Announcement

Collapse
No announcement yet.

Help me understand power section of AC30

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    With cathode biased amps that are idling too hot, the root cause is often that the HT voltage is somewhat excessive. That seems to be the case here.
    The problem with cooling the idle point by increasing the cathode resistor value is that when the HT supply is poorly regulated, as with most guitar amps, as the cathode resistor value is raised, the decrease in HT current draw results in the HT voltage increasing. The end result being that the actual idle dissipation is little reduced, but the tone and response of the amp may be worse, due to reduction in the conduction angle and consequent increase in bias shift.

    A better approach may be seen to address the root cause by reducing the HT voltage a little, eg add resistance into the HT supply, rather than the power tube's cathode return. A few volts off the HT will reduce the plate voltage and plate / cathode current. So adding x ohms into the HT supply has a double whammy effect on plate dissipation and no effect on the amp's tone / response, whereas adding x ohms into the cathode resistor has a 'half whammy' on plate dissipation and may worsen the amp's tone / response.

    Hence rather than increasing the cathode resistance 21 ogm to 68 ohms, I suggest to add a suitably beefy 22 ohm resistor between the rectifier and reservoir cap.
    Which accords with some other versions of silicon rectified AC30s, eg https://el34world.com/charts/Schemat...x_ac301986.pdf
    My band:- http://www.youtube.com/user/RedwingBand

    Comment


    • #17
      Hence rather than increasing the cathode resistance 21 ogm to 68 ohms, I suggest to add a suitably beefy 22 ohm resistor between the rectifier and reservoir cap.
      I agree that reducing HT might be the better solution. Especially as this also reduces screen voltage. Some AC30s used a 47R HT dropper. But with around 200mA current draw at idle this means a HT reduction of only 9.4V or 2.6%. That won't significantly bring down the 150% plate dissipation at idle. HT would probably need to be dropped by 30..40V to get the original 120% dissipation.
      Last edited by Helmholtz; 02-21-2020, 10:09 PM.
      - Own Opinions Only -

      Comment


      • #18
        There are merits of both approaches. As I was seeing thermal runaway having more resistance in the cathode seemed like the prudent approach. In addition, no loss of output power.
        Experience is something you get, just after you really needed it.

        Comment


        • #19
          My guess is that adding ~20 ohms into the the tubes’ HT current path will result in a similar degree of power output loss, whether that resistance is in the ‘feed to the plate’ or the ‘return from the cathode’. As either way, the resulting voltage drop reduces the available effective HT.
          My band:- http://www.youtube.com/user/RedwingBand

          Comment


          • #20
            either way, the resulting voltage drop reduces the available effective HT.
            Agreed, but I don't see how 22R could do much to avoid the replating. You'd probably need something like a 100+R HT dropper.
            Last edited by Helmholtz; 02-22-2020, 02:09 PM.
            - Own Opinions Only -

            Comment


            • #21
              Or perhaps a sting of Zener Diodes to reduce B+, just another thought, although resistors it's my understanding work equally as well if the use the correct value and wattage.

              Here's Rob Robinette's excellent layout diagram. I did this for one of my amps and it works well. I had to resort to dropping B+ as pdf64 suggested because of the plate and screen voltage gain when I initially raised the cathode resistor size a lot. The bigger cathode resistor pushed the plate and screen voltage a bit too high for my 6973 tubes. Just another thought on how and why you might want to adjust the B+ lower. You could do a little of both as you see fit.

              https://robrobinette.com/images/Guit...iode_Stack.pdf
              " Things change, not always for the better. " - Leo_Gnardo

              Comment


              • #22
                Or perhaps a string of Zener Diodes to reduce B+, just another thought, although resistors it's my understanding work equally as well if the use the correct value and wattage.
                The difference between the use of resistor(s) and zeners is that the resistor increases sag, zeners don't. So if you adjust for equal B+ at idle, there will be more compression and power loss (at full power) with the resistor (in a class AB or B design).
                E.g., if you use a 150R/40W! resistor to drop B+ by 30V at idle, the total voltage drop across the resistor might increase to over 40V at full power, while zeners would have a constant voltage drop. So the resistor solution gives less power output.

                The AC30 is a very hot class AB design, so the difference in current demand between idle and full power is rel. small (around 25%).
                With colder biased amps the difference between zeners and resistor will be stronger.

                Because of the high dissipation involved with B+ dropping, it's no real fun in amps having an output > 20W.
                - Own Opinions Only -

                Comment


                • #23
                  I noticed a typo in my post #22 above: Max. resistor dissipation in the example should read 10W (actually more like 11W). Power output will be reduced by around 6W and tube dissipation at idle will be lower by around 1.5W/tube.
                  Last edited by Helmholtz; 02-25-2020, 12:49 PM.
                  - Own Opinions Only -

                  Comment


                  • #24
                    OK. I changed the bias resistor to 73ohm. Now I have
                    - 355V on the plates
                    - about 340 on the screens
                    - 13V across the cathode resistor t

                    Thats still 125% hotter...
                    Change to 100ohm?

                    Comment


                    • #25
                      I'd leave it at that, if it sounds ok.

                      From your data I calculate a total dissipatin of 15.2W. As this contains the screen dissipation of around 1.7W, I get a plate dissipation of 13.5W or 112%.
                      - Own Opinions Only -

                      Comment


                      • #26
                        Okay, Finally, I have left 100ohm resistor there. Becuase after 5-6 minutes having 73ohm one I saw a bit of redplating in the dark...
                        After change to 100ohm the plate voltage was only 2-3 volts up (from 73ohm) and there was a ~14v across the resistor. That counts 100-105% plate dissipation now.

                        Now, two questions come to my mind:

                        1) Why in this amp I put such large (I mean, doubled from original design) cathode resistor and is it gonna sound worse or what in general are the downsides of that move?
                        2) The amp have speakers connected to + (green wire) and - (black) that comes from OT, but there is also second super-short black wire which is grounded to chassis, also wired to the minus along wit other black one from OT. I found it being the reason for a bit hum in this amp and disconnected it. Why would someone put minus to chassis ground?

                        Comment


                        • #27
                          The amp have speakers connected to + (green wire) and - (black) that comes from OT, but there is also second super-short black wire which is grounded to chassis, also wired to the minus along wit other black one from OT. I found it being the reason for a bit hum in this amp and disconnected it. Why would someone put minus to chassis ground?
                          Please post a picture showing the ground wire.
                          It is common practice to ground one of the OT secondary wires. It has a safety function.
                          You should be able to minimize hum by adjusting the "humdinger" (heater balancing pot).
                          I don't recommend to change the wiring.
                          - Own Opinions Only -

                          Comment

                          Working...
                          X