Announcement

Collapse
No announcement yet.

Capacitor Selection: Electrolytic v Poly

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

  • Capacitor Selection: Electrolytic v Poly

    Sorry if this issue has been covered before, but I'm making some speakers for a friend which have a 50 microfarad capacitor in series in the tweeter circuit; however, the poly capacitor called for in the design costs more than the tweeter! The tweeter circuit is a third order with the 50 microfarad before the tweeter.

    Would a NPE do the job instead, please? Is it possible to mix NPEs and poly caps in series?

    I know that it's OK to use an NPE in parallel, but I've not seen one used in series.

    Thank you

    Geoff

  • #2
    Done that - with the appropriate Working Voltage ratings
    * I've paralleled NP with poly ( aka bypassing/cascading ) to build up to large value - the NP providing approx 90% of target value
    "Not a Speaker Designer - Not even on the Internet"
    “Pride is your greatest enemy, humility is your greatest friend.”
    "If the freedom of speech is taken away, then dumb and silent we may be led, like sheep to the slaughter."

    Comment


    • #3
      I'd be interested to see that schematic. Not sure how/why a series cap in an HP filter would be that large for a tweeter (mid in a 3-way, more understandably).
      EYE would go w/a quality npe (the ones PE sells are pretty good) for a value that large. You CAN mix & match. SOME think a 25uF npe + a 25 poly would be better (or even a 40 npe + a 10uF pp). The 2 caps would be paralleled (so their values add) of course.

      Comment


      • #4
        Didn't I see someone somewhere along the way say you could have the sound of a poly by doing just what Sydney did???
        http://techtalk.parts-express.com/fo...khanspires-but
        http://techtalk.parts-express.com/fo...pico-neo-build
        http://techtalk.parts-express.com/fo...ensation-build

        Comment


        • #5
          Originally posted by Kornbread View Post
          Didn't I see someone somewhere along the way say you could have the sound of a poly by doing just what Sydney did???
          North Creek Music technique
          "Not a Speaker Designer - Not even on the Internet"
          “Pride is your greatest enemy, humility is your greatest friend.”
          "If the freedom of speech is taken away, then dumb and silent we may be led, like sheep to the slaughter."

          Comment


          • #6
            Originally posted by Chris Roemer View Post
            I'd be interested to see that schematic. Not sure how/why a series cap in an HP filter would be that large for a tweeter (mid in a 3-way, more understandably).
            EYE would go w/a quality npe (the ones PE sells are pretty good) for a value that large. You CAN mix & match. SOME think a 25uF npe + a 25 poly would be better (or even a 40 npe + a 10uF pp). The 2 caps would be paralleled (so their values add) of course.
            Such quick replies!

            I can't post the schematic as it's the designers' intellectual property, the design choice was based on drivers already on hand, I searched for a DIY design which used them. I haven't yet modelled the design in Xsim, I'll give that a try also.

            Thank you all again for being so helpful and knowledgeable

            Cheers

            Geoff

            Comment


            • #7
              Originally posted by Geoff Millar View Post
              a 50 microfarad capacitor in series in the tweeter circuit;
              That's nuts, what tweeter goes to 800Hz or thereabouts? In any event if you go with NPE and poly parallel the usual value ratio is 10:1. It's not ideal, as the poly will have a lot lower ESR than the NPE, but it's better than an NPE alone.

              www.billfitzmaurice.com
              www.billfitzmaurice.info/forum

              Comment


              • #8
                So a cheap tweeter crossed around 800 Hz? A electrolytic should be fine Seriously though the electrolytic could add some graininess but that might not be that noticeable depending on the rest of the design.
                John H

                Synergy Horn, SLS-85, BMR-3L, Mini-TL, BR-2, Titan OB, B452, Udique, Vultus, Latus1, Seriatim, Aperivox,Pencil Tower

                Comment


                • #9
                  Originally posted by billfitzmaurice View Post
                  That's nuts, what tweeter goes to 800Hz or thereabouts? In any event if you go with NPE and poly parallel the usual value ratio is 10:1. It's not ideal, as the poly will have a lot lower ESR than the NPE, but it's better than an NPE alone.
                  The crossover point is 3,000Hz: but I've played around in Xsim and reducing the capacitor value doesn't seem to have bad effects on the FR, given a little tweaking of other parts. However, the designer's measured FR with the drivers in the cabinet looks really nice.

                  Geoff

                  Comment


                  • #10
                    Originally posted by Geoff Millar View Post

                    Such quick replies!

                    I can't post the schematic as it's the designers' intellectual property, the design choice was based on drivers already on hand, I searched for a DIY design which used them. I haven't yet modelled the design in Xsim, I'll give that a try also.

                    Thank you all again for being so helpful and knowledgeable

                    Cheers

                    Geoff
                    Two regular (polarized) electrolytic capacitors can be connected in series (back to back) which then basically becomes, functions the same as a Non Polar Electrolytic. When done this way, you end up with half the capacitance. Two 50uF in series gives you 25uF total capacitance.
                    Last edited by AEIOU; 08-25-2018, 01:23 AM.

                    Comment


                    • #11
                      A 50 uF cap in a tweeter circuit as you've described is not "nuts" as Bill said or whatever Chris was trying to say with those funny capitalized words. It's a third order electrical HP filter with a small series cap first, followed by a shunt inductor, then the series 50 uF. I've used that topology on tweeters several times. The first cap and inductor set the corner frequency and Q. The large cap tailors the slope and adds 90 degrees of phase rotation which can make a huge difference in the phase relationship between the drivers but has little effect on the on-axis frequency response. Model it in XSIM or WinPCD. Look at the response, phase tracking, and reverse null with and without that cap in the circuit. I think you will see why the designer uses it.

                      As for whether you should use a poly cap or a non-polarized electrolytic... your budget kinda determines that. If I'm using very nice tweeters I buy the polys. If I'm using a $15 tweeter I might lean towards the NPE.
                      Craig

                      Comment


                      • #12
                        Originally posted by PWR RYD View Post
                        A 50 uF cap in a tweeter circuit as you've described is not "nuts" as Bill said or whatever Chris was trying to say with those funny capitalized words. It's a third order electrical HP filter with a small series cap first, followed by a shunt inductor, then the series 50 uF. .
                        How is it better than a conventional 3rd order high pass? Especially as a typical 3rd order HP at 3kHz with an 8 ohm tweeter wouldn't use more than a 16uF cap in the circuit.
                        www.billfitzmaurice.com
                        www.billfitzmaurice.info/forum

                        Comment


                        • #13
                          Did you even try modeling it? And not every tweeter is 8 ohms yoda.
                          Craig

                          Comment


                          • #14
                            Originally posted by PWR RYD View Post
                            A 50 uF cap in a tweeter circuit as you've described is not "nuts" as Bill said or whatever Chris was trying to say with those funny capitalized words. It's a third order electrical HP filter with a small series cap first, followed by a shunt inductor, then the series 50 uF. I've used that topology on tweeters several times. The first cap and inductor set the corner frequency and Q. The large cap tailors the slope and adds 90 degrees of phase rotation which can make a huge difference in the phase relationship between the drivers but has little effect on the on-axis frequency response. Model it in XSIM or WinPCD. Look at the response, phase tracking, and reverse null with and without that cap in the circuit. I think you will see why the designer uses it.

                            As for whether you should use a poly cap or a non-polarized electrolytic... your budget kinda determines that. If I'm using very nice tweeters I buy the polys. If I'm using a $15 tweeter I might lean towards the NPE.
                            Thank you, it models well in Xsim, although reducing the value of that capacitor doesn't seem to change the FR that much. However, I'm very new to speaker building and using Xsim, so there's more to this than meets my eye. And there's much more to Xsim than just FR.

                            I appreciate it would be easier to comment if I'd posted the crossover schematic, but I couldn't do that out of respect for the designer and his IP.


                            Thanks (as always) to the advice from PETT members I think I now know how to go about this.

                            Cheers

                            Geoff

                            Comment


                            • #15
                              Standard analog filters would not be protected, assuming you are not sharing (or haven't elsewhere) the drivers, box, etc.
                              John H

                              Synergy Horn, SLS-85, BMR-3L, Mini-TL, BR-2, Titan OB, B452, Udique, Vultus, Latus1, Seriatim, Aperivox,Pencil Tower

                              Comment

                              Working...
                              X