Connect with us

A question for the hobby builders here

Discussion in 'Electronic Design' started by Larry Green, Nov 5, 2003.

Scroll to continue with content
  1. Larry Green

    Larry Green Guest

    Hi Folks,

    I wonder if any of the hobby project designers on here can help me with a
    hobby project headache I have created for myself ;-)

    I have been away from designing hobby circuits for a number of years and I
    am obviously very rusty. I have been asked to come up with a design for a
    hobby related circuit and while the majority of the circuit is fine (after
    some help from the guys here) the power supply side remains unfinished. I
    have spent over 70 hours researching power supply options and I am now
    suffering from major information overload to the point where I can no longer
    see the woods for the trees (I am also disabled and the brain fogs
    associated with that are not helping either!)

    Basically I have a battery powered device that uses a PIC microcontroller
    with a few switches and a piezo buzzer (plus a few diodes, caps, resistors
    etc.). I am sure there must be other people out there designing simple
    battery powered PIC projects but what are you guys using for a regulated
    supply for the PIC? The options I have looked at are below with comments as
    to my findings. I will be building on 'Vero' type protoboard BTW as I do not
    have the equipment to make my own boards and it is an extra cost which I
    cannot justify with this project.

    Linear Regulator
    Pros - Cheap, readily available, the old faithful regulator.
    Cons - Supply voltage is not high enough at 6V to give a regulated 5V
    output.

    LDO Regulator
    Pros - Fairly cheap, readily available.
    Cons - I am not sure if my supply voltage is enough to give a regulated 5V
    output.

    Charge Pumps
    Pros - Ideal for the job, readily available.
    Cons - Only available as tiny SMT components (MSOP), 'surf boards' are
    expensive, parts are too small for my aging eyes and hands to hand solder,
    increased expense.

    DC-DC Converters
    Pros - Ideal for the job, readily available in DIP packages.
    Cons - Difficult to source inductors, Schottky diodes and low ESR 'through
    hole' capacitors, increased expense.

    'Roman Black' Regulator
    Pros - Simple, effective, uses discrete components, cheap.
    Cons - Supply voltage is not high enough at 6V to give a regulated 5V
    output.

    So there are my findings! I have put a circuit diagram on the web so that
    you can see what I am trying to do and have an idea of load in the circuit
    etc. It can be found at the address below......

    http://www3.sympatico.ca/larry.green/circuit_diagram_v4.gif

    Do any of you guys have any suggestions/ideas/comments regarding a suitable
    power supply for this project.

    Any comments/help will be gratefully received as I am now at the point where
    I am starting to tear my hair out.....lol.

    Larry Green VE3YET
     
  2. Arch-lab

    Arch-lab Guest

    If you are going to have that range on your battery then you will have to go
    with some type of boost circuit.
    As you have already said, you ruled out the other regulators due to your 4.5V
    input and wanting 5V output... you just have to put up with what you called
    cons on your charge pump and DC/DC and pick which has the least cons now.

    How many of these are you going to build?

    As far as inductors go, they aren't that hard to find online etc.
    I also have found most datasheets will tell you what company and model number
    they used for the inductor.

    Low ESR parts etc. if this is just a circuit for yourself etc. you might want
    to test with the cheaper parts. I have had no problems when I didn't use ESR
    parts, so I would say weight what the mfg. says you gain vs. what the circuit
    actually does, and what you need it to do.

    SMT problem... hmmmmm... other family member who could do it? Or maybe I should
    sell you some pre-build ones ;)

    Or just make it a 6V battery...
     
  3. PICs can run at lower voltages than 5V, in particular, 4.5 volts is
    appropriate. They can go down to around 2V. So, I'd try it without any
    regulator. Just use a capacitor across the power/ground of the PIC to smooth
    out any spikes. The max for the PIC is 5.5V, so don't run it from 6V.

    Also, you can get 3.3V LDO regulators if you find that the PIC is not
    running properly. Run it from 4.5V, and you have it made. Try a Seiko S817
    series regulator. They have them for lots of voltages between 6 and 1.1V,
    and they have a very low dropout voltage (like 160mV at 5V using 10mA)

    you can get them at mouser, etc.

    Regards
    Bob Monsen
     
  4. "Larry Green" <> schreef in bericht

    [snip]
    Since you are still throwing away power with those leds and their 330
    ohm resistors, I assume you don't need to get the last drop out of
    the battery. A ldo regulator will work fine. If the device runs on
    4 batteries of 1.5V, I bet two diodes in series work fine too. The
    PIC won't mind. If you run it on 3 batteries, you don't need anything
    at all.

    For the leds, I'd use high-efficiency leds, with a 1K or 1K5 resistor.
     
  5. Are you sure your circuit won't run straight off the battery, no regulator
    required? Check the VCC specs for the PIC.
     
  6. Why regulate? If the PIC can function over the discharge curve of the
    batteries, it's not neccesary. Instead, you would will need an undervoltage
    lockout so that the PIC does not hang when the voltage becomes too low.
     
  7. Check It Out

    Check It Out Guest

    I would say use no regulator. If you use 6V, just put a diode in series with
    the pic VCC and strap a cap there, too.

    High efficiency led's with higher res. or pwming them would save power. I
    assume red led goes with buzzer sounding, could use that pwm for both.
    Efficcient use of SLEEP modes will save more, too.

    REALLY want a reg, use a LDO at like 3v for just the pic, change outputs on
    pic so they sink (active low) for the led and buzzer.
     
  8. First off, replace the 1N4002 with a 1N5817 for a half volt less drop.
    Or if you want to get zero drop, put the 1N4002 across the battery
    cathode to positive, and put a 1/10A fuse between it and the battery.
    If the battery is reversed, the fuse blows. But you get zero drop.

    Then the Q is why do you need a regulator to begin with? Why not run it
    straight from the battery?


    [snip]
     
  9. Larry Green

    Larry Green Guest

    Hi Folks,

    Thanks for all the very helpful replies to my original question. A number of
    you raised questions so if you don't mind I will try to answer them all in
    one post as some questions cover the same areas.

    "Arch-lab" asked/wrote:
    At the moment I have been asked to build two prototypes for 'field testing'.
    If they are successful then there could a need for 40-50 more.
    should
    sell you some pre-build ones ;)"

    LOL......nobody else in the family builds circuits but there could be a huge
    market out there for ready built PS units using SMT parts that could then be
    plugged into through hole proto boards. If I had the finances to set myself
    up making circuit boards and the ability to use reflow techniques I might
    even consider it myself...lol. Trying to hand solder 8 legs on a 3mm x 3mm
    chip was beyond my capabilities though!


    running properly. Run it from 4.5V, and you have it made. Try a Seiko S817
    series regulator. They have them for lots of voltages between 6 and 1.1V,
    and they have a very low dropout voltage (like 160mV at 5V using 10mA)"

    4 batteries of 1.5V, I bet two diodes in series work fine too. The
    PIC won't mind. If you run it on 3 batteries, you don't need anything
    at all."

    on
    pic so they sink (active low) for the led and buzzer.

    I am definitely begininning to think this is the direction I should head in
    as they are readily available, cheap and 'through hole'. The idea of running
    at 3.3V makes sense too as I can still use 4 'AA' cells. Finding a 3 cell
    holder is proving difficult (I am trying to source all the parts through as
    few suppliers as possible to save shipping costs/min. order charges etc. and
    there aren't that many online suppliers here in Canada (trying to avoid
    customs charges too!)) plus most rechargeable battery chargers only charge
    in 2 cell units so having 3 rechargeable cells could turn out to be a royal
    PITA!


    "Richard Henry" asked/wrote:
    regulator
    required? Check the VCC specs for the PIC."

    "Frithiof Andreas Jensen" asked/wrote:
    batteries, it's not neccesary. Instead, you would will need an undervoltage
    lockout so that the PIC does not hang when the voltage becomes too low."

    with
    the pic VCC and strap a cap there, too."

    "Watson A.Name" asked/wrote:
    straight from the battery?

    I originally considered this and posed the same question to Microchip's Tech
    Support group. This is part of the reply I got back.........

    "The PIC16LF818 can operate with Vdd from 2.0 to 5.5vdc, so operating
    with 2 or 3 batteries directly should work. 4 cells fully charged is a
    little
    high at 6vdc. 3 cells should provide 4.5v max(1.5v/cell) and 2.25v
    min(0.75v/cell).
    You have to understand some of the limitations of the PIC16LF818 under these
    variable
    VDD conditions. For example, the part is rated at 10MHz at 5vdc and is
    derated
    for lower VDD. This means that if you are expecting to run full speed, you
    may
    not be able to achieve this at 2.0vdc. If you limit your max speed at 4MHz
    then
    you are fine across the entire Vdd range when providing an external clock,
    using a crystal or external RC mode. If you decide to take advantage of the
    internal
    oscillator block, you can expect some variation in clock frequency over
    temperature
    and Vdd."

    This was one of the main reasons I wanted a regulated 5V supply. I did
    intend to use the internal oscillator block (to save costs......why add
    extra parts if it is already built in) and I will be using a pseudo random
    time delay generated by the PIC. Variations in the clock speed may have
    unpredictable results when generating the time delay. I want it pseudo
    random but it must fall between an upper and lower limit (1-60 seconds). I
    am now thinking that if I went with a regulated 3.3V supply and the chip
    does run slower then I can at least compensate for the error in the code.


    ohm resistors, I assume you don't need to get the last drop out of
    the battery.......For the leds, I'd use high-efficiency leds, with a 1K or
    1K5 resistor."

    assume red led goes with buzzer sounding, could use that pwm for both.
    Efficcient use of SLEEP modes will save more, too."

    The circuit I sent you to is actually one revision lower than the 'latest
    and greatest' version.........but it had the missing PS block which is where
    I directed the initial question. My latest design uses a single bi-colour
    LED to replace the two single ones. Both colours will now be controlled by
    the PIC and will flash to conserve power. My idea is this.......

    The flashing green LED is used to indicate power on. When the device goes
    into its pseudo random delay mode it will switch from a flashing green to a
    flashing red to indicate that it is 'thinking' (the operator needs to know
    that his button push actually started the timer).

    Once the delay time is over the red LED will go out and one of two tones
    will come out of the buzzer (a continuous tone or a series of beeps). After
    5 seconds of tone(s) the device will reset to the start condition and flash
    the green LED again to indicate that the power is still on. The operator
    should power down at this point until the device is required again in order
    to conserve battery power.

    Hmmmmmm.......I suppose I could even make it beep once every 30 seconds or
    so to remind people to turn it off!

    My original circuit was built around a 9V battery (PP3 type) with a 7805
    regulator but I was told in a previous post to this group that it was being
    very wasteful of resources. That was why I started looking at 3 or 4 'AA'
    cells and then hit the wall with regard to finding a suitable method of
    regulating the supply to 5V. If the batteries lasted for a full weekends use
    that would be fine. The device would NOT be on continously for that time but
    could be required dozens of times for up to 2 minutes per use.

    Once again thanks for all the help and at least I am beginning to see a way
    forward again ;-)

    Larry Green VE3YET
     
  10. Consider to use a colour-tv colourburst crystal for - it's 3.(sumething)
    MHz - for the oscillator or maybe a 33 kHz watch crystal instead. Depends on
    your speed requirements - the PIC can run down to DC.

    The colourburst xtal frequency is spec'ed to a great number of decimals,
    Five! (it think), and quite cheap for such a tightly-spec'ed xtal (just
    compare some of the other frequencies); the watch crystal is even cheaper
    *BUT* may vary a bit with temperature as it is designed for a constant 36'C.

    Nevertheless, it will be more accurate than the internal oscillator block!
     
  11. N. Thornton

    N. Thornton Guest

    no, extremely annoying. It will result in damaged units being returned
    and disgruntled customers. Not smart.


    Re your PS choice, how much does the internal osc frequency vary over
    2v to 3v? Variation may be small enough to be not a problem.

    Also you're doing random times: how would a limited shift in clock
    freq be a problem there? You should be able to make the times still
    fall in range.

    Thirdly, I dont know the int osc design, but in some cases you may be
    able to stabilise it to some extent with an external component like a
    varicap diode, or a zener, etc. Maybe, anyway.


    Regards, NT
     
  12. N. Thornton

    N. Thornton Guest

    Just a thought...

    Or put diodes/resistor onto a couple of input pins so the PIC can
    distinguish 2 or 3 batt voltage ranges, and compensate in software.

    Using the R drop on the LEd feed would give you an already trimmed V.

    Regards, NT
     
  13. You won't notice the difference. Run it off the 6V battery, one diode
    in series.
    Use a PNP in series with the plus. Put a momentary switch over the PNP, to
    turn on the circuit. The PIC's first job is to set an output low, to
    keep the PNP turned on - you can now release the 'start' button. Once
    the thing has timed out, it sets that output high, and turns off it's
    own powersupply. The battery will work for years.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-