dissipation is about 25 watts. It should work fine but that rating is the maximum for an EL34. I was lead to believe we should run them about 80% of maximum. I set my EL34s for about 40mA to match the 80% point. The 6550s are fine at 50 mA as they are a 35 watt tube.
    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Mon Mar 11, 2019 9:25 pm

    WLT wrote:Just to clarify. In your latest comment when you say 520 mA it is actually 52 mA per tube not 520 mA per tube. Agreed?

    I went back to look at your data and your latest comment. I think you are using a 10 ohm resistor to measure across. The JJ EL34 (kit 1) has the same mV/mA issue. Agreed?

    One other item. At 52mA and almost 480 VDC across the tube the plate dissipation is about 25 watts. It should work fine but that rating is the maximum for an EL34. I was lead to believe we should run them about 80% of maximum. I set my EL34s for about 40mA to match the 80% point. The 6550s are fine at 50 mA as they are a 35 watt tube.
    Yes, it's 520mV and 52mA / tube. Sorry for the confusion.
    This totals to 25w , split between plate and g2.
    Max Pa=25w and max g2=8w which sum's up to 32w.
    Measuring across a cathode resistor will include both Ia and Ig2.
    PeterCapo
    PeterCapo

    Posts : 678
    Join date : 2008-12-05

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by PeterCapo on Tue Mar 12, 2019 8:39 am

    All interesting, thank you.
    avatar
    BNR_1

    Posts : 110
    Join date : 2013-06-11

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by BNR_1 on Tue Mar 12, 2019 10:55 pm

    Nice....real data.

    Was the JJ GZ34 operated for the same 4325 hours? Any quantitative readings on the rectifier before and after?

    Thanks for tracking and sharing.
    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Wed Mar 13, 2019 1:32 am

    BNR_1 wrote:Nice....real data.

    Was the JJ GZ34 operated for the same 4325 hours?  Any quantitative readings on the rectifier before and after?

    Thanks for tracking and sharing.
    No, the JJ GZ34 was installed when the original chinese gz34 was broken, i guess
    at about 2000h. The pre-tubes was replaced at i guess 3000h due to one of the
    (5963? from bob) developed a hum.
    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Thu Mar 14, 2019 2:52 pm

    About Gm readings on my used 6550 tubes :
    Maximatcher (seems to) measure Gm at the only working point they have : 400V and g1=-48V
    At that point a very low reading is obtained. This is partially confirmed with my uTracer. Still
    the readings are useful for comparing and matching tubes which is what i use it for.

    Examining the tubes in my uTracer will make comparation with Tunsol ( the original US tungsol,
    not the russian copy) possible. Measuring at the "standard" 250V plate and g2 , -14v on g1

    Testing 2 of the 6550 ( #3 and #4)
    #3 10.18
    #4 9.44

    and a new JJ 6550 grabbed randomly from my pile 11.5

    Ia at the "typical" conditions is
    #3 81% 112mA
    #4 79% 108mA
    and the same new 6550 86% 122mA

    Thus, we observe that after 4000h of use emission has gone down a few %, Gm has been
    reduced more ( 10-15% ) . Max power measured in the amp has gone down slightly, unfortently
    i have no values from when they were new.
    There is still no gas in any of the 4 , getter is still silvery, thus
    i would not expece flashover or bias runaway yet. But i have no urge to test this. I will
    save the tubes in case some other observations are needed.

    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Thu Mar 21, 2019 2:31 pm

    An GZ34 flashed today, also one EL34 draw much to much and also flashed. Both replaced.
    I guess 52mA bias is too much, will back down to 48.

    PeterCapo
    PeterCapo

    Posts : 678
    Join date : 2008-12-05

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by PeterCapo on Thu Mar 21, 2019 2:38 pm

    Peter, have you tried any of these tests and measurements with KT66?
    peterh
    peterh

    Posts : 1134
    Join date : 2012-12-25
    Location : gothenburg, sweden

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by peterh on Thu Mar 21, 2019 2:59 pm

    PeterCapo wrote:Peter, have you tried any of these tests and measurements with KT66?
    No. I only have one quad of these currently and i do not want to experiment with tubes
    that is to be sold as new.

    Sponsored content

    vta70 jj-6550 evaluation and optimal bias Empty Re: vta70 jj-6550 evaluation and optimal bias

    Post by Sponsored content


      Current date/time is Mon May 20, 2019 5:44 am