Loading ...
Sorry, an error occurred while loading the content.

Re: Spark unit - coil driver

Expand Messages
  • Steve
    This page has a great 555 javascript calculator, you can enter in scientific notation in the form 390e-12 for 390pF. And it will let you enter values that
    Message 1 of 7 , May 3, 2008
    • 0 Attachment
      This page has a great 555 javascript calculator, you can enter in
      scientific notation in the form 390e-12 for 390pF. And it will let you
      enter values that other calculators won't allow.

      http://www.csgnetwork.com/ne555timer2calc.html

      It comes up with 5.4uS On time, 2.7uS Off time, 123KHz.

      For a spark coil, you want a fixed On time, and if you want to vary
      the power, vary the Off time. But with the 555 timer as connected in
      this circuit, varying the upper resistance varies both On and Off
      times, varying the lower resistance varies just the Off time.

      I see so many problems with this, I am daunted by the prospect of
      typing it all in. No offense meant, but perhaps a less complex project
      as your first try?

      Steve Greenfield

      --- In Electronics_101@yahoogroups.com, Ben <ben@...> wrote:
      >
      > Hi folks,
      >
      > I'm trying to get a spark unit working. The essential problem is to
      > have a box that hooks up to a 24V current source and outputs a pulsed
      > signal to a spark coil (from an Otto cycle engine, like a car), which
      > then makes sparks. The challenge is that the voltage spike (aka back
      > EMF) that comes off the coil when the pulse ends must be guarded
      > against, or it will fry the current switching element and cause noise
      > in other circuits hooked up to the battery.
      >
      > I got a schematic for just such a thing from a friend, and he said it
      > is known good. In essence, it is a 555 for timing driving a MOSFET.
      > This is a 12v circuit, I was trying to get it working before I adapted
      > it for 24v. I uploaded it to the group's files area, it is
      > sparkschematic.JPG in 555_Timer_IC. http://tinyurl.com/6bul57 or
      >
      http://groups.yahoo.com/group/Electronics_101/files/555_Timer_IC/sparkschematic.JPG
      > should work.
      >
      > I made two changes to it. As it is the 10k pots would just be 10k
      > resisors. I connected the center tap to one side on each, so they both
      > go from 0-10k. They are to set the timing on the 555, which is very
      > fast in this circuit. I also substituted the Zener with a higher
      > voltage one, planning ahead to move to 24V, but from what I've read on
      > zeners the one in this circuit isn't actually doing anything anyway.
      >
      > I've looked at it with a digital scope, and the 555 is oscillating
      > fine. With the scope between the drain of the MOSFET and +V, with no
      > load, it doesn't look like the voltage drops all the way down on the
      > off part of the square wave. Running it on a variable power supply I
      > got it to make sparks, but only by limiting the current. On a battery,
      > the MOSFET just gets hot.
      >
      > On a MOSFET, the pins are Gate-Drain-Source, when looking at the
      > plastic side, correct? And in this schematic, the left connection
      > would be G, top D and bottom S, right?
      >
      > Here's the list of parts I bought, by Digikey p/n. It's possible I
      > made an incompatible substitution; Digikey isn't always very clear
      > about what the differences are in similar part numbers.
      >
      > IRFP450APBF-ND MOSFET N-CH 500V 14A TO-247AC
      > 568-3430-5-ND DIODE RECT UFAST 200V TO220AB
      > 1N4752AFSCT-ND DIODE ZENER 33V 1W 5% DO-41
      > 1N4007FSCT-ND DIODE GPP 1A 1000V DO41
      > ICM7555IPAZ-ND IC TIMER RC GEN PURP CMOS 8DIP
      > 575SF103-ND POT 10K OHM COND PLAS LIN TAPER
      > 490-4244-ND CAP CER 10000PF 2KVDC E3 RAD
      > 490-3840-ND CAP CER 390PF 50V X7R RADIAL
      >
      > The resistors are all 5% carbon film 1/2 watt.
      >
      > I bought enough parts to make 10 of these, so I'd really like to get
      > this circuit working if possible, so as to not end up tossing them all
      > in the misc parts box. If anyone has any ideas for
      > improving/testing/debugging, please let me know.
      >
      > I also tried changing the resistors and caps on the 555 to run it at a
      > lower frequency (around 60hz), no luck. I started on a solderless
      > breadboard but ended up soldering it to perfboard to make sure it
      > wasn't stray capacitance issues.
      >
      > Thanks,
      > Ben
      >
    • Ben
      ... Vehicle main power bus is 24V, plus you get a bigger spark out of a 12V coil by running it on 24V (for a reasonable duty cycle). My plan was to run the 555
      Message 2 of 7 , May 3, 2008
      • 0 Attachment
        > First question I have is, why do you want to run that circuit off 24V? I
        > just looked at the datasheet I have handy for the 555, which in my case is
        > National's LM555, and they give an absolute maximum supply voltage rating of
        > 18V, which means that you're exceeding it already by a significant margin.
        > I don't think I've ever seen one running at more than 15V supply or so,
        > often less.

        Vehicle main power bus is 24V, plus you get a bigger spark out of a
        12V coil by running it on 24V (for a reasonable duty cycle). My plan
        was to run the 555 part either off the 5v signal power bus or use a
        zener diode or dc/dc converter to supply power to the 55, while
        leaving 24V across the MOSFET and coil.

        > What you can do, if you need to use a 24V supply for some reason, is to use
        > a zener diode or 3-terminal regulator to drop the supply voltage to the 555
        > chip to something lower, like 12V or 15V. I wouldn't go much lower than
        > that because you want plenty of gate drive for the MOSFET.

        I'll try that when I get to that point.

        > That network used there is interesting, though I'm not clear why they do
        > things that way. I've seen diodes across a device like the MOSFET (and
        > indeed some of those have the diode built-in, though those are reportedly
        > not very fast), and I've also seen the diode across the coil being used in
        > other situations.

        Yeah, it's a bit different than other designs I've seen for the same
        purpose, but I was hoping that the person who designed it knew what
        they were doing. It remains to be seen.

        > As far as not putting noise from this circuit into others, a decoupling
        > network consisting of a bypass capacitor on the incoming power and some small
        > value of resistance in series with the power (low enough to not cause a major
        > voltage drop) should help.

        Is there a rule of thumb or set of equations for sizing those
        components? I looked at the noise coming off another working spark
        unit and it was going over 50V on the spikes.

        > So did you try it on 12V first?

        Yes, all the testing so far has been on 12V.

        > It's not clear to me why pots are specified in there in the first place.

        To vary the frequency into the coil. I believe they were trying to get
        some kind of resonance going. In the final design I'll be replacing
        them with fixed resistors, once I've decided what frequency I want.

        > > They are to set the timing on the 555, which is very fast in this circuit.
        >
        > I haven't done the math, what speed is it operating at?

        Theoretically, 112khz to 1.23Mhz. I was running it on the lower end
        when I was running it with those values installed.

        > Do you mean one of those "pc scopes" or something different here?

        I was using a two channel portable Fluke oscilloscope, I don't recall
        the model. Now I have a Tektronix TDS 2002B two channel oscilloscope.
        They can be hooked up to a computer to transfer stored data, but are
        standalone scopes.

        > I don't think you can do that with no load, the MOSFET being basically a
        > varying resistance I'm surprised you see any change in voltage there at all,
        > with the "load" in this case being the input impedance of your scope or
        > something? If you want to try something, try substituting some value of
        > resistance in there instead of the coil...

        I wondered about that. I'll try that test again and report back.

        > > Running it on a variable power supply I got it to make sparks, but only by
        > > limiting the current.
        >
        > Limiting it to what? And at what voltage?

        The power supply was unhappy about it, and I didn't have other meters
        hooked up to it to see what it was really doing, but... it was set for
        12V, but turning up the current knob it was sparking at 2 or 3 A and
        displaying something in the area of 5V. Twiddiling the voltage knob
        didn't change it.

        >
        > > On a battery, the MOSFET just gets hot.
        >
        > What do voltage measurements and the scope show you then?

        A small change in voltage, but no real oscillation between drain and +V.

        > I'd say refer to a datasheet for the part, and though I looked I don't seem
        > to have one for that particular MOSFET, though I have lots of others.

        I did, and that's what it says, I just wanted to make sure that didn't
        sound unusual.

        > Did you get actual pots or trimmers? No matter, overall...

        It's a pot.
        http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=575SF103-ND

        > Any change in behavior at all?

        Oscillator still oscillates, at the lower frequency, spark still doesn't spark.

        > I don't see stray capacitance as likely being an issue with this circuit,
        > mostly I'd worry about that with RF stuff, or very high impedance stuff
        > where leakage can become an issue, though some folks tend to have
        > intermittent or incorrect wiring issues with those things.

        Well, with the cap on the 555 being only 390pf, I figured it couldn't hurt.

        Thanks for the constructive response, I'll do some more experiments
        and get back to you.

        Ben
      • Ben
        ... Yeah, it did look like a fairly useless implementation. ... There is a 1k resistor in series with each pot on that circuit. Ben
        Message 3 of 7 , May 3, 2008
        • 0 Attachment
          On Fri, May 2, 2008 at 8:33 PM, rtstofer <rstofer@...> wrote:
          > If that zener ever fires, it will go up in smoke! What is going to
          > limit the current? The battery? You need some kind of current
          > limiting device in that branch.

          Yeah, it did look like a fairly useless implementation.

          > I don't like circuits where pots can actually go to the rail. There
          > is no way to limit current. It is better to put a 1k fixed resistor
          > in series with the pot to make sure there is always SOME minimum
          > resistance.

          There is a 1k resistor in series with each pot on that circuit.

          Ben
        • Roy J. Tellason
          This is my second attempt at a reply, a power loss caused by a 16-month-old lost my first one...! ... Ok, though 12V would probably be better than 5V, since
          Message 4 of 7 , May 3, 2008
          • 0 Attachment
            This is my second attempt at a reply, a power loss caused by a 16-month-old
            lost my first one...!

            On Saturday 03 May 2008 19:07, Ben wrote:
            > > First question I have is, why do you want to run that circuit off 24V?
            > > I just looked at the datasheet I have handy for the 555, which in my
            > > case is National's LM555, and they give an absolute maximum supply
            > > voltage rating of 18V, which means that you're exceeding it already by a
            > > significant margin. I don't think I've ever seen one running at more than
            > > 15V supply or so, often less.
            >
            > Vehicle main power bus is 24V, plus you get a bigger spark out of a
            > 12V coil by running it on 24V (for a reasonable duty cycle). My plan
            > was to run the 555 part either off the 5v signal power bus or use a
            > zener diode or dc/dc converter to supply power to the 55, while
            > leaving 24V across the MOSFET and coil.

            Ok, though 12V would probably be better than 5V, since you want to make sure
            that you have sufficient gate drive ofor the MOSFET. Not enough and it'll
            run hot. I wouldn't run the 555 on more than 12-15V either, so you have a
            couple of constraints operating there.

            > > What you can do, if you need to use a 24V supply for some reason, is
            > > to use a zener diode or 3-terminal regulator to drop the supply voltage
            > > to the 555 chip to something lower, like 12V or 15V. I wouldn't go much
            > > lower than that because you want plenty of gate drive for the MOSFET.
            >
            > I'll try that when I get to that point.

            Ok.

            > > That network used there is interesting, though I'm not clear why they
            > > do things that way. I've seen diodes across a device like the MOSFET
            > > (and indeed some of those have the diode built-in, though those are
            > > reportedly not very fast), and I've also seen the diode across the coil
            > > being used in other situations.
            >
            > Yeah, it's a bit different than other designs I've seen for the same
            > purpose, but I was hoping that the person who designed it knew what
            > they were doing. It remains to be seen.

            Yup.

            > > As far as not putting noise from this circuit into others, a decoupling
            > > network consisting of a bypass capacitor on the incoming power and some
            > > small value of resistance in series with the power (low enough to not
            > > cause a major voltage drop) should help.
            >
            > Is there a rule of thumb or set of equations for sizing those
            > components? I looked at the noise coming off another working spark
            > unit and it was going over 50V on the spikes.

            That may be some reason for the zener across the power input. I typically use
            a series resistor and a bypass cap (to ground) for decoupling. The limit
            there is that you don't want too high a value of resistance or you lose
            voltage supplied to the circuit. Too small and it's ineffective. Take the
            time constant of the two parts (RC, where R is in ohms and C is in _farads_)
            and the inverse of that (1/RC) should be a frequency with at least five times
            longer a period than the operating frequency of the circuit.

            > > So did you try it on 12V first?
            >
            > Yes, all the testing so far has been on 12V.

            Ok.

            > > It's not clear to me why pots are specified in there in the first place.
            >
            > To vary the frequency into the coil. I believe they were trying to get
            > some kind of resonance going. In the final design I'll be replacing
            > them with fixed resistors, once I've decided what frequency I want.

            I've seen something of the sort done before, particularly involving battery
            desulfators (a search on the net will yield a few commercial products and a
            bunch of other stuff that all eventually seems to lead back to the same
            sites). In those cases they kick a lead-acid battery with a pretty hefty
            pulse and it "rings" at some kind of a self-resonant frequency up in the
            region of a few MHz. For this sort of thing to work here you'd need to be
            using the same coil in all cases, are you doing that?

            > > > They are to set the timing on the 555, which is very fast in this
            > > > circuit.
            > >
            > > I haven't done the math, what speed is it operating at?
            >
            > Theoretically, 112khz to 1.23Mhz. I was running it on the lower end
            > when I was running it with those values installed.

            Ok.

            > > Do you mean one of those "pc scopes" or something different here?
            >
            > I was using a two channel portable Fluke oscilloscope, I don't recall
            > the model. Now I have a Tektronix TDS 2002B two channel oscilloscope.
            > They can be hooked up to a computer to transfer stored data, but are
            > standalone scopes.

            Ok, gotcha.

            > > I don't think you can do that with no load, the MOSFET being basically
            > > a varying resistance I'm surprised you see any change in voltage there at
            > > all, with the "load" in this case being the input impedance of your scope
            > > or something? If you want to try something, try substituting some value
            > > of resistance in there instead of the coil...
            >
            > I wondered about that. I'll try that test again and report back.

            Yeah, you could use something vaguely close to the DC resistance of the coil,
            or as much as 10 times higher, to get an idea.

            > > > Running it on a variable power supply I got it to make sparks, but
            > > > only by limiting the current.
            > >
            > > Limiting it to what? And at what voltage?
            >
            > The power supply was unhappy about it, and I didn't have other meters
            > hooked up to it to see what it was really doing, but... it was set for
            > 12V, but turning up the current knob it was sparking at 2 or 3 A and
            > displaying something in the area of 5V. Twiddiling the voltage knob
            > didn't change it.

            I've not done much with variable power supplies nor current limited ones,
            though I suppose I could see their usefulness in some contexts. In the case
            of this circuit I'd put a fuse inline with the power lead and hook it up to
            something that'll give it some amps, and let 'er rip. :-)

            > > > On a battery, the MOSFET just gets hot.
            > >
            > > What do voltage measurements and the scope show you then?
            >
            > A small change in voltage, but no real oscillation between drain and +V.

            Eh? With good gate drive? Something's not right there. Maybe a wiring
            error, or maybe a problem with the part, or something.

            > > I'd say refer to a datasheet for the part, and though I looked I don't
            > > seem to have one for that particular MOSFET, though I have lots of
            > > others.
            >
            > I did, and that's what it says, I just wanted to make sure that didn't
            > sound unusual.

            Not offhand but then I haven't done _that_ much with MOSFETs anyhow. I have
            been filling in on my datasheet collection of those particular parts lately,
            prodded into it by a guy's experience with another datasheet site that got me
            going on that again, my "transistors-2sh-k.html" file having had almost a
            couple of thousand lines added to it in the past week or so. :-)

            > > Did you get actual pots or trimmers? No matter, overall...
            >
            > It's a pot.
            > http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=575SF103-N
            >D
            >
            > > Any change in behavior at all?
            >
            > Oscillator still oscillates, at the lower frequency, spark still doesn't
            > spark.

            Well, you should be seeing a fairly high-amplitude pulse on the gate of the
            MOSFET. If you're not find out why not. Once you are, then you should see
            a similar pulse on the drain lead, subject to the modifications added by
            those diodes and whatnot.

            > > I don't see stray capacitance as likely being an issue with this
            > > circuit, mostly I'd worry about that with RF stuff, or very high
            > > impedance stuff where leakage can become an issue, though some folks
            > > tend to have intermittent or incorrect wiring issues with those things.
            >
            > Well, with the cap on the 555 being only 390pf, I figured it couldn't hurt.

            Yeah, and it probably pays to have good solid connections anyhow.

            > Thanks for the constructive response, I'll do some more experiments
            > and get back to you.

            Sounds like you're having fun, anyhow. :-)

            --
            Member of the toughest, meanest, deadliest, most unrelenting -- and
            ablest -- form of life in this section of space,  a critter that can
            be killed but can't be tamed.  --Robert A. Heinlein, "The Puppet Masters"
            -
            Information is more dangerous than cannon to a society ruled by lies. --James
            M Dakin
          Your message has been successfully submitted and would be delivered to recipients shortly.