Zener + Polyfuse circuit protection

Discussion in 'Electronics & Electrics' started by imgod22222, Jun 18, 2013.

  1. imgod22222

    imgod22222 Member

    Joined:
    Jan 4, 2006
    Messages:
    729
    Location:
    Florida, America
    I wanted to check with y'all that this is a sane voltage implementation.
    [Yes, I know that circuit protection is a subject that has been beaten to death by the internet, but please bear with me.]
    Where the zener is rated for the max voltage the load can take
    And the resettable polyfuse is rated for min(Load_max_current,zener_max_current)

    [Ramblings below. Above is my real question.]
    And ideally, the zener is rated for a current higher than that of the load. Maybe a darlington pair with a zener connected (anode to base, cathode to collector). In that case, the zener would be rated for (load voltage + turn-on voltage of darlington pair), correct?
     
    Last edited: Jun 18, 2013
  2. callan

    callan Member

    Joined:
    Aug 16, 2001
    Messages:
    4,955
    Location:
    melbourne
    The Zener only has to be rated to carry a current significantly higher than the fuse (which by implication is higher than the load current of the device being protected.
    I suppose theoretically you could rely on the load helping to pop the fuse, but I wouldn't. Generically, the load could vary, and could be a bad mistake (the load could be in standby mode, for example)

    Callan
     
  3. Blue_Yoda

    Blue_Yoda Member

    Joined:
    Oct 19, 2006
    Messages:
    18
    So as I understand it, what you are trying to do is supply a regulated voltage and have over current protection for the load. This is what you are proposing:
    Circuit 1: http://imgur.com/8hwb44l

    Is it sane? Not really. For starters you're missing a series resistor:
    Circuit 2: http://imgur.com/sSOCvcL

    But Circuit 2 is still not sane. Because your fuse now needs to blow at Izener(max) + Iload(max). But Izener = (Vin - Vout) / R so it will vary depending on what Vin is.

    Better way to do things is shift the location of the fuse:
    Circuit 3: http://imgur.com/Cu8im8x

    So now the polyfuse's trip current is simply your Iload(max).

    Hope that helps.
     
  4. aXis

    aXis Member

    Joined:
    Jun 27, 2001
    Messages:
    5,548
    Location:
    Kalgoorlie, WA
    It might have some use in specific situations but it's a long bow to draw to call it "Better". Personally I think it's terrible - it's adding series resistance to the circuit which just wastes power all the time and impacts on Vout, and in the event of overvoltage allows the circuit to continue running whilst dissipating copious amounts of heat though the series resistance and the zener.

    The OP's origonal circuit looks fine. Rate the polyfuse appropriately for the load current, and rate the zener something like 200% or 300% of the polyfuse current so that it can trip quickly.
     
  5. Blue_Yoda

    Blue_Yoda Member

    Joined:
    Oct 19, 2006
    Messages:
    18
    It will not impact Vout at all. Vout is set by the Zener. The resistor is needed to balance the circuit. If Vin is higher than Vzener then the voltage drop (Vin - Vzener) must occur somewhere (if not at R then it will be up stream somewhere probably inside the power supply or you will just blow the zener if Vin is high enough).

    By placing a resistor you choose where that happens, how much power can be dissipated safely and prevent the Zener from suffering an overcurrent condition.

    Some power is wasted but that's the trade off you get for this very simple regulator.

    You will not get an overvoltage condition because the Zener sets Vout. As Vin rises the current through the Zener will increase to maintain.

    Some sources:
    http://en.wikipedia.org/wiki/Linear_regulator#Simple_Zener_regulator
    http://www.electronics-tutorials.ws/diode/diode_7.html


    OP:
    To help you in picking components, this calculator can help you out:
    http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/zenereg.html

    Also note, this kind of voltage regulation is generally for lower currents as (like aXis pointed out) you will be dissipating power across R and your Zener's max current rating comes into play as well.
     
  6. gdjacobs

    gdjacobs Member

    Joined:
    Apr 3, 2007
    Messages:
    2,567
    Location:
    MB, Canada
    Linear voltage regulators all waste energy. It's unavoidable. Having a series resistor will give you clean, regulated output voltage and protect downstream components. Having no series resistors may result in nuisance tripping with low voltage or short duration transients. Current would be limited by wire, fuse, and supply impedance, so it might be quite high.

    Fuses don't limit current. Impedance limits current.

    If you want to implement overvoltage protection without a resistor, use a MOV. However, MOVs are not meant to be loaded continuously, so don't use them for regulation.
     
  7. aXis

    aXis Member

    Joined:
    Jun 27, 2001
    Messages:
    5,548
    Location:
    Kalgoorlie, WA
    Blue Yoda & gdjacobs - Why are you talking about linear regulators? The OP clearly asked for a "protection circuit" - i.e. his circuit demonstrates overvoltage and overcurrent protection.
     
  8. gdjacobs

    gdjacobs Member

    Joined:
    Apr 3, 2007
    Messages:
    2,567
    Location:
    MB, Canada
    Then use a MOV and size the fuse for your conductors and load.
     
  9. Blue_Yoda

    Blue_Yoda Member

    Joined:
    Oct 19, 2006
    Messages:
    18
    Apologies. My mistake. I interpreted from OPs body text that he might be wanting a zener regulator when title is clearly "Zener + Polyfuse circuit protection". :/

    OP: What are you protecting against? And what are you protecting?
    That will allow us to help you decide what's best to be used. Your current solution with a Zener is valid but will not cover you for all situations. There are other devices to consider such as TVS diodes, and as imgod22222 suggested, MOVs.
     
  10. Symon

    Symon Castigat ridendo mores

    Joined:
    Apr 17, 2002
    Messages:
    4,820
    Location:
    Brisbane QLD
  11. OP
    OP
    imgod22222

    imgod22222 Member

    Joined:
    Jan 4, 2006
    Messages:
    729
    Location:
    Florida, America
    I'm protecting an electrically-tunable lense (it's current-controlled, with a max continuous current of 300mA (400mA peak) and a max voltage drop of 5V)

    I've made a circuit that should let me adjust a linear voltage regulator (configured to act as a adjustable current source via a digital POT), and has a high-side current meter that outputs to a 16-bit ADC, so I can monitor and adjust the current through the lens.

    But I need to ensure that in the event of a fault-condition the lens is okay because the lens passes a laser which, even diffracted, could potentially cause blindness to an operator in the same room, or potentially cause burns. (Yes. A super powerful laser. All for science.) Given a fault-condition, an electrically-disconnected lens actually allows the lens to sag (due to gravity) and "bend the beam down" where it is then absorbed by a surface with great thermal dissipation. And as such, being able to fuse it/volt-protect it is important to me.
    Power consumption? Not so much. Noise? Totally.
    1. I don't think I need a TVS as the load (lense) is sitting downstream from a linear voltage reg.
    2. I have no experience with MOVs nor IS barriers, so I'll start doing some reading.
    3. I don't want to use a resistor in the lens supply path--my current source has a max of 500mA. I need that headroom to encourage faster switching times in the lens.

    But please! If you have something else to suggest that may perform better, be my guest and tell me. I'm adequately inexperienced compared to some of you, undoubtedly. Based off what I think I know polyfuse+zener seems like a fine idea. But in the mean time, I'll start reading to further educate myself as per your suggestions thus far.

    ::Also worth noting:: Since it is now obvious that I am not using a fully analog solution, there exists code where in the event a failure is detected (loss of connection from the ADC gathering current data/current sensed exceeds a predefined threshold) to disable the current source/send the actuating/controlling hardware into a safe OFF state.
     
    Last edited: Jul 11, 2013
  12. Blue_Yoda

    Blue_Yoda Member

    Joined:
    Oct 19, 2006
    Messages:
    18
    Cool :)

    Sounds good.

    So you are ultimately trying to prevent human injury. This is when part of me says "Tell this guy to hire a professional". But I assume that you don't have the budget for something like that?

    So, as I understand it, you actually have three problems to solve:
    (1) Protection of the lens from a supply fault (over voltage you've mentioned is there also an under voltage condition though?)
    (2) Cutting power to the lens if it fails (worst case is a short circuit it seems)
    (3) Immediately putting the whole system into a safe state (shutting down the laser?) if the lens loses power and/or does not behave as you intend it.

    I agree. A TVS is for suppressing transients normally from an ESD bolt delivered by a human touching stuff. I assume no humans are allowed to touch this when it is in operation. Also these bolts have very little energy and I think your lens would just gobble it up no problems. Finally, like you said, it is down stream of a linear regulator.

    I'll have a dig around for some good quality reputable articles for you as well. They're buried somewhere on my work drive.

    I agree. I only suggested resistor because I thought you might be regulating a voltage down no protecting from over voltage (Sorry about that. It was late at night). If protecting from overvoltage for a high current device like this, using that series resistor gives you nothing and hampers performance.

    Zener:
    1) How much of an overvolt at a known time can this lens take? It sounds like exotic kit. A very very narrow gotcha that comes to mind is ensuring that the Zener turns on in time (generally nano to micro seconds). Most probably not an issue as it sounds like a higher energy device, but just pointing it out, just in case. Let us know if you are not sure, there are ways to work around this.

    2) Make sure you take the tolerances into account, especially the clamping voltage (so many people forget this).

    Polyfuse:
    I would recommend no. Polyfuses are good for consumer products (or generally "don't care situations") where if a fault occurs, you just want to protect the device, and then automatically clear the fault condition, thus allowing the device to recover on its own.

    But this sounds like a specialist setup. I would want something that blows irreversably (like fast blow glass fuses). This is so that when a fault occurs, an operator is forced to diagnose it and the approximate area of the fault is more apparent.

    In your setup, this is key to keeping people safe IMO. Is this a custom setup?

    My understanding is that: Power loss to the lens is bad because gravity deforms it and you are not longer controlling the direction. And any exposure to a human results in harm very quickly. You should have a very fast and very good watch dog circuit poised to cut power to the laser unless your data sanity checking kicks it to keep power going (and a windowed watch dog does not count as a good watch dog). There is one particular article that tackles this very well. I'll try find that for you too.

    EDIT1: Here it is: http://www.ganssle.com/watchdogs.htm

    Also, since you are getting help from the Internet, you should triple check everything :)
     
    Last edited: Jul 11, 2013
  13. gdjacobs

    gdjacobs Member

    Joined:
    Apr 3, 2007
    Messages:
    2,567
    Location:
    MB, Canada
    This isn't clear. If current is shut off to the lense, will the lense (by design) focus the laser onto a beam dump, giving you a safe condition?
     
  14. OP
    OP
    imgod22222

    imgod22222 Member

    Joined:
    Jan 4, 2006
    Messages:
    729
    Location:
    Florida, America
    *cough* I am the professional. Scary, isn't it?

    Exactly true. There's no way to automate shutting the laser off without possibly damaging the >$50k USD laser.

    Implemented on the microcontroller.
    EDIT: HOLY COW. I am reading the article now, and am in love with the wealth of information. Thank you very much for this! +1 internets to you, good sir.
    ===
    Fast-blow glass fuses are an option. I was justifying the polyfuse because I know that there's nothing that needs to be replaced [and there will never be a difficulty with sourcing the part that needs to be replaced]. I suppose fast-blow glass fuses are common enough--finding a fuse specific to my application that isn't only provided by a single supplier in a cave in the Himalayas is why I immediately gave up on trying to do that.
    ===
    And yes, it is a custom set-up. Always fun.

    With any luck you guys/gals can push me toward the right direction. And of course, I won't take any advice at face-value. I'll seek to confirm I'm not getting bogus advice.
    It is the internet after all :p
     
    Last edited: Jul 13, 2013

Share This Page

Advertisement: