First try for DC heater supply - voltage too high

Discussion in 'Shock Brother's DIY Amps' started by NSB_Chris, Oct 11, 2021.

  1. printer2

    printer2 Poster Extraordinaire

    Posts:
    7,583
    Joined:
    May 24, 2010
    Location:
    Canada
    Marshall runs their 12V terminals at 8.5V dc. Just saying.
     
  2. NSB_Chris

    NSB_Chris Tele-Holic

    Age:
    56
    Posts:
    620
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Really?! So just put that across pins 4 & 5.
     
  3. wallybob

    wallybob TDPRI Member

    Posts:
    17
    Joined:
    Sep 7, 2013
    Location:
    Iowa
    I have dropped DC heater voltages in the past by running the DC through series 1N400X diodes. You get 0.7 to 1V drop per diode. You have to consider diode dissipation for higher heater currents.
     
    NSB_Chris and andrewRneumann like this.
  4. printer2

    printer2 Poster Extraordinaire

    Posts:
    7,583
    Joined:
    May 24, 2010
    Location:
    Canada
    Look beneath V1.

    [​IMG]
     
  5. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    4,475
    Joined:
    Dec 30, 2019
    Location:
    california
    The *12* in 12A_7 tubes refers to the heater requirement of 12.6 volts to heat the cathode. (Likewise the *5* in 5Y3 denotes a 5 volt heater supply.)

    To use 12.6 volts heaters, the voltage is supplied to pin4 and pin5. Pin9 is not connected. As you know, the 12A_7 tube can also use a heater voltage of 6.3 volts when wired pins4&5 and pin9.

    12.6 volt heaters will be noisier than 6.3 volt heaters but may not be noisy enough to notice the difference.
     
  6. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    793
    Joined:
    Nov 23, 2020
    Location:
    Finland
    12**7 series tubes cathode heating is designed so that they lessen AC humming when 6,3VAC is used. Obviously the two filament wires are wound opposite direction? If heater voltage is 12,6VAC then this cancellation does not happen and it is not good idea to use!

    When power transformer has two 6,3VAC outputs using them series and rectifying, filtering and regulating using TO-220 chip should make quietest option. When 12,6VAC is available there is enough voltage so that filter capacitor does not be huge.

    6,3VAC can be rectified and filtered to 6,3VDC but often there is not enough DC to regulate it because regulator require at least 1,2V more input than target output voltage is.
     
  7. dougsta

    dougsta Tele-Meister

    Posts:
    407
    Joined:
    May 25, 2015
    Location:
    Edinburgh, Scotland
    12.6v dc is used in the preamp section on many printed circuit board based amps.
    The main design decision driver is supplying enough current for all the tubes from pcb tracks. Some even use a serial heater chain for all the preamp tubes, pull a tube and they all go dark.
     
    2L man and NSB_Chris like this.
  8. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    793
    Joined:
    Nov 23, 2020
    Location:
    Finland
    Because it would be very challenging to wound PCB filament the DC is good choice :)

    Doubling the voltage to 12,6VDC halfs the needed current, which leads to half PCB print power losses, regulator and diode current rating and filter capacitor value.

    Obviously somewhere in Europe mains voltage was chosen later than USA and it is about double and there is only half the copper mass in our homes wiring :)
     
    NSB_Chris likes this.
  9. NSB_Chris

    NSB_Chris Tele-Holic

    Age:
    56
    Posts:
    620
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    very interesting! Thanks for pointing that out. So they are just running the bridge into the 6.3v heater configuration (pins 4&5 together and pin 9) not the 12v configuration (pin 4 and 5 leaving 9 open). Either their bridge rectifier gets them closer to 6.3VDC or they just let it be higher than spec.
    Looks like they have capacitors to ground but do not reference to the chassis after the rectifier.
    They also fuse the crap out of it. They have one fuse for each power tube heater leg and then one for the preamp leg.

    In the lower left they have an info statement "... for rats nest purposes". Seems and odd statement.

    Fun facts! I did not know that!
     
  10. andrewRneumann

    andrewRneumann Tele-Holic

    Posts:
    970
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    What purpose does the connection to “W1 Chassis” serve?

    And this supposedly puts out 8.5VDC?
     
  11. Ten Over

    Ten Over Tele-Holic

    Posts:
    889
    Joined:
    May 13, 2015
    Location:
    Central California
    W1 Chassis connection balances capacitors C1 and C3. The heater winding has an artificial center tap with two 100R resistors to ground. When the junction of C1 and C3 is connected to the chassis, that is the same as connecting a center tap to that junction just like we've all seen in countless bridge rectified HT supplies. It also references the DC supply to ground, even though it seems to me that the supply is already referenced to ground by the artificial center tap on the winding. Maybe they got a better result by doing it this way.

    I would bet money that this DC heater supply puts out 6.3V.
     
    Last edited: Oct 15, 2021
    andrewRneumann and NSB_Chris like this.
  12. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    4,475
    Joined:
    Dec 30, 2019
    Location:
    california
    I am not sure.

    To quote Blencowe, "A common beginner's error is to try to add a ground reference on both sides of the rectifier. This will short out the rectifier! Only one ground reference is required (centre tap, humdinger, elevation, whatever). It can be on either side of the rectifier; the other side will received its reference through the rectifier. (Actually, you can spot this mistake in the Mesa Dual Rectifier and Carvin Legacy amps which foolishly use artifical centre taps on both sides of the rectifier, so maybe it's not such a beginner's error!)"

    Some of Marshall's earlier amps had a ground reference on both sides of the rectifier. Perhaps the C1/C3 connection to chassis was a move to mitigate the heater bridge rectifier problems that were happening. I have a JTM60 that had a blown heater bridge rectifier when it was orphaned at my door. Reading up on DC heater supplys, I am thinking maybe I have to remove one of the taps on the JTM60!:eek:

    EDIT: The heater voltage is high on my JTM60. High 7's iirc.
     
    andrewRneumann likes this.
  13. andrewRneumann

    andrewRneumann Tele-Holic

    Posts:
    970
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    @Ten Over I never considered using two balanced caps to provide a ground reference. Thanks.

    @Lowerleftcoast you were reading my mind with that Merlin passage. I wonder if in the case of having balanced resistors (or caps, it seems) on both sides of the supply, it isn't as big of a deal as long as they remain balanced. Still, it does seem unnecessary to have 2 ground references.

    @NSB_Chris back to your original question. I think you got scared off too easily. I think you saw a lot of volts because practically no current was flowing through the bridge (after the capacitor topped up) and there were only very small diode drops. I have a very similar layout and all I needed was one R47 1W resistor in the supply and I nailed 6.3vac and 6.3VDC.
     
  14. NSB_Chris

    NSB_Chris Tele-Holic

    Age:
    56
    Posts:
    620
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    I may play around with this and try a few things now that it is up and running. The amp sounds good but there is more hum than I like with the volume turned up for a decent growl. Not sure how much DC on the input stages will help with that or if it is coming from the longer unshielded runs coming from from the PI to the power tubes which run under the power supply filter board. Will be fun to experiment.

    First I will just hook the DC supply back up to the input socket and see what the supply levels out at with some load. If I did the math correctly, with 300mA demand from the 12AX7 tube, I figured I would have to have 3.3 ohm to drop 1 volt that the heater sees. Maybe 300mA is conservative and the actual current draw will be less. I may just purchase a variety of resistors so that I can dial it in.
     
  15. printer2

    printer2 Poster Extraordinaire

    Posts:
    7,583
    Joined:
    May 24, 2010
    Location:
    Canada
    You would lose. Where is the extra voltage going? If you do rectify the 6.3V and it is high (which it will be) you can put a few diodes in series, each produce a 0.7V drop. Might not be a good idea to use a zener ubless you actually do a RC network incorporating it. Or you can get a low-dropout regulator (LDO regulator). This is what I would do if I was using AC to supply the heaters and rectifying the preamp supply. Marshall only does the first two tubes, I think they might have thought to do it for the first one and then added the second. It doesn't cost them much more to do it right, they have had the odd problem with it enough that it is a known problem. On fusing everything, cheap protection, so why not do the dc right for the heaters?
     
  16. 2L man

    2L man Tele-Holic

    Age:
    61
    Posts:
    793
    Joined:
    Nov 23, 2020
    Location:
    Finland
    To tube amp chassis t is very easy to install TO220 L7812 +12V (perhaps 1...3$) regulator to metal chassis drilling one hole and two diodes between ref and ground to lift its output voltage to 12,6VDC. It gets warm when it heats two 12**7 tubes (2x150mA pins 4 to 5) but there is no need to install heat sink component.

    I did rectify and filter 12,6VAC near PT and wired DC to pre amp wounding wires although it is not necessary for DC and mounted regulator between tubes. Also using reg output 0.1uF capacitor is good practice.

    Two 6,3V series no load output might be 13,5V and after bridge rectifier and electrolyt about 18VDC. Power loss to regulator comes (18V-12,6V) * 0,3A = 1,6W.

    I don't remember what is needed electrolyt value but it is possible to calculate and it does not have to be as high as when 6,3VDC filament is used because on 12,6V use there comes more "spare" voltage when rectifier loss halves and current comes half as well.
     
    Last edited: Oct 16, 2021
    NSB_Chris likes this.
  17. NSB_Chris

    NSB_Chris Tele-Holic

    Age:
    56
    Posts:
    620
    Joined:
    Sep 2, 2017
    Location:
    New Smyrna Beach Florida
    Ok, so I hooked the DC heater supply to the input valve again and the voltage was about 6.9VDC. It must have pulled the 2A supply down enough that with the drop across the diodes, it put it barely within safe limits. I put a 2ohm 3W resistor in line with the DC supply and it dropped to 6.28 VDC or so. Pretty good!

    Problem is that the hum did not diminish, and may have even become worse. I don't think the heater supply is my main issue. I chop-sticked around to see if moving some of the wires would show areas where wire routing was an issue, but nothing showed any affect.

    I have some low noise 12AX7's on order so we will see if that helps. Can't wait to roll them into some of my other amps to see what effect they have.

    Next on this amp, I will play around with some of the grounding change options to see if that helps.
     
    andrewRneumann likes this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.