marshallf3 - problems on your livewell circuit

Discussion in 'The Projects Forum' started by seabear, Mar 1, 2011.

  1. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    If anyone can answer, please do. A while back I got a diagram from "marshallf3" for a variable timing device for my live-well aerator (see ref below). I am trying to breadboard it one segment at a time to check function and delays and I have questions about the first segment 555 timer circuit. I find that my pin 3 output times are significantly different than what I should get as per some 555 circuit calculators that I have used. When I use all the suggested resistors (checked them with my digimeter), and just one capacitor (no toggle switches) at 82uf, I get an average output ON time of 12min35sec, and OFF time of 4min45sec.... and the calculator says it should be ON for almost 6min and OFF for 4.45min. Tried it again with 27uf instead and got 3min ON and 1.38 OFF and the calculator said it should be about 2 min ON and 1.30 OFF. I am concerned that when I get to the really high values (220 and 440 uf) I will be way off on timing. I tried swapping out the 555 chip and different caps. Volt source is regulated 12 supply.

    ref information::

    AS WRITTEN BY MARSHALLF3

    K1, 2 & 3 set the on interval and K4 & 5 set the run time per interval.

    In other words if you just switch on K1 you'll get around 1 hour, K2 only equates to 28 minutes, K3 to around 10 minutes. K1 + K2 = 1-1/2 hours, K1 + K3 = 1 hour, 10 minutes etc. Same applies to the others:

    K4 = 5 minutes, K5 = 2 minutes, both on would give you 90 seconds. The power transistor comes in a plastic case so it won't need an insulator but it will need heat sink compound and some way to dissipate the heat generated. If you choose a thick die-cast aluminum box it should handle it, if you choose thin metal or plastic one a decent size heat sink will have to be added. I prefer the cast ones anyway as they're far more durable and easy to seal. They are also available with mounting flanges and my personal choice would be a 546-1590TF

    The circuit should be good to drive at least 3A worth of 12V motor which I assumed would be plenty for a simple aerator pump. If it needs more current we'll probably want to add a driver transistor in there as I'm close to the maximum driving capability of the 555 IC if the transistor gain comes in at the absoulute minimum. Chances are it'll do 6A with no problem but when you're breadboarding something trying to replace a failed part can be a hassle.

    http://www.innoengr.com/examples/aerator_timer.jpg
    http://www.innoengr.com/examples/aerator_timer.xls


    In essence, when the circuit is powered up the output of the 555 goes high and triggers the second one through C6 which will run the motor for the selected time then turn it back off. As the first timer continues timing it will go into a low output state for a while then when the full amount of time has elapsed it again switches its output high and that pulses the second one into its timing cycle again.

    Switch K6 turns everything on or off, and K7 is a manual on switch which, if on, allows the motor to run all the time ignoring the timer circuit.

    Parts can of course be substituted, a 25V or 35V electrolytic will take the place of a 16V one but they're larger and more expensive. As always I do recommend using high temperature rated electrolytics, they tend to last about 10 times as long as the regular ones and only cost pennies more. Likewise the .01 uF caps can be ceramic or of a higher voltage, they're non-critical in their applications here. The values were carefully chosen to get as close to your desired times as possible using standard off the shelf values, and the part #s I used in the BOM were the least expensive options that were shown to be in stock at the time I chose them. I know you can find multitudes of less expensive 555 timer ICs but I tend to specify TI because #1 - never got a bad one and #2 - when they claim a 200 mA output capability you know you're going to get it. I've accidentally overdriven them by quite a bit a few times and they never complained.
     
  2. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    and by the way, I am assuming that the problem is something I am doing or not doing and not a fault of the excellent circuit diagrams from marshallf3.. (thank you very much for them!!)
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    What is the exact part number on your 555 timers that you have been using?
     
  4. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    its from TI and it is NE55P
     
  5. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    make that NE555P
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The original thread is here:
    http://forum.allaboutcircuits.com/showthread.php?t=41775
    Reading through the thread, appears that Marshallf3 didn't build the circuit, however 555 timer circuits aren't very complicated.

    I made a simulation of the circuit, and the times didn't quite add up to what Marshall thought they'd be either.

    Unfortunately, I'm really pushed for time today, so will have to visit this later. Marshallf3 hasn't been on since the 27th.
     
  7. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I take it back. The first complete cycle at power-up should take about 13.3 minutes, ON-time being ~9.1 minutes, OFF time being ~4.2 minutes. Subsequent cycles should be about 5.8 minutes on, 4.2 minutes off.

    The first cycle is longer because the timer cap C3 is completely discharged on start-up. After the 1st cycle, the voltage on the cap cycles between 1/3 and 2/3 Vcc, or 4v and 8v.

    It sounds as if your 82uF capacitor might be a few percent larger than specified (this is expected; aluminum electrolytic caps have a fairly wide tolerance) but the leakage rate is a bit high.

    Take C3, the 82uF cap, out of the circuit and connect it in series with a 1 MEG resistor across your 12v supply. Measure the voltage across the 1 MEG resistor over a period of time using a digital multimeter. If the capacitor has no leakage, the voltage should drop to zero. I'm betting you'll see around 0.63v across the resistor, even after waiting a good while (an hour or so).

    With large timing resistors and capacitors, even very small amounts of leakage in the cap will throw your timing off considerably. An ~18 MEG resistor in parallel with C3 will cause the kind of timing error that you're seeing - the leakage rate of the cap would be too small to be measured by an Ohmmeter, and would be acceptable for it's typical use as a filter capacitor.
     
    Last edited: Mar 2, 2011
  8. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    Thanks for the information, I will try the cap leakage test in a bit... So for my timing circuits would you suggest a different type of cap that is more accurate? What type would that be? thanks again
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Other types of caps in that uF range would be large and very expensive. Here's one that is $25.13, and 1-1/4"x1-1/4" x 13/16" (appx): http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=478-5356-ND

    [eta] Datasheet for that 82uF cap is here: http://www-images.panasonic.com/industrial/components/pdf/ABA0000CE22.pdf
    Acceptable leakage rate is <3uA. However, even 1/10 that amount would cause the timing error you're seeing.

    A least-parts-count (and very stable) solution would be a microcontroller, as proposed in the other thread.

    If you want to stick with a 555 timer, you'll really need to use much smaller resistor values (so that the leakage doesn't affect the timing as much), and have a divide-by counter on the output to slow the time down; something like a 4040 or 4060, or perhaps some 4017's chained together.
     
    Last edited: Mar 2, 2011
  10. MMcLaren

    Well-Known Member

    Feb 14, 2010
    759
    116
    I suspect a microprocessor based design is still a better choice for this application. If this is for a commercial product, I would be happy to discuss licensing terms and provide you with a prototype.

    Cheerful regards, Mike
     
  11. seabear

    Thread Starter Active Member

    Feb 20, 2009
    34
    0
    Ok, the cap test shows about .90 volts dropped across the resistor from a 10v supply using the 82uf cap, after about an hour.

    To MMcLaren and SgtWookie-- this was to be a one-off device just for me and something cheap so no thoughts of going commercial - - and I'm not real thrilled about the $25 caps.

    The original premise was to address my fishing boat sitting tied up at the dock overnight with a live-well containing my live minnows and maybe a few fish I didn't feel like cleaning that night. My live-well does have a switchable water pump which pumps fresh water in and sort of aerates at the same time (pretty standard). But I didn't want to leave the pump on all night or I would drain my 12v main battery. So I figured I could cut into the pump voltage line and put a circuit that I could set for a few simple settings depending upon the temperature at the time and the amount of fish in the tank. (e.g. turns on once per hour for duration of 5 min). If you all think I could do this more simply and still cheap using a micro proc, I guess I would need a bit more guidance in that area since I am not an EE. thanks
     
  12. MMcLaren

    Well-Known Member

    Feb 14, 2010
    759
    116
    Hi seabear,

    Sorry for late reply. I forgot to subscribe to the thread for notification.

    Depending on where you're located (you didn't fill in your location information), a few of us here on the forum might be able to help you with a one-off microcontroller based LiveWell Timer project (programming, etc.).

    I can't imagine parts being very expensive. The 8-pin PIC microcontroller is a couple dollars. A 7805 regulator and filter/bypass caps should be another couple dollars. Not sure what a pot or relay costs now days. You could build it on a $3 Radio Shack prototyping board. An enclosure may cost a bit.

    Cheerful regards, Mike
     
Loading...