My first foray into the KT88 world

Discussion in 'Tube Audio' started by mroboto, Jul 6, 2018.

  1. mroboto

    mroboto Super Member

    Messages:
    1,387
    The schematic says 120V for one of the cathodes of each 6N8P , the pin 6 of one tube and the pin 3 of the other.

    On that position I have 100V electrolytics.

    100V may be short, I think about replacing them.
     

     

    Please register to disable this ad.

  2. dowto1000

    dowto1000 Active Member

    Messages:
    188

    Yes, but that Rk cap is IN the audio path, so realize, no electrolytic will sound that great.

    Use a few film cap(s) , is a humble suggestion.

    As part of your future audio building arsenal of parts, experiment and look into using WIMA DC LINKs with 4 leads, aka MKP 4s, as a main Rk film cap. If on a budget, try that main value DC LINK WIMA bypassed with a couple of smaller WIMA FILM caps, say a 0.22 uF at 1250 FKP 1 and a 0,022 uF at 2,000 VDC FKP 1.

    Let us know, if you experiment, HOW that sounds to you. Have fun listening.

    Dowto1000
     
  3. mroboto

    mroboto Super Member

    Messages:
    1,387
    I already ordered 100uf 200V Nipon Chemicon 105°, and I will bypass them with 4.7uf 250V CBB Metalized film that I already have.

    They will replace 100uf 100V Nipon Chemicon.

    There is also some Nichicon and Elna capacitors in this amp.
     
    Last edited: Jul 10, 2018
  4. dowto1000

    dowto1000 Active Member

    Messages:
    188
    Well, respectfully submitted, here is my take on the Rk bypassing circuit parameter

    Do a little thinking before copying the Chinese engineers' values. Try to convert the entire front end of that amp, to all film caps for Rk bypasses. The schematic you shared is very telling !!!

    On the input stage, they have a 2.7K Rk with a 220 uF lytic !! That is 0.27 hZ. Very poor, useless values and design, for SE - tubes IMHO.

    Use a 20 uF film, on the first gain stage, and you have a 2.95 hZ turnover. Hell, I'd be tempted to try two ten uFs WIMA DC LINKS, single first. listen, and then doubled, and select whats best, by ear.

    Even more laughable is their middle stage, which you are addressing. 33K and 100 uF is 0.05 hZ. Again, IMHO, very very poor design for SE - tubes

    A nice WIMA DC link at only 5 uF will get you down to 0.96 hZ.

    I would not worry about the Finals tubes lasting. The KT88 is rated 42 Watts maximum, and you are operating at 72 mA. and ONLY 24 Watts of plate dissipation, about 57% of maximum. I usually shoot for a Golden Ratio percentage of maximum, which is 62 %. They did real good there !!

    Have fun.

    Dowto1000
     
  5. mroboto

    mroboto Super Member

    Messages:
    1,387
    Thanks Dowto for your suggestions, but I don't want to modify too much the circuit values, I will slowly get acquainted to this amp, and I will see.

    I bypassed the B+ cap, there is only confortable place for one 4.7uf 630V CBB.

    I also bypassed the first stage cathode cap with 4.7uf 50V multilayer ceramic caps.

    I will probably wait for the caps I ordered today prior to continue.

    I will not bypass the output tubes bypass capacitors, space is limited, and I think this could not affects as much as bypassing the others primary stages cathodes bypass caps.

    One thing I wish, is to ground the chassis if this cause no trouble, I feel like I will be more secure.
     
    Last edited: Jul 11, 2018
  6. dowto1000

    dowto1000 Active Member

    Messages:
    188
    As a final thing to try on that circuit, that has always served me excellently since 1981, would be to eliminate the 5 HY 250 mA. choke. The amp draws 77.4 mA. total, per channel, as schematically shown. That would be 72.3 mA. for the finals, 3.6 mA. for the middle stage, and 1.5 mA. for the input stage. Amp's TOTAL current draw for B+, as stereo amp, is 154.8 mA, or 0.1548 A.

    The choke is dropping ( 390 V - 382 V ) or 8 VDC across it. Eight volts divided by 0.1548 A. is 51.6 Ohms of DCR, which is typical, and that 5 Hy. Choke is at least 2 times too high in DCR, for a choke to KT-88 finals, IMHO, IME. Most audio designers are unconscious about DCRs in chokes, but good.

    I could re-design that for you, ( likely using two Hammond off -the-shelf chokes, 159Ys, 0.6 HY at 750 mA. and only 11 Ohms DCR ) and get the KT-88 amp to WAY outperform the stock 51.6 Ohm choke, on the basis of perceived drive, dynamic contrasting, and just plain ole general FUN to listen to !

    Do the other obvious stuff first. Get rid of those electrolytics on the front-end's Rks totally, use the film types and values I suggested. If you have a means of volume control, replace the amp's nasty and very cheap sounding pot, with two really high-QUALITY 100K Rs, and re-position the RCA jacks very close, two inches or less, to the two driver tubes.

    This is almost free performance improvements, and simple common-sense engineering.

    When I get some time, I will mess-around with a PSUD2, I may look "one day" to re-do the too-high DCR choke. Not today !!

    Have fun.

    Dowto1000
     

     

    Please register to disable this ad.

  7. mroboto

    mroboto Super Member

    Messages:
    1,387
    I read that the more H the choke is , the more the bass response is effective.

    My theory on electrolytics and films, is that film and ceramic caps with low ESR corrects the flaws of the electrolytics when added to, there is no point to remove any lytics.

    And adding too low value bypass cap contribute to increase inductance.
     
  8. dowto1000

    dowto1000 Active Member

    Messages:
    188
    What
    The books one reads on power supplies for audio amps is out dated and out-moded. In the last couple of decades, new and far higher performing power supply topologies have been developed, by people who " read it all " but also listen to their builds.

    I will not debate EE theory with others, sorry. The best performing amps are built by listeners, not theorists, in all of my experiences.

    One has to know the theory, and LISTEN and determine when not to pay attention to theory, but rather, heed what they hear. In the end, performance determined by listening, not simply one's reading of old theory, counts the most to me, personally. YMMV, fine.

    Enjoy your time spent with that amp. Have fun. Good luck.

    Dowto1000
     
    Last edited: Jul 11, 2018
  9. mroboto

    mroboto Super Member

    Messages:
    1,387
    Yeah, I listen a lot,

    I started decoupling electrolytics using 2.2uf film capacitors, and I realized later how much of an improvement the 4.7uf does.

    The 0.1uf 1000V Ceramic disc capacitor does another type of improvement.

    Both combined can give one of the finest sound in my oppinion.
     
  10. dowto1000

    dowto1000 Active Member

    Messages:
    188

    Sure, I have used up to seven in parallel in different circuits, across a " main" film cap. Each value plays a different part of the audio spectrum. I sometimes start with a 5 uF at 1200 VDC, a GTO cap which can do 1,500 AMPERES instantaneous peaks, and use this on speaker crossover caps and caps going to my SE amps' finals stage, as the first of several bypasses.

    Dowto1000

    My own Altec A7-800 VOTT crossover shown, to a 515B and 802D. Large round Gray caps are the GTOs. Lovely to get the leading edge, of initial transients.



    P1010017.JPG
     
  11. mroboto

    mroboto Super Member

    Messages:
    1,387
    Worked my brain and finally find a way to install 2 x 4.7uf 630V CBB + 1 x 0.1uf 1000V Ceramic Disc under the resistors.

    The main B+ work is finished.

    Listening now for a few minutes after the work done, sound improved a lot.

    b.jpg

    Thermistors mounting looks flimsy that way, I ordered terminal strips to mount the them on, waiting for them to arrive.
     
    Last edited: Jul 11, 2018

     

    Please register to disable this ad.

  12. triode17

    triode17 Well-Known Member

    Messages:
    973
    Cathode poisoning takes quite a long time to happen. Days to weeks. Don't worry if it was off for a hour or two.
     
  13. triode17

    triode17 Well-Known Member

    Messages:
    973
    Your schematic looks odd. Shouldn't the 470 go to the plate? Plus, why not earth ground it. It only makes it safer to run.
     
  14. triode17

    triode17 Well-Known Member

    Messages:
    973
    From the GEC KT88 data page. "the outside of the bulb should not exceed 250 deg.C." That's 482 deg. F. Why are you so surprised ? These suckers get damn hot !
     
  15. mroboto

    mroboto Super Member

    Messages:
    1,387
    Yes, but if I forget one night, and another night, and another night...

    This seems OK.

    I will ground it.

    If I have ground loop problem I will revert.

    Yes, but runned at maximum plate dissipation.
     
  16. triode17

    triode17 Well-Known Member

    Messages:
    973
    Well, then grab it when its running and let us know how hot it gets.
     

     

    Please register to disable this ad.

  17. mroboto

    mroboto Super Member

    Messages:
    1,387
    You can quickly slip a finger on the side of a tube which operates softly without getting seriously harmed.

    I do not recommend doing the same on a pan in the oven at 350°F !

    The temperature difference between both sides is too much to be the glass temperature, at 340°F on one side and 295°F on the other, focused like this, I beleve it's directly on the plate, the hot spot.
     
    Last edited: Jul 12, 2018
  18. mroboto

    mroboto Super Member

    Messages:
    1,387
    Today I started hearing small ticking noises, swaped my source from my TubeDAC to my small USB DAC and heared again after about 3 minutes warm up of the amp,

    I swaped the stock russian 6H8C (those without top sides micas) fot others 6H8C , (with top side micas) , and no thicking noises for now.

    Edit: After contacting the seller, he proposed to send replacement tubes.
     
    Last edited: Jul 12, 2018
  19. mroboto

    mroboto Super Member

    Messages:
    1,387
    I still hear occasional ticking sounds since I added decoupling caps, but very occasionnal, I think to do a better grounding scheme.

    1. Tie ground prong of the power supply cord to chassis.

    2. Tie high voltage center tap to ground.

    3. Tie the negative side of the B+ cap to ground.

    4. Maybe another circuit grounding point to ground if not already, I see a ground wire on my photos, but I don't see where it leads to.

    That make sense, I think.

    PS:

    And remove the capacitor at the right from the circuit.

    Clic on the link at the right of "Document joint" to see full size image.

    https://www.cjoint.com/c/HGkriBpp0qs

    PS2:

    Maybe it's the tube that ticking, even if it works fine in my other amp, because ticking sounds seems to only appear on the left side, but as this is very occasionnal, that's bizarre.

    I will invert left and right tubes to see if ticking follow the tube.
     
    Last edited: Jul 13, 2018
  20. mroboto

    mroboto Super Member

    Messages:
    1,387
    I realized I hear ticking noises in both channels.

    Sounds like a bad grounding sheme to start with and decoupling capacitors I added prior to trouble does not help , in fact, they transmit noises instead of filtering them ?

    For history:

    I added 4.7uf 50V multilayer ceramic to first stage 220uf 25V E.C cathode bypass capacitor.

    And 2 x 4.7uf 630B CBB + 0.1uf 1000V ceramic disc to the B+ 330uf 450V E.C.

    The rest is thermistors added.
     
    Last edited: Jul 13, 2018

Share This Page