Announcement

Collapse
No announcement yet.

switching power supply for tube amps

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

  • switching power supply for tube amps

    There is a thread over on Doug's forum about a battery amp. The original poster wanted to get rid of heater hum and the like but I took it in a different direction.

    I gutted a tiny little cig lighter 12v to 120v 100watt inverter. It had two halves with only a transformer connecting them. I desoldered the transformer then cut away the 120v half of the PCB and was left with a current limited 12vdc in 275vct out supply. The 275vct out was in the form of a 100khz square wave- so I connected some diodes and now I have a DC powered 275 volt supply for tube amps.

    I hooked the thing up to a small amp I built a while back- it's got a 5879 out front, a gain stage and CF ala marshall, a LTPI and a cathode biased pair of 6aq5's. I only used the battery to supply the B+ voltage- future versions will supply B+ and filaments.

    It sounded great, better than I expected. I was rocking out for about 45 seconds when the my cheapo bridge rectifier blew up. I swiped some ultrafast diodes from a junk computer supply and built a new bridge.

    The inverter was surprisingly stiff. It idled around 71 ma, rising to around 95 ma with the output section clipping heavily- voltage dropped only about a volt between rest and maximum output. I might have to reduce the 1k screen grid and 1k screen supply dropping resistors and try for more power! On the battery side of things I was drawing 1.75 amps at rest and about 2.5 at full load. Very nice!

    My goal is to build a small cathode biased tube amp with 8 to 15 watts that can run using one or two valve regulated lead acid batteries. With this little inverter I should be able to make something that draws between 3.5 and 4.5 amps total and have a nice long run time of 2-3 hours using inexpensive surplus rechargable batteries. A smallish pedalboard will cover the back of the combo when not in use so I can leave the thing in my trunk for impromptu jam sessions. I could include an inexpensive 6 amp battery charger or 12v supply to use the amp "plugged in" when needed.

    Here is my other thought- why haven't we been using switch mode stuff for tube amps? I'm about to build a large bass amp and it would seem that a very stiff B+ and screen supply would be a wonderful complement to a large output transformer while cutting back on the heft required to run 4 to 6 kt88's at high voltages.

    A switching supply for a guitar amp needn't be advanced. I'm picturing something not too far off from a bigger version of my inverter- little in the way of voltage feedback, basic current limiting in the event of a dead short, windings for 6.3 volts, 400 volts and 650 volts. Perhaps a rewound computer transformer or two.

    I've read about using dimmable electronic transformers but I thought I'd get some opinions before I try to build anything else.

    jamie

  • #2
    Sounds very interesting.
    I have been toying with a similar idea for a long time, but now that you graciously offer to do the research work , you are welcome!!
    I have more modest ambitions, would just love to turn some of all those PC supplies that get junked every day into stiff +/-40V supplies for my 100W SS amps. At least in theory, they should do fine.
    Keep us informed.
    PS: I guess you will feed 6 or 12 VDC to those filaments, not 100KHz squarewaves
    Juan Manuel Fahey

    Comment


    • #3
      Originally posted by imaradiostar View Post
      Here is my other thought- why haven't we been using switch mode stuff for tube amps? I'm about to build a large bass amp and it would seem that a very stiff B+ and screen supply would be a wonderful complement to a large output transformer while cutting back on the heft required to run 4 to 6 kt88's at high voltages.
      Why haven't we indeed?

      1. It's more complicated and harder to troubleshoot. Even professionals get SMPS designs tragically wrong. Not naming any names, but, ahem, Coldamp.

      2. You need to know math. If you're willing to settle for an unregulated SMPS, that helps, because you don't need to know control theory.

      3. It's less reliable. A well-designed SMPS can be perfectly reliable, but not many people know how to design one well.

      4. A big bass amp needs a humongous, heavy speaker cabinet. You're going to get a workout carrying that anyway, so what's one more transformer?

      5. Transformers can be protected by a fuse, but SMPS transistors sacrifice themselves to save the fuse.

      6. (Corollary of 5) Transformers can be undersized because of their large short-term overload capacity.

      7. Keeping ultrasonic crap out of the audio path. If you thought ground loops were hard...

      Having said all of this, it can be done. I know a guy who designed and built a 500 watt Class-D bass rig with a switching power supply, as a hobby project. However, he works on the accelerator at Fermilab and is something of an egghead.
      "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

      Comment


      • #4
        so not worth it?

        I was picturing a switch mode supply in a small box that would mount outside the amp chassis.

        I don't know if I lucked out or what but I'm not having any problems so far with the battery switcher. Maybe I got lucky.

        I also thought it would be neat to modify old computer power supplies and use them for guitar amps.

        Steve- you've got to admit though, running a tube amp from a large battery, however impractical, is pretty neat. I'm looking forward to showing up somewhere and confusing people.

        jamie

        Comment


        • #5
          Well, I think your project is neat. But in the general case, I'm not sure that SMPS are worth it.

          I'm surprised that you cut the 120 volt bit off, since it would have contained a bridge of fast diodes and a smoothing cap. That's how these things work, they go from 12V DC to high frequency AC, step it up, rectify it to 160V DC, then chop it back up to 120V AC with another inverter.

          I've taken the ferrite transformer out of a PC power supply and run it backwards as a step-up before. It'll give you 160 or 320V DC after rectification, give or take. I used a self-oscillating push-pull topology, what oldies would call a Royer oscillator, with two big power MOSFETs.
          "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

          Comment


          • #6
            Although not many find that a reason, I always suspected SMPS trouble killed the otherwise excellent Crate Power Block 150.
            Some described a "tick tick tick"noise and amp refusing to stay on, which smells strongly of that.
            Even if 90% of them work perfectly, 10% end-user failure kills any product.
            I am thinking of using a simple unregulated push pull or half bridge switcher-PSU, as (un)regulated as any conventional iron power supply.
            EDIT:
            Hey!!, that's about what I want to do: rectify 220 (or double 120) VAC to get around 300VDC and chop and transform that into, say, +/- 40 to 50 VDC.
            If you have something to suggest, you are welcome.
            meanwhile I'll google that "Royer oscillator" stuff.
            Last edited by J M Fahey; 04-21-2010, 03:32 PM. Reason: Read simultaneous post
            Juan Manuel Fahey

            Comment


            • #7
              Originally posted by J M Fahey View Post
              Hey!!, that's about what I want to do: rectify 220 (or double 120) VAC to get around 300VDC and chop and transform that into, say, +/- 40 to 50 VDC.
              If you have something to suggest, you are welcome.
              meanwhile I'll google that "Royer oscillator" stuff.
              check out the Vicor front ends:
              VICOR front end VI-FKE6-CUX 250W 300VDC 47-63HZ (3 UNIT - eBay (item 360191009086 end time May-17-10 14:24:27 PDT)


              or the much smaller units:
              VICOR AC INPUT FRONT END MODULE VI-AIM-C1 (2 PCS) - eBay (item 190390040849 end time May-18-10 09:17:17 PDT)

              which are unisolated rectifier circuits with ~50mV ripple in the HV DC

              matched to a few of their dc/dc converters, or the RO Associates or Astec AMPSS line.

              I just got 2 of these:
              ~$15 for 2 x 24VDC at 10A for a chip or class D amp
              and you can use the 300vdc for the preamp!

              Comment


              • #8
                Thanks, although I'd rather build them myself.
                Remember you can not use *those* 300V for the preamp because they are "mains live"
                Of course you can add an extra winding and get as much isolated safe +B as you like.
                Thanks.
                Juan Manuel Fahey

                Comment

                Working...
                X