Solution for 2270 high bias - early serial number

Discussion in 'Marantz Audio' started by hirscwi, May 9, 2017.

  1. hirscwi

    hirscwi AK Subscriber Subscriber

    Messages:
    1,628
    Location:
    Indianapolis
    I just finished restoring an early serial 2270 for a friend (s/n 2801). The bias on one channel would start at zero and climb to about 19mV over the period of an hour. I had originally replaced the bias transistors (H760 - each board) with ksc3503. Following the recommendation of an earlier thread, I tried a D669 and the problem was solved.

    I did not develop this solution. It was originally proposed by member ecluser in a thread titled "Marantz 2270 bias won't go below 15mV". The solution works because the 669 has a Vbe of .68 while the 3503 Vbe is .74 and this is enough difference to solve the problem. Many thanks to Glen (ecluser) for this solution. Please refer to his thread for a very thorough discussion.

    I have seen quite a bit of interest recently in the 2270 and thought this might be helpful to resurrect an excellent solution from the past.

    BTW, it's been my experience that the early 2270s take at least an hour to stabilize the bias.
     
    rhodus, Hipocrates, ranch 22b and 2 others like this.

     

    Please register to disable this ad.

  2. Steven Tate

    Steven Tate AK Subscriber Subscriber

    Messages:
    2,481
    Location:
    N. Richland Hills, TX
    Good information. Thanks for posting! :thumbsup:
     
  3. rBuckner

    rBuckner Luv 2 Restore Subscriber

    Messages:
    9,512
    Location:
    Prescott, Arizona
    ^^^Indeed.
     
  4. sssboa

    sssboa Well-Known Member

    Messages:
    513
    ok, and what about s/n 55000 and higher, do they have originally different transistors there than the early numbers? My bias is not that stable either.
     
  5. hirscwi

    hirscwi AK Subscriber Subscriber

    Messages:
    1,628
    Location:
    Indianapolis
    The original thread I referenced contains information for the units with serial above 3900. The recommendation is to use D669 for the bias transistor, as a starting point.

    Instability of the bias current might be some other problem. There are quite a few threads in this forum about bias problems. Worth reading.
     
  6. sssboa

    sssboa Well-Known Member

    Messages:
    513
    Will check it out. For the start, isn't D669 obsolete now? Any equivalent?
     
  7. sssboa

    sssboa Well-Known Member

    Messages:
    513
    As I looked on ebay you may buy NOS Hitachi, today's Chinese production as NOS Hitach or openly today's production - not branded or by Hi-Sincerity Microelectronics from Taiwan.
     
  8. hirscwi

    hirscwi AK Subscriber Subscriber

    Messages:
    1,628
    Location:
    Indianapolis
    Try B&D Enterprises. They had some not too long ago. If you're looking for substitute, you want the lowest Vbe - D669 is .68v
     
  9. hirscwi

    hirscwi AK Subscriber Subscriber

    Messages:
    1,628
    Location:
    Indianapolis
    If you can't find D669, the following work but don't have as low Vbe, I don't think. KSC3503; JE243G; 2SC495; 2N4923
     

     

    Please register to disable this ad.

  10. sssboa

    sssboa Well-Known Member

    Messages:
    513
    I think Hitachi stopped making them decades ago. So it's risking buying fake Hitachi or buy this

    d669.jpg
     
  11. sssboa

    sssboa Well-Known Member

    Messages:
    513
    What is Vbe of original transistors 2SC984 and 2SC496 ?
     
    Last edited: May 11, 2017
  12. Steven Tate

    Steven Tate AK Subscriber Subscriber

    Messages:
    2,481
    Location:
    N. Richland Hills, TX
  13. catrafter

    catrafter Marantz Specialist Subscriber

    Messages:
    5,998
    Location:
    Montana
    I made Vbe measurements a while ago using an Atlas Peak DCA55:
    2SD669=0.68V
    MJE243G=0.70V
    2SD612=0.70V
    2SC495=0.72V
    KSC3503=0.76V

    YMMV,
    Tom
     
    iceman57ad and runningdog like this.
  14. sssboa

    sssboa Well-Known Member

    Messages:
    513
    Looks like MJE243 will do the job :)
    And I know what I'm buying.
     
  15. sssboa

    sssboa Well-Known Member

    Messages:
    513
    Turns out I already have 2N4923.
    I did some measurements with my Chinese MK-328. It gave me some results in different scale but consistent i.e. 2N4923 Vbe would be around 10% less than KSC3503 which I have too, so around 0.68V (if KSC3503 has 0.76V)

    BTW. Do you know anything in TO-92 form that would do the job? I have some slip-on radiators for TO-92 that maybe I would attach to the radiator.
     
  16. dgunnr

    dgunnr Marantz Owner Subscriber

    Messages:
    170
    Location:
    Southern NJ, USA
    Unrelated, apologies but could u confirm this is the champagne engraved face plate or silver model? It should be the champagne model per my survey.

    Other readers please do not drift or take this bias thread into another direction due to my question. Look for 2270 engraved face plate thread if you have questions or input to preserve this important thread. Thank you!
     
  17. rBuckner

    rBuckner Luv 2 Restore Subscriber

    Messages:
    9,512
    Location:
    Prescott, Arizona
    I'm going to guess that since it was always a transistor in a medium power package that a TO-92 won't be safe to use.
     

     

    Please register to disable this ad.

  18. hirscwi

    hirscwi AK Subscriber Subscriber

    Messages:
    1,628
    Location:
    Indianapolis
    Yes, the unit in question is champaign faceplate. s/n 2801
     
  19. dgunnr

    dgunnr Marantz Owner Subscriber

    Messages:
    170
    Location:
    Southern NJ, USA
    Thank you and good luck getting your questions answered. Dennis
     
  20. runningdog

    runningdog AK Subscriber Subscriber

    Messages:
    337
    Location:
    Sonoma, CA
    This chart helped me solve an issue with a 2245 amplifier, which is similar to the 2270. It had widely drifting bias, which started after replacing the original H760 bias transistor with a MJE243G. And bias could not be set correctly at R763. After viewing Tom's list, I decided to try the KSC3503. It's now possible to adjust the bias at R763, and it is reasonably stable now. With the MJE243G it was drifting from 2.0mV to 8mV.

    This 2245 is an early champagne face with the older style 2245/2270 amp board. Tomorrow I'll try this on the other amp and see if this fixes the same issue with it.

    Thank you Tom!
     

Share This Page