Controlling Large Voltage with Smaller Voltage or Using Transistor/MOSFET as Switch

Discussion in 'General Electronics Chat' started by elec_mech, May 24, 2011.

  1. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Hello All,

    I'm working on a circuit for a Nuts & Volts reader request for a 2-digit up/down counter display with flashing capability. While I feel clever for successfully implementing a remote function, the flashing function is driving me to distraction. In essence, I need to control a large voltage (15VDC) with 3VDC (ideal) or 5VDC. I've tried a IRLD120 MOSFET and 2N3904 and 2N2222 transistors with limited luck.

    It seems whenever I apply 5VDC to the base of the transistors through a 1K resistor and 15VDC to the collector, the most voltage I see is about 4.2VDC. I'm checking this by leaving the emitter unconnected and measuring the voltage between the emitter and the ground. Please see the attachment. I notice if I increase the base resistor value, the voltage goes down. I can't recall if I've tried lowering the resistor beyond the 1K, so maybe that is my problem. Is it possible to use a small voltage such as 3V or 5V to switch 15V using a transistor (I understand the actual output will be closer to 13.5-14.3V depending on the voltage drop between collector and emitter)? I thought it was, but most of the research I've done on forums and the web regarding transistor circuits show the base and collector voltages being the same.

    On the MOSFET side of things, I've hooked it up the same way minus the resistor at the gate. I also see about 4.2V out instead of the expected 13V-plus. I understand transistors need different base currents to get achieve different outputs, but I thought MOSFETs only needed a high-enough voltage at the gate to allow full output at the source. Again, I'm using a IRLD120 which states on the datasheet it will accept 5V TTL voltages, but applying 5V at the gate only gets me 4.2V at the source. Is there a way to use a small voltage at the gate of a MOSFET to switch a larger voltage? Am I using the wrong MOSFET?

    I'd like to stick with transistors for this circuit to keep parts simple (most of us have these in our shops), but I thought using a MOSFET would solve my transistor woes since I wasn't having luck. Apparently, I know as much about transistors as MOSFETs which seems to be very little at the moment. Thanks to everyone in advance for your help and input. I'll gladly post the completed circuit once finished if anyone is interested.
     
  2. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,789
    945
    Okay. connect the transistor emitter to ground. The leg with the arrow that you have the meter on in the drawing.

    Put a small value resistor between the 15 volts and the collector around 50-100 ohms.

    Put your meter across the resistor.

    Now when you apply a voltage to the base, the voltage across the resistor will be close to 15 volts. When you remove the base signal the voltage across the resistor should go to zero. That resistor represents your LOAD.
     
  3. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    I will give that a shot tonight, thank you Kermit2. To get the 15V to where I need it (to either a PNP transistor or MOSFET controlling the common cathode connection of the LEDs from the display), do I then leave the 50-100 ohm resistor connected to ground and wire the emitter (before the resistor) to the base/gate of my PNP?

    If so, and please forgive the newbie question, do I need to worry about the resistor shorting out or becoming hot over time if the base is in the ON position for a long time, so to speak? I don't plan to draw much current, but I'm curious if that is okay too for future applications.

    Lastly, because I like beating a dead horse, I assume I'm not getting the 15V because I'm not applying a load? I tried connecting the emitter output directly to the PNP and the PNP did not fully activate because the voltage was too low. I'm sure there is little to no current draw from the PNP. Does a transistor need a reasonable load in order to pull current which then allows for the full voltage to come through?

    Thanks again!
     
  4. Jaguarjoe

    Active Member

    Apr 7, 2010
    770
    90
    The original circuit you posted could almost be an "emitter follower". Its output voltage mirrors the input voltage minus the Vbe of the transistor (0.7v). Without a resistor between the emitter and ground, your circuit was simply acting like a diode between input and output. An emitter follower has current gain but no voltage gain.

    I don't follow what you're saying in your last post. Can you provide a sketch or drawing?
     
  5. hobbyist

    Distinguished Member

    Aug 10, 2008
    764
    56
    I'll let kermit2 answer your questions,

    Just a side note:
    The transistor circuit in the schematic will give you ambiguous readings, because for a bjt transistor to conduct, it needs to be forward biased from emitter to base, and bjt. transistors only develope voltage across resistances, your 15v. will not be felt across the collector emitter junction, because a bjt. transistor is not a voltage device, but a current device.

    A base input voltage needs to draw (electron) current from emitter to base for the collector emitter junction to conduct.

    Again the collector emitter junction conducts current, so any voltage being felt at the emitter can only be by using a emitter resistance to allow a voltage drop to occur.

    If you were to use a resistor of appropriate value in the emitter leg to ground, and then apply the base input voltage, if the transistor saturates, youll get close to 15v. around 14.7 or so, (due to the voltage drop across the collector emitter junction) across this emitter resitor, because now the base has a forward bias due to the emitter resistor.
     
    Last edited: May 24, 2011
  6. magnet18

    Senior Member

    Dec 22, 2010
    1,232
    124
    Just to throw it out there, make sure they have a common ground...
    I forgot that once...
    Took me forever to figure it out.
     
  7. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,789
    945
    Try it this way. When you apply 3.3 volts to the base, the transistor will turn on. The voltage of the 12 volt battery will then make current 'flow' in the resistor named LOAD. By putting your meter across the resistor(or light or motor) as shown you will see the voltage going to the LOAD. When the 3.3 volts is present at the base of the transistor, the 'switch' is ON, and you will read 12 volts on the meter. When you remove the 3.3 volt control signal, the switch will be off, and the meter will show Zero volts.


    [​IMG]
     
  8. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Everyone,

    Thank you for your replies and advice. I went home and tried several things, ended up burning out a wallwart and singeing a finger or two when trying to locate the burning resistor.

    First off, yes the 15V and 5V grounds are shared. I burned out the unregulated 13.5V, 1.3A wallwart, it supplied 15V plus and was attached to a 7805 which provided the 5V. This was also hooked up to my master circuit which is a bit of a rat's nest, so I don't know if the burnout is a result of the transistor testing or knocking a wire somewhere.

    In any event, I switched to my benchtop power supply and used a single breadboard connected to nothing but the transistors. I carefully watched the 15V and 5V from here on out. Again, I made sure both grounds where tied together.

    Note, transistors used were PN2222 and 2N3904.

    Results - please see the attachment. Well, whenever I used a 100 ohm resistor, it wanted to burn up whether I put it on the collector side or the emitter side. It burned up immediately on the collector side regardless of the state of the base (0V or 5v). It only burned up on the emitter side if the base was at 5V.

    Logic lead me to try a higher resistor value, so I went with a 220 ohm on the emitter side. No burning, yay, but I only saw about 4.5V across the resistor.

    So, I also tried experimenting with a MOSFET since I had not tied the source to ground previously. Also bad results, the 15V supply went down to 10 and I only saw 2V across the 220 ohm resistor before I immediately shut off the supply.

    Am I doing something epically wrong here? To the best of my knowledge, I have all of the connections right. If the figures look right but the results look wrong to everyone, I'll take some pictures of my setup and post them to make sure I'm not doing something dumb.

    Figure 5 shows what I'm trying to accomplish at the moment. Ultimately, I think I can safely apply 15V to the rest of my circuit with the exception of the remote control circuit which is taken from a 3V toy. In that case, I'll still need to use 3V to switch 15V which is why I'm really trying to understand how to make transistors and MOSFETs work like switches.

    Thank you again.
     
  9. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,789
    945
    [​IMG]

    Keep at it. This is the 'learning' part of electronics. Puffs of smoke,Lots of funny smells and burned fingers. Good times. :)
     
    YorksireDave likes this.
  10. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Thanks Kermit2.

    So, it appears I'm doing everything correctly, at least based on the schematic then? I tried a couple of PN2222 and one 2N3904, but I have several so I'll give it the ol' college try. I think I bought most of them from those Jimpacks sold in local electronics shops (when you can find one), but who knows about the quality. It didn't occur to me that the parts themselves could have gone bad.

    Okay, so I'm not going crazy, this is good. Burning, smoke, and smells - and we wonder why the greater population stays away from electronics. :)

    Thanks again. I'll report my findings tomorrow.
     
  11. hobbyist

    Distinguished Member

    Aug 10, 2008
    764
    56
    Hi,

    Since this is not in the homework section, then I can give you circuit values and you could build it and test it to get you in a right direction.

    First using a 2n3904 transistor, you cannot get 280mA. witrhout excedding its max values.

    However I used a 2n3904 transistor to supply around 20mA. through 5 LED's in a series string, just to show you what needs to be done.

    1.) Put your 5 LED's in series from positive 15v. supply to the collector of your transistor.

    2.) Then put a resistor of any value from 100 ohms to 180 ohms, between the emitter and ground.

    3.) Put a 240 ohm resistor connected to the base terminal, and connect the other end to your 5v. squarewave oscillator.

    And see if the string of LED's lights up and blinks for you.

    If you need to drive 14 strings through one transistor, you will need to use a transistor with higher max. power ratings.
     
  12. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    I feel like I'm Wile E. Coyete, transistor concepts are the Roadrunner, and the transistor circuits I'm trying are the constantly malfunctioning ACME products.

    Okay, so I toasted, um, I mean, tested 8-9 NPN bipolar transistors in similar configurations with three different models - 2N3904, 2N4401, and 2N2222. In all but one case, the transistor got hot. In the one case where it didn't, I still couldn't see more than 4.5V across the resistor connected to emitter and ground.

    This leads me to believe one or more the following:
    1) I'm doing something terrifically wrong. I'm using a RadioShack transistor tester to determine which lead is the collector and base. Maybe it's misinforming me?
    2) All the transistors I've tested are bad - I find it unlikely, but it's not outside considering.
    3) My power supply is bad - unlikely since it seems to work fine when tested with a meter.
    4) My meter is bad - unlikely since I'm feeling and witnessing parts getting too hot when my meter tells me something is wrong.
    5) The circuit isn't correct for some reason or I'm not hooking something up properly.

    If I may beseech, impore, beg, etc., could someone attempt to repeat my tests in the attachment using a dual power supply and breadboard? This would immensely help me keep my sanity and hopefully help me hone in on the problem. I would ask you to use only a 2N2222, PN2222, 2N3904, or 2N4401.

    Any additional advice would be greatly appreciated. I simply cannot get this to work and it's driving me mad.
     
    Last edited: May 26, 2011
  13. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Hi Hobbyist,

    I just saw your reply. Thank you for the advice. I'm driving the strings of LEDs with a UDN2982 trasistor array on the anode side. Tentatively, I've been using either a IRF9540 PNP MOSFET or a 2N2907 PNP (rated about 600mA) to control the common cathode where all of the cathodes are attached. The problem is I need 15V to go to the gate/base in order to fully shut off the MOSFET/transistor whereas I only have roughly 3V coming from the 555 astable circuit. This may become a non-issue once I attempt to power everything from 15V, except for the 3V remote control. In that case, I need a way to use 3V to control 15V. I thought I could do this with a transistor, but so far no dice.

    Thank you.
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    You are using most of the transistors as emitter-followers so the emitter does the same as the base except its voltage is a diode voltage drop (0.7V) less than the base no matter of the collector voltage. The emitter will be about 4.3V max.
    The current is 4.3V/47 ohms= 9.1mA and the heating in the transistors is (15V - 4.3V) x 9.1mA= nearly 0.1W.

    You correctly connected one transistor as a common-emmiter type (emitter connected to ground) but selected a collector load resistor value much too low so the current from 15V was trying to be 15/47= 319mA and the heating of the 47 ohm resistor was trying to be 15V x 319mA= 4.8W! But the base resistor value was not low enough for the transistor to completely turn on.
     
  15. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,789
    945
    Agreed with Audioguru. Stop putting resistors in the emitter portion of your circuits.

    Go back and look at my post (#7). connect the emitter to ground. Directly to ground. No resistors in that leg of the circuit. Got it?

    Use your 15 volt value and calculate the current in a load, which you place between your supply positive and the collector of the transistor. If you have to much current for the transistor to handle increase the resistance of your load.(place a resistor in series with it)

    keep any base resistor at 1K or less. If the transistor is dropping more than .1 or .2 volts then it is not fully turned on. Lower the resistance in the base leg in this case, but not too low..
     
  16. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Alright. Resistor between emitter and ground bad, got it.

    I think I've gotten too focused on one thing, so let me revisit the original issue.

    I would like to use a transistor to flash on and off the common cathode of several LED strings. When there is no pulse, I would like the LEDs to remain lit. To this end, I was originally using a PNP MOSFET, now a PNP transistor. In either case, the LEDs never went fully off, they lighted fully then dimmed. I noticed if I put 15V to the gate/base, the LEDs went off completely. This led me to think if I could turn the 2-3V pulse from the 555 into a 15V pulse, the problem would be solved. In hindsight, that seems redundant if I can control the PNP directly with the 555 signal.

    So, if possible, I'd like to use the 2-3V pulse from the 555 to turn on and off fully a transistor as shown in the attachment. If I understand everyone's input, I can achieve this if the base resistor is low enough, but not too low, correct?

    Kermit2, when you say the transistor should not drop more than 0.1-0.2V, where are you referring? Across the base resistor? Between collector and emitter?

    I've added a 10K resistor between base and ground so the transistor is conducting and the LEDs remain lit whenever I have the 555 turned off. Is 10K okay, should I use a different value, or will I be inviting disaster again?

    If the LEDs do not turn fully off, then I should lower my base resistor, correct? Is there a minimum value or calculation I should be taking into account?

    Thank you everyone again for your patience and assistance.
     
  17. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Again, you have an emitter-follower transistor except this time it is a PNP type.
    It should be an NPN type with its emitter connected to ground and the LEDs connected to its collector. It needs a 1k series base resistor.
    When the output of the 555 goes high then the transistor lights the LEDs and when the output of the 555 goes low the LEDs turn off.
     
  18. elec_mech

    Thread Starter Senior Member

    Nov 12, 2008
    1,513
    193
    Audioguru,

    Thank you for the input. I understand I'm not using the transistor properly, I don't know why it makes a difference what side the load is on, collector vs emitter, or what an emitter-follower is (a quick search reveals I need a LOT more time to study these).

    Attached is my understanding of what you have suggested plus a pull-up resistor to keep the LEDs lit whenever the 555 is off. Is this correct? Will the pull-up prevent the LEDs from going fully off? If yes, what might I do to keep the LEDs on when the 555 is off, but not interfere with the switching operation?

    Thank you again.
     
  19. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    A normal 555 (not a low power Cmos one) has an output that powerfully goes high and low, it never goes OFF. Your 10k resistor will do absolutely nothing except waste a small amount of power.

    Why do you want the LEDs to light when the output of the 555 is low? They already light when the output of the 555 is high so if they also light when the output of the 555 is low then the 555 is not needed.

    Get rid of the 555 and connect the LEDs and their current-limiting resistor directly to the power supply and they will light continuously.
     
  20. hobbyist

    Distinguished Member

    Aug 10, 2008
    764
    56
    I think he wants the LED's to be continually on, when the 555 is not operating, but when the 555 is turned on, then the toggle output from the 555 will connect signal to a transistor to blink the LED's.

    If this be the case, then you may need to make a seperate transitor circuit that will keep the LED's lit during no pulse, and another transistor that will couple to the LED driver transistor, that will take the 555 output signal and cause the driver transistor to blink in unison.

    Maybe, something like this.?
    pulse led.jpg

    The one LED is just for proof of concept.
     
    Last edited: May 26, 2011
Loading...