Announcement

Collapse
No announcement yet.

Quadrupling filamen to drive IC?

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

  • Quadrupling filamen to drive IC?

    Would this work to get a split supply(+/-) to drive some ic and still drive the filament without any problems(like hum).
    If the answer is yeas, how big should the 8 caps inte the two quadruplers be?
    Last edited by Tubis; 05-15-2008, 01:28 PM.

  • #2
    Should work fine. I've used a tripler to get +12V for channel switching relays. The filaments will still be centre-tapped to ground, so hum shouldn't be a problem in that respect.

    I only had hum troubles when I tried to run a digital reverb off the +12V too. You must keep the pulsating current that flows between the quadruplers and the filament centre tap under control, and make sure it stays out of the ground line of your signal circuit. I think my problem was due to ripple current flowing to ground through the reverb board's supply decoupling cap.

    The size of the caps will depend on how much current you want: I'd start with 470uF.
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

    Comment


    • #3
      So each of the 8 caps should be 470uf? Not counting the ones after the zeners.

      Comment


      • #4
        Yes. 470uF 6.3V caps shouldn't take up a great deal of room.

        The only other "gotcha" is that each multiplier stack only has 3.15V to drive it, and the 0.7V drop of the diodes actually wastes quite a lot of this. I developed another circuit that gets round this by driving the diodes with the full 6.3V, and posted it on the forum somewhere, but I can't remember where.
        "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

        Comment


        • #5
          Yea, the diods takes some but i want to get +/-9V after the zeners so i would have to quad the 3.15 anyway so the zener have something to work with.

          Comment


          • #6
            You know, if your op-amp is happy with any supply up to +/-30V (some are) or at least +/-15V, and has a really good PSRR, AND you filter the daylights on of the supply right at the chip (+ to -, and + to GND and - to GND), you don't technically need the zeners (but leave room for them).

            Your next question is, 'so what's a really good PSRR?' ... Look for the biggest number you can conveniently source.

            Hope this helps!

            Comment


            • #7
              Originally posted by Don Moose View Post

              Your next question is, 'so what's a really good PSRR?' ... Look for the biggest number you can conveniently source.
              I donīt know what PSRR stands for i canīt find it in the datasheet(tl072).

              And what is the difference in putting the caps from + to - instead of from + to gnd

              Comment


              • #8
                Power Supply Ripple Reject?

                (just guessing)

                Cheers,
                Albert

                Comment


                • #9
                  Mr. Kreuzer is right - Power Supply Rejection Ratio (in dB). It has been approximately forever that I went shopping for op-amps, and maybe that number lost popularity.

                  I agree that a cap from + to - along with +/- to GND seems redundant, but it's something I used to see a lot in instrumentation circuits for hypersensitive applications. I guess the idea is to make multiple paths for noise to disappear into.

                  Comment


                  • #10
                    TI specs 80 dB min, 86 dB typical for PSRR, which they call kSVR.

                    Caps to ground don't affect PSRR, since the IC has no ground pin. Caps betwen the supplies reduce noise between the supplies, improving PSRR. It does nothing for common mode noise on the supplies, and immunity to this noise is covered by CMRR, the common-mode rejection ration (which is really measured by changing the inputs common-mode, which is kind of like common-mode on both supplies. Again, TI specs 80 dB min, 86 dB max. caps from the supplies to ground will improve this performance even more.

                    I think the TL072 appeared around '81. They were pretty revolutionary with the JFETs up front. National just shook the earth and introduced new audio op-amps (first ones in years), pin compatible with TLO72 and most of the others, and thankfully available in DIP for us throwbacks. Part number is LM4562. They have a THD around .00003%, 140dB of open loop gain, equivalent input noise density of 0.65uV/rootHZ, 10Mhz full power bandwidth, 15V/uS slew rate, .00005% IM distortion, +/- .7 mV offset voltage, 110 dB PSRR and CMRR, drive 600 Ohm loads... all of which means they're pretty much perfect, and National can measure very small things, but they draw up to 12mA of supply current instead of 1.2mA.

                    I tried the LM4562 replacing TLO82s, and you can't hear them. DigiKey has them. Kind of pricey at a couple of bucks a piece.

                    Comment

                    Working...
                    X