1. Time for some upgrades in server hardware and software to enhance security and take AK to the next level. Please contribute what you can to sales@audiokarma.org at PayPal.com - Thanks from the AK Team
    Dismiss Notice

AES Six Pac project assist

Discussion in 'Tube Audio' started by cwh, Dec 2, 2018.

  1. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,987
    Location:
    Southern NJ
    ^^ this, basically not something you'd deal with in an amplifier outside of an across the line or line to chassis cap.
     

     

    Please register to disable this ad.

  2. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    I got the amps ready for lift off. Any hand holding to avoid certain death? Or just plug in tubes and go?
     
  3. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,987
    Location:
    Southern NJ
    maybe feed 12v in and verify you've got approximately 1/10 the voltage you should have at the power supply points. Most importantly you'd want to confirm you have a workable bias supply. I'd set the bias trimmer so its as negative as it can be. Should be right around -.65v. The other way should give zero volts, which will be no good.
     
  4. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    Yeah it's got a bias jack and I just plug meter in. They are definitely all the way down, even as it sits.

    PS points would be across the main filters?

    Thanks gadget!
     
  5. gadget73

    gadget73 junk junkie Subscriber

    Messages:
    36,987
    Location:
    Southern NJ
    yes

    The bias test point won't read anything until the tubes conduct. I was thinking the measurement at the grid of the output tube socket to confirm where that pot is. Might also be worth sweeping the pot to make sure it doesn't have dead spots. If that pot is flaky it could lose bias to the tubes and cause you some grief.
     
  6. kward

    kward AK Subscriber Subscriber

    Messages:
    3,506
    Location:
    Utah
    Here's roughly what I do:

    (this first step assumes power supply caps are rated at least 1.5x more than the secondary voltage (i.e., 500V caps with a 300-0-300 secondary). Take all the tubes out and power it up. Check bias voltage that it's sufficiently negative, and make sure nothing else smokes. Check voltage on center tap of OPT. Should read 420V to 450V or so for this amp. Check bias adjust range by rotating bias pot all one way, then all the other way and check voltage range. Check that the bias voltage is reading on each control grid pin of each tube socket. Check that you have screen voltage on the screen pins of each output tube socket. Screen voltage should measure same as CT voltage at this point.

    Next, put only the power tubes in (no small signal tubes yet). Set your meter to measure bias voltage on output stage. Power it up. If the meter reads significantly more than what it should, somethings wrong...power it off and debug. If all goes well, then adjust bias pot until you get the bias reading the manufacturer recommended. If all goes well, then...

    Power it off, put all tubes in and power it up.
     

     

    Please register to disable this ad.

  7. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    A6918CA6-4B7E-413A-B462-7CE3FAE75191.jpeg
    @kward
    asaume as you were graciously typing I was turning up the voltage on the second amp...and by no means are your words lost, they’re in my book now!

    So I did slowly bring them up monitoring the bias current and the amp draw. I think things are good. At the manufacturer’s voltage (117v) I biased them to factory 230ma. The Kill a Watt showed about 1.75a.

    With that said should I take other measurements as it sits? If so where and how?
     
    Last edited: Dec 5, 2018 at 7:16 PM
  8. kward

    kward AK Subscriber Subscriber

    Messages:
    3,506
    Location:
    Utah
    Yes I think it's still a good idea to do as much bench testing as you can. For the DC conditions you can use a volt meter as descriibed previously. I think it's a good idea to write those voltages down and file them in case you need to validate something later. You need to turn the amp over and prop it up on a suitable support and remove the bottom plate of course. Its always a fun experience to take those measurements on a live amp without getting shocked, err I mean killed. I usually do this by clipping the negative volt meter probe on ground so its hands free. Then carefully with the other probe test the voltages at those points discussed. One tends to compromise safety when both hands are in the amp while the amp is running.

    aC conditions should also be checked if you have the means. You need an oscilliscope and a way to generate a sine and square wave for this part
     
  9. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    Thank you kward! I assume all checks are referenced to ground.

    Your previous post (without tubes) mentioned bias check. That would be at one of the power tube control grids?

    Bench testing is possible on my O'scope. My function generator might need some help though. I did check output with 1k iphone sine but that didn't seem to have enough drive for the amp. Not sure if there an input impedance incompatibility?

    This is my first high voltage tube adventure so I'm playing it green and taking all the help I can get.
     
  10. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    Also I'll note that I found amp #2 had more transformer hum at 230ma bias than #1. Taking it down to 160ma made it significantly quieter. In the Mr. Carlson Six Pac video he actually found 160ma to be a good bias current so there's that.

    Not sure if this hum is indicative of something else at play. Are there noises to look for that help assessing amplifier health?
     
  11. kward

    kward AK Subscriber Subscriber

    Messages:
    3,506
    Location:
    Utah
    Yup, all DC voltage checks are referenced to ground, except perhaps the method to measure bias. It's probably measured across a small value resistor instead of from the high side of that resistor to ground.

    The larger hum in amp #2 may be due to more apparent mismatching of output tubes. This amp doesn't appear to have a way to adjust balance of a single pair independently, and apparently only has the means of measuring the aggregate quiescent current of the entire stage.

    The amp should be dead quiet at idle, when connected to a test speaker, even with your ear up within a few inches of the cone. On an oscilloscope, if you connect the amp to a dummy resistor load (say 8 ohm load on 8 ohm tap), and probe the signal across that load with the oscilliscope when no music is playing and with the input shorted, you will probably read a 60 Hz sine type wave at less than 2 mV peak. That value or less is approximately what it should read if there are no hum issues with the amp.

    If there is a ground issue in the amp's internal wiring, or any other sort of hum issue, it will show up quite obviously on the scope trace. Hum will be probably either 120 Hz or 60 Hz. You can determine which it is by finding the period of the hum signal (how many miliseconds from crest to repeating crest) and taking the reciprocal of that. If you have 120 Hz hum, it would be indicative of power supply switching coupling into the output stage or into the audio circuit. If you have 60 Hz hum, it would be indicative of mains voltage coupling to the grid of the voltage gain stage, such as through a long or unshielded wire from RCA input through to the grid of the first amplfiication stage tube. A bad ground (sloppy solder joint) or less than stellar ground system wire dress could also be the cause of hum. First place I'd look though is at the potential mismatch of output tubes. You might try swapping tubes from the good unit that is "hum free" into the other unit and see if that changes anything. That will tell you if it's the tubes or the amp.
     
    Last edited: Dec 6, 2018 at 1:52 PM

     

    Please register to disable this ad.

  12. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    Excellent. I'll get them back on the bench and scope the output.

    New matched tubes from Jim McShane. New main PS capacitors. Three other electrolytic filter caps haven't been changed yet.

    I'll check the star grounding for good metal to metal contact.
     
  13. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    I shorted the input and scoped the out. Couldn't find a sine wave. I also feed in the function generator at -20db slowly increasing the amplitude. I get audible (ringing?)without very much amplitude. Something isn't right.

    Might've mentioned previously that the hum is made more or less through the bias adjustment. 230ma is factory suggested and I'm at 160 during last test.
     
  14. kward

    kward AK Subscriber Subscriber

    Messages:
    3,506
    Location:
    Utah
    For the amp that's working correctly, the noise is going to be way way down there in the 1 to 2 millivolt range. You need to turn the volts/div up to the most sensitive setting, and you also need to turn the time/div down to the 2 mS or 5 mS setting. For the amp that is humming is should be easier to find--but you may need to turn the volts/div up a bit to 20 mV or 50 mV or even higher.

    When applying a test signal, it's normal to hear it through the vibration of the output transformer, even at low wattage output.

    If the output stage bias adjustment changes the hum level, it's likely rooted in the output stage, bias circuit, or output tubes, BUT I'd still try swapping ALL tubes between hum free unit and humming unit. It's the obvious first thing to try. It's 5 minutes of time well spent. You have to systematically eliminate possible causes, one by one.

    Your case is the ideal test setup. You have one amp that is quiet and one that hums. It's a lot easier to compare good against bad to see what's different when you have examples of both on your bench at the same time.
     
  15. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    18F803E6-32FA-4FFD-9289-7A49A2A29F09.jpeg

    Before I dig in and start from the beginning taking measurements with tubes out of circuit I did swap tubes. Hum stays with amp. It’s a hum just a bit louder than the amp.

    Signal into #2 looks good. Square is pretty square.

    #1 though has less ideal square.
     
  16. kward

    kward AK Subscriber Subscriber

    Messages:
    3,506
    Location:
    Utah
    Figured, (hum stays with amp) but now you know for sure and can focus on the amp as the problem.

    What frequency is the previously posted square wave?
     

     

    Please register to disable this ad.

  17. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    1k although as mentioned my function gen might be out of spec.
     
  18. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    Besides hum I'm hearing a noise generated in the power tubes. Sounds like water going through copper pipe or like the noise a filament makes when tap on an incandescent light bulb only continuous.

    ??

    Thanks for sticking with me guys, guys? :)
     
  19. BinaryMike

    BinaryMike Pelagic EE Subscriber

    I think the next step is to yank the small tubes and determine if the hum persists with only the finals in place. I hate to say this, but if it does, we could be looking at a failed OPT. You'll want to give the power supply plenty of time to discharge and check it with a voltmeter before proceeding with an ohms check on the primary winding.
     
  20. cwh

    cwh AK Subscriber Subscriber

    Messages:
    787
    Location:
    Oakland, Ca.
    I've got my questions asked in previous posts still lingering but......

    Since I've got stable bias at the amp's bias jack, nothing smoking, and wave forms that aren't too troubling to me I went ahead and hooked her up,

    After playing through Sgt. Peppers and Abbey Road at 1/2 volume all I can say is holy shit. My Bozak 302s are doing just fine.

    I've got equal balance and no perceptible distortion. Is this sustainable? This goes back to how do I test in a more complete and thorough manner.
     

Share This Page