Help-Need advice on using zener to drive mosfet

Discussion in 'The Projects Forum' started by InventPeace, Jul 3, 2011.

  1. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Hello,
    I trying to build a battery overvoltage "battery charger" cutoff circuit. I must be in error because what I keep trying isn't working. I am trying to build a power cutoff when a battery reaches a predetermined higher voltage. I wish to use a zener diode of perhaps 16 volts to provide a trigger to the mosfet, which in turn drives a relay to cutoff the 110v to the charger.
    Jeff Zener Trial  Untitled.jpg
    I got the mosfet to turn on with low voltage at the gate and off with high voltage at the gate, so I think that part is ok. (hope you can see the attachment diagram) ; but when/if the zener breaks down it doesn't give a positive enough signal to the gate to of the mosfet to turn it off?
    I've tried all different voltage at supply from 12 - 15.6v , and many different resistors for R1 (100, 200, 15k, 47k , etc ), also tried zeners rated 10, 12 and 15 volts. It may be that the Vgate has to be higher than the Vdrain but I'm unclear on howto do it?
    Any help would be appreciated. I guess I'm really rusty anymore (getting old), but rather than waste another day trying myriad of different setups, i thought I'd just ask the experts.
    I got another zener setup to work fine using an LED that tells me when batteries are low voltage (LED when voltage drops to 11 volts on 12v batteries). But this is a little beyond that.


    **To limit a 14.4 volt NIMH battery from being overcharged, unfortunately normal battery chargers when they "seem" to be at 0 amps are really still feeding about "up to" 1 amp (shumacher etc all have same problem), this can in the case of lead acid 12v batteries boil away the water ruining the battery over a week or two (so don't ever leave the charger on for more than it takes to charge the battey fully or about 6-10 hours using 15 amp charger). In the case of NIMH batteries its even worse, the nimh actually can overheat and catch fire (nimh voltage actually goes down just very slightly when fully charged which is very hard to detect with); this slight v drop can result in metling as NIMH batteries can tend to hog the juice once fully charged and most chargers aren't smart enough to know its already fully charged.
     
    Last edited: Jul 3, 2011
  2. #12

    Expert

    Nov 30, 2010
    16,283
    6,794
    Ignoring the fact that many labels are missing, the problem is that mosfets require significant voltage change really to turn on. This would work much better with a bipolar transistor because they switch from nearly nothing to full on in about a tenth of a volt.

    Is that enough to get you going?
    Reply if necessary.
     
    InventPeace likes this.
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    You're trying to detect when NiMH batteries are charged using a crude voltage cutoff type of thing. It won't work.

    NiMH batteries can't be charged the same way you'd charge a lead-acid battery. You need to look for that "bump" you mentioned, AND monitor the battery temperature. Otherwise, you can wind up having a fire.
     
  4. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Please read carefully orginal post:

    Actually, we thought of using a logic chip and writing the assy language for it (already mostly written) but really , with all the surges we get around here , not to mention other surges from charging system and batteries as well, we'd prefer something more basic at alot less hassle and cost, thus: the system similar to the one described or a better one if you know of one. SINCE we don't plan to go to full voltage (or full charge) we can just pick a voltage at about 80 or 90% of full charge and cut if off there. (TOYOTA says that if you keep NIMH batteries between 40% and 80% there is no degradation (thats why toyota RAV-4 EV batteries last so long; 1998 models still running on original ev1 nimh batteries and still able to take charge in only 15 minutes WOW !! Now you know why Texaco dealed with GM to undermine ; buy and suppress the full large format NIMH such as was in EV1; someone appears to doing similar undermining to the NIMH battery inventors later invention, the solar shingle ECD UNI-Solar, which is now in bankruptcy?? at a time when we desperately need alternative energies to flourish? Ovshinski's solar shingle can even run in the rain still giving power, where si based solar cells quit unless in nearly direct or direct sunlight).
    Your statement about charging is similar to what we had in original post? (nimh charging needs are more finicky than lead acid and could melt down if overcharged or overdischarged or overheated.; "....the nimh actually can overheat and catch fire ...."
    1. I think a zener can be used as a clamp to limit voltage to a set level as described below the max. Just looking for cheap simple effective way to do it.
    2a. We also need a method to cut off charging at least to individual battery or charger system in the case of about 120- 130 Deg F (approx). The thermistors we have now are way off , more like 180 deg F and higher before they go to zero resistance, which is way to high for our needs. (*better suited to logic chip opreation probably).
    2b. SO MAYBE someone of you knows of a good way to cut off power if thermal overheat occurs at around 120-130 Deg F?
    We need both voltage and thermal cutoff independent of each other for more insurance and safety.
    Of course you Don't have to charge at high amps but its nice to know it could be done safely. (Thinking that high amps leads to the meltdowns much more often than low amps , just seems sensible).

    The 2nd poster had a good reply, yes bipolar may very well be the solution, we'll be back on it in a couple days and let you know. THANKS !!
     
  5. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    Ok, keep your hair on, but according to your diagram you managed to get the FET connected in common source, which pretty well knackered any chance of the thing working, irrespective of the exact device characteristics. The output from a common-source is less than its input, and with the circuit you were using the relay coil could never have seen more than a few volts, maybe nothing at all if the FET required a largish voltage to turn on.

    Will you make a similar error and put the BJT in common collector? If you get it in common-emitter, do you know that something has to be added to keep the circuit safe if the charger cuts off suddenly?

    In short, don't despise expertise: you came here asking advice, so do not be surprised if it is offered.
     
    InventPeace likes this.
  6. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Oh, ok.. I won't use bold or caps or underline.. guess people think we are yelling/mad,, no, just trying to make specific points stand out.
    "Just looking for cheap simple effective way to do it". Humble is our name or we would not ask, we don't despise expertise , but in fact admire it.

    IF you know of a way using a zener(s) please put the circuit up for us. thanks. If you know of any cheap simple way , please put that up as well.
    Also the thermal issue would be nice as well to have a cheap simple effective way to prevent thermal runaway. We are rusty with our electronic knowledge and really wondering if it can be done that easily, THANKS !
     
  7. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    A diode with cathode to positive side of relay coil would protect the circuit from back emf, if thats what you are asking.
    Maybe someone has access to SPICE program and can quickly come up with a solution, the transistor should act as a simple on/off switch triggered by increase/decrease from zener reaction to voltage changes. Need power on charger until approx 15.6 volt or higher is reached on batteries, then transistor cuts power to relay and in turn cuts off power to charger. The battery voltage should hold the transistor off for awhile until battery v falls enough to cause it to come back on.
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Your idea to energize a relay to cut the mains power when the battery is charged is somewhat self-defeating.

    Relays are notoriously power-hungry, and will discharge your batteries in fairly short order. You'll wind up with the charger on initially, but when the batteries become charged enough to energize the relay and turn the mains power off, you'll wind up discharging the batteries back down again until the battery voltage falls to the relay dropout voltage - then charging starts again in a seemingly endless cycle - until your batteries hit their charge/discharge cycle limit.

    What you need is something more like a circuit that is latched by the charger output; when the batteries hit the terminal voltage, then the latch is broken until it is manually reset.
     
  9. wayneh

    Expert

    Sep 9, 2010
    12,114
    3,039
    A thermostat based on an
    LM35 thermometer and a comparator would do the job. Having the "spare" comparators left over - if you got a quad such as the LM339 - would allow using it for more precise battery tending strategies.
     
    InventPeace likes this.
  10. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    I guess all the kings men and all the kings horses can't figure a way to use a zener in a simple circuit to trigger a mosfet or bipolar or any other transistor which in turn could turn on (of off ) a relay? The challenge has been issued!
     
  11. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Hey "wookie" did you write this back in 07 ? If you look at the circuit (below in your text) it looks like its a zener circuit that you say triggers a transistor? Yet earlier you said it wouldn't work ( in this thread), can you explain? What we need is a circuit diagram that would tell us exactly what values for what parts (we are too rusty to try to do it all on just suggestions and have no access to SPICE programs at this time either also slow internet and 40zillion other problems to work out like installing batteries in boxes not big enough, and also trying to make a living). Please see next post.
    Also on the "self defeating" aspect of not having a latching disconnect, we thought that we could add a latch somehow after we got the overvoltage control inplace we could add a latch; this could be done perhaps with just an additional relay perhaps that could be set (on) with a manual help (push button), to allow current to flow, then when transistor turns off, the primary relay cuts power to the charger and the extra relay which prevents power from battery to circuit, thus preventing current drain.
    [​IMG]SgtWookie [​IMG]
    Senior Member
    Join Date: Jul 2007
    Location: Orlando, FL USA
    Posts: 17,319


    [​IMG]
    See the attached.

    On the left is just a simple model of a 60mA current supply whos' voltage output ramps up as the capacitor charges across the resistor; you can ignore it.

    On the right is a voltage clamp. The three Zener diodes are arranged in series so that basically, the sum of their voltages is equal to the voltage at which you wish to start clamping.

    If there is no current flowing through the Zener diodes, R1 keeps the base of Q1 low, so that Q1 does not conduct.

    When the voltage from the current supply increases the voltage beyond the sum of the Zener diodes, Q1 turns on and begins conducting current. This transition is rather sudden.

    Q1 will dissipate around 3.7 Watts, so it must be heatsinked.
    D1-D3 will dissipate around 580mW apiece, which is why there are three in series and not merely a single Zener.

    Instead of a 2N3055, use a 2N3902. I didn't have the 2N3902 in my PSPICE library, which is why the 2N3055 is shown instead. The 2N3902 has a far higher Vce rating than the 2N3055 does. As shown, the 2N3055 is being operated right at it's absolute maximum ratings, and would not last long thus being unreliable.
    Attached Thumbnails[​IMG]

    __________________
     
  12. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    TO WayneH: THanks for advice; can you give us a complete circuit perhaps with adjustment for temp between 100deg F and 120deg F approximately ? thanks if you can, we are so strapped for time between family, work, installing batteries and other rewiring needs, that it would just save us alot of time.
     
  13. dollat7852

    New Member

    Sep 22, 2010
    5
    0
    Try using a bjt with appreaciable gain n dat may be all
     
  14. wayneh

    Expert

    Sep 9, 2010
    12,114
    3,039
    Take a look at this project for some ideas.

    Here's an example of using the LM35 and a comparator as a thermostat to control a MOSFET and high current devices.
    Picture 2.png
     
    Last edited: Jul 8, 2011
  15. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    SgtWookie,
    Did you write the below in 07? It looks like you are triggering a transistor with zeners? Does this work? Didn't you say earlier in this thread that "...it won't work"; this is very similar to what I posted originally. I still think there must be a way to use a simple zener to trigger a mosfet or other maybe bipolar transistor on or off ; to in turn, turn a relay on/off.

    [​IMG]SgtWookie [​IMG]
    Senior Member
    Join Date: Jul 2007
    Location: Orlando, FL USA
    Posts: 17,330


    [​IMG]
    See the attached.

    On the left is just a simple model of a 60mA current supply whos' voltage output ramps up as the capacitor charges across the resistor; you can ignore it.

    On the right is a voltage clamp. The three Zener diodes are arranged in series so that basically, the sum of their voltages is equal to the voltage at which you wish to start clamping.

    If there is no current flowing through the Zener diodes, R1 keeps the base of Q1 low, so that Q1 does not conduct.

    When the voltage from the current supply increases the voltage beyond the sum of the Zener diodes, Q1 turns on and begins conducting current. This transition is rather sudden.

    Q1 will dissipate around 3.7 Watts, so it must be heatsinked.
    D1-D3 will dissipate around 580mW apiece, which is why there are three in series and not merely a single Zener.

    Instead of a 2N3055, use a 2N3902. I didn't have the 2N3902 in my PSPICE library, which is why the 2N3055 is shown instead. The 2N3902 has a far higher Vce rating than the 2N3055 does. As shown, the 2N3055 is being operated right at it's absolute maximum ratings, and would not last long thus being unreliable.
    Attached Thumbnails[​IMG]

    __________________
     
  16. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Still waiting for "anyone" to post a very simple zener circuit to trigger a transistor. Since this is so simple we'd expect the submitter to actually test it ahead of time. Again battery will be connected to charger that need to be powered off when 15.6 volts is reached. We feel that we should be able to think of it ourselves but we are so rusty, and we just don't have the time to put on it, hoping some of you experts would be able to post a good circuit for this type operation. THANKS !
     
  17. InventPeace

    Thread Starter New Member

    Jun 20, 2011
    13
    0
    Thanks WAYNEH, it looks great but its really too involved for what we want to do, thanks anyway. We are looking for just a few parts involving a zener dioded to trigger a transistor, that in turn would turn a relay on/off (which controls 110v ac to batttery charger, later we can add a latch mechanism to prevent battery running down after charger is turned off, repeating, cycling).
    Advice to all: don't leave chargers plugged in for days or weeks, with some batteries it can boil away the water ruining the battery. (ie PbA flooded types). We are actually working with nimh at this time but I think it would be good to have the charger off as much as possible just for additional safety.
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Yes, I did.

    The person in question wanted a "voltage clamp" for a current limited source. From memory, their current source had the capability of outputting a pretty high voltage; perhaps 400v or so, but their variable load would be damaged if the voltage across it exceeded approximately 60v. That wouldn't be exact, but it would be within a few volts.

    It worked for that circuit.
    Nope, sorry - it's very different. The transistor is in cutoff (not conducting via the collector) when Vbe = 500mV, but is in conduction by the time Vbe reaches around 620mV. That's a pretty small difference; 120mV.

    You're trying to switch a MOSFET using a Zener. a MOSFET gate needs to be below the Vgs(threshold) to be considered turned off, and at one of the specified Vgs's for the Rds(on) specification in order to be turned ON. That can be several volts' difference. The threshold voltage alone can generally vary by a couple of volts. That makes it problematic to give someone an exact voltage when a MOSFET will start to conduct.


    There might be. I don't have a good solution for you.
     
  19. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Just FYI, all members who help on here are purely volunteers who may have some slices of spare time on their hands.

    My time in particular has become very limited, and there are LOTS of people who want help.

    Now, expecting people to actually design a circuit for you, and then buy parts to test that circuit, etc; - well some members do that, but they might have a lot of extra time on their hands.

    Zener diodes have a bit too much tolerance to expect a precise voltage out of them. The lack of adjustment can be a problem; once you build something, it's hard to band-aid it later.
     
  20. wayneh

    Expert

    Sep 9, 2010
    12,114
    3,039
    I hope you realize that a lot of the parts in the schematic I posted for you might be unnecessary for your usage. For instance you may not need the 5V regulator, any of the LEDs with associated resistors, and obviously you are not switching a TEC. That leaves a sensor, a comparator, and a MOSFET at the core of the circuit, plus a few resistors. I don't think you'll get much simpler.
     
Loading...