Announcement

Collapse
No announcement yet.

Battery sucking device needed

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

  • Battery sucking device needed

    I'm open for suggestions on a boost regulator that will suck batteries dry. Preferably with self-starting oscillation AND sync input, so that it cooperates with the rest of the rig, preferably low side switcher (polarity is irrelevant - it will suck batteries), and preferably with polarity protection, and preferably not picky for the input voltage.

    My idea is to supply my rig with 12V regulated, and any ripple to be in sync with the rest of the rig, and to feed it with any battery available by the point they entirely give up the ghost. Say, a charge of 4 AA batteries, or a single LiION cell ... whatever.

    I have some ideas in my head, but it was 12 years ago that I played with such regulators and I'll appreciate someone with new mindset on the problem to join in.

    Any takers?

  • #2
    Isn't "sucking batteries dry" most relevant to dry cells? If you use rechargeables, NiMH or Li chemistry, they tend to dump quickly at the end of life, so sucking-dry will gain you very little. You should also be able to make the supply clean enough that synchronising becomes unnecessary.

    Comment


    • #3
      There are now quite a few boost regulators that will operate down to less than a volt for single-cell operation, so this is not a difficult challenge. Skippy is right, most batteries fall off a cliff at the end, so for a 4-cell (6V nominal) pack the difference between 3V and 1V is about 1 minute.

      Comment


      • #4
        Try the "Joule Thief" ->
        http://en.wikipedia.org/wiki/Joule_thief

        Comment


        • #5
          Yeah, that's the spirit

          I'm well aware of the batteries care and feeding, but for my rig to behave I'll need a fairly accurate 12V regulator and it will need to synchronise to a Tx oscillator once it starts. So trivial solution would be adding a battery or two and put a stupid 78xx regulator. But even then it will start misbehaving when the batteries go down.

          I can't foresee continuous use of the device - I don't have that much time, so a logical choice are off the shelf dry batteries that will be abused till the last moment when they completely give up the ghost.

          So the primary goal is to maintain the regulated voltage as long as possible, and then completely switch off. Not the long agony of slowly dying batteries, but a clean cut at the end.

          Initially I thought of something like a device suggested by Qiaozhi, but these are either self-oscillating, or externally triggered, and I need both.

          As Carl says, this is not a difficult challenge, so if anyone knows of such a device already in existence - please post the schematic here.

          Comment


          • #6
            What you probably want is a buck-boost regulator. Take a look at the LM5001, which I've used. It can maintain 12V out with down to 3V on the input. Can free run or synchronize.

            Comment


            • #7
              What about this boost converter:

              2A Converter with 150µA Quiescent Current (4.5V-8V to 12V)
              http://cds.linear.com/docs/Application%20Note/an29.pdf page 9
              Attached Files

              Comment


              • #8
                If you look on eBay, you will find many ready-built DC/DC converter boards based around common SMPSU IC's, often LT ones. It would help if you told us how much current output you want, what batteries you actually want it to work on, and what frequency you are wanting to sync it to.

                Comment


                • #9
                  My bad. Missed to mention the purpose explicitly.

                  It is supposed to supply an IGSL board, so it will need less than 200mA in peaks. Guess most of our toys are of the same appetite, so 2A is an overkill. Mustafa's circuit is nice, but I'm not sure how to sync it with my gear. In fact, due to the low power consumption a SEPIC solution with LM5001 could do just fine. Carl, if the circuit you mention is not a problem to share, please do :wink-nudge:

                  I'll look deeper into this, but if someone already has a circuit schematic of a working device - even better.

                  Comment


                  • #10
                    LT3580 is similar to LM5001.

                    Pin 5 Shutdown/undervoltage lockout
                    Pin 8 SYNC
                    Attached Files

                    Comment


                    • #11
                      Originally posted by Davor View Post
                      Carl, if the circuit you mention is not a problem to share, please do :wink-nudge:
                      I believe I used one of the TI apps circuit straight out of the data sheet. There are better buck-boost solutions, but they don't have a sync pin. And there are better sync'd boost regs, but they don't do buck-boost.

                      Comment


                      • #12
                        Yes, buck-boost SEPIC solution has its appeal. It is inherently short circuit protected. OK, it requires 2 coils, which makes a boost regulator a more straightforward solution.

                        Maybe I'll have to check with the local component dealers what they have, and go from there.

                        Any solution without coils? Switching capacitors?

                        Comment


                        • #13
                          Originally posted by Davor View Post
                          Any solution without coils? Switching capacitors?
                          Yes, but they tend to be very limited in I/O voltage ranges and the max output current. Most of the switched-L regs run at high f and therefore use very small L's.

                          Comment


                          • #14
                            So if I reduce my appetite to, say, 6V input (4 dry cells) I can benefit from low frequency operation and no coils. That could do.
                            My rig runs at ~8kHz, so L-s could become little monsters for operation that low. Even for as little as 200mA.

                            So in essence cascading two LM7660, synchronising them, and stabilising the resulting voltage with a 7812 will give me what I need in a quick and dirty way. I'm sure there is a better approach to this.

                            So now we have a somewhat different ballgame here. I need a simple solution to a switched capacitor step up voltage regulator that will accept 6V input and keep rock solid 12V @ 200mA. It will have to work on it's own, and when a device's oscillator kicks in it will have to synchronise to it at about 8kHz.

                            Comment


                            • #15
                              You can also synchronize to a harmonic of 8kHz, like 1.6MHz. Then the L's are small.

                              Comment

                              Working...
                              X