Charge pump sanity check

Discussion in 'General Electronics Chat' started by Maccara, Feb 10, 2012.

  1. Maccara

    Thread Starter New Member

    Aug 22, 2009
    Hi, complete newbie in electronics trying to learn a bit about transistors and CMOS 555 circuits.

    I attached schematic what I came up with so far, but I'm not able to build/test it for the moment (I realize I would learn better by doing that) so I would like to have a sanity check if it is something that actually should even work on a breadboard without blowing up the transistors.

    It is just a basic charge pump which could be directly driven from 12V battery with output of NE555 without the transistors q1/2 & m1/2 and have some ten(s) of mA usable on the negative rail.

    I'm trying to make something working down to 5V supply and have as little voltage drop as possible on the outputs and possibly little better driving capability on the negative rail. Therefore ICM7555 and the use of MOSFETs & schottkys.

    Transistor models currently chosen are mostly random what was available from LTSpice library, but as I understand it, M1&2 need to be logic level MOSFETs if I hope them to be switched with 5V and the purpose of Q1&2 is to minimize the switching time since ICM7555 has pretty limited sinking/sourcing capability. It is likely this will not work well from just 5V and little higher supply is needed?

    I parametrized the schematic to play with LTSpice, but I give values below which seem to work ok in simulation. They're mostly ballpark choices (for now) to get simulation going.

    R1 = 1M, R2 = 10M, C1 = 1.2p
    R3 = 470 Ohm
    C5&8 = 40u (arbitrary choice)

    Real use of 7555 in astable config might need bigger capacitance (C1) and smaller resistors (R1&2) to be stable outside of simulation?

    Do I have things wired up right? Is it safe to use schottky's like that? Other handy tips to make this something that would actually be buildable/usable? (this seems to work in LTSpice, but since I know so little myself, I don't trust my interpretation of the results, therefore asking for sanity check)

    I realize there are ready chips to do the same (inverter/voltage doubler) more efficiently, but what is the fun in that. :)
    Last edited: Feb 10, 2012
  2. jimkeith

    Well-Known Member

    Oct 26, 2011
    Looks to me like it will work OK. The only potential problem I see is that during the transition, all transistors will be conducting simultaneously and will tend to 'glitch' your power rail for perhaps 50 to 100ns--peak current could be in amps--you might want to consider a gate driver to do this. Of course, C1 will have to be in the range of about 100 to 1000pf. Yes, this is the proper application of schottky rectifiers.

    After you are done with that one, check out the schem I just came up with as I was thinking about yours. This one is for strictly 5V and below--it uses a cheap logic buffer array with all sections connected in parallel. Diodes Inc /Zetex makes some dandy high current bipolars that are well suited for synchronous rectification for low voltages (5V and below). Of course the ubiquitous 2N4401 will work too. It all works in my cranial simulator--do not know if it will work in the real world.
  3. Maccara

    Thread Starter New Member

    Aug 22, 2009
    Thank you very much for your input!

    Now that you mentioned it, I did originally think about all of the transistors conducting at the same time, but forgot about it completely while figuring out how to put everything together...

    Your schematic will be nice one to think over (at first glance I think I mostly understand how it is supposed to work, but I'll ask if I don't :)).