Opto Isolator Problems

Discussion in 'The Projects Forum' started by Mechatronic, Oct 17, 2009.

  1. Mechatronic

    Thread Starter New Member

    Oct 15, 2009
    9
    0
    I am having troubles with my Opto in my circuit. On the the output side i am getting a very large voltage drop across the collector to emiter. I have a source of 6 V wich goes through a 100ohm resistor to pull down the current. any suggestions.
     
  2. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    You have a few things wrong here. I recognize your spice as similar to Tina so R4 is shorted, as I can visibly see the wire running through it. I'm surprised your spice didn't throw a net list error. Even if it wasn't shorted, 35 Ohms is far to low for a base bias resistor. It should be closer to 1K at your supply voltage. I would also change R3 to 1K. Lastly, the neg end of your 6V Vcc is not connected. Still no net list error?? Is this the free TI version of Tina?
     
    Last edited: Oct 17, 2009
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Your R4 has a short across it.

    Try it more like this:

    [​IMG]

    In this version, R2 replaces your R4. You're using the optoisolator as a voltage follower anyway, might as well limit current via the collector.

    The R4 in this version ensures that Q1 still turns off quickly, but doesn't load the optoisolator output as much.
     
  4. Mechatronic

    Thread Starter New Member

    Oct 15, 2009
    9
    0
    just drew it in tina i didnt run it at all.
     
  5. Mechatronic

    Thread Starter New Member

    Oct 15, 2009
    9
    0
    will it work if r3 is 7.2 ohms?
     
  6. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    I thought I was crazy answering posts at 0100hrs! What are you guys doing here? Hey, I'm hitting the sack....
     
  7. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    That value R would require a hefty power transistor. Hey, more in the morning!;)
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Oops Cdrive - I cross-posted ;)

    I chose 200 Ohms for my R2, as that allows around 24mA base current for Q1, allowing Q1 to sink up to around 240mA (which includes 84mA via R3) and still be well saturated.

    If our OP really only needs to sink 84mA via R3, they could increase my R2 to 560 Ohms without changing R4.

    As far as our OP's 6v supply not being grounded; I think that got cropped off of the right side.
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Oops, I didn't see the decimal point. :eek:

    You'll need a more capable transistor than a 2N2222.

    6v across a 7.2 Ohm resistor will have 834mA current.
    It will require a base current of 83.4mA to saturate the transistor.

    You'd really be pushing beyond the limits of an optoisolator to put that much current through it. I was pushing it already with 25mA current.

    You'll need to either go to a Darlington transistor (and have a Vce of around 1v) or go to a power MOSFET.
    [eta]
    There's another option; see the schematic on page 2.
     
    Last edited: Oct 17, 2009
  10. Mechatronic

    Thread Starter New Member

    Oct 15, 2009
    9
    0
    Sorry Guys im in australia didnt think of time difference. That fixes my opto problem but now my transitor isnt saturating completly. It has a beta of 40. im loosing half the voltage across the colector to emmiter.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Had to go offline for a bit to fix a problem with the computer.

    That's basically what I was talking about earlier; your optoisolator would be too hard-pressed to supply enough base current.

    But here's a way to use a PNP driver to supply the base current without having to resort to a Darlington:

    [​IMG]

    R2 limits Q2's base current. [eta] R2 should be increased to 560 Ohms.
    R4 keeps Q2 turned off when the optoisolator is turned off. [eta] R4 should be increased to 2.2k Ohms.
    R5 limits Q1's base current to around 90mA.
    R6 turns off Q1 when Q2 is off.

    [eta]
    Since you're in Australia, you should put that in your profile. It will help when people are trying to recommend parts or suppliers. I think you're about 15 hours ahead of me.
     
    Last edited: Oct 17, 2009
  12. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Well, our OP would either need to use a logic-level MOSFET (which Radio Shack doesn't carry) or have a higher voltage available to drive the gate. Depending on their desired switching rate, a gate driver circuit may also be required.

    Using a Darlington would result in a lot of power dissipation in the transistor, as Vce(sat) would be somewhere around 1v. A Darlington would be an "easy way out", but at the cost of expending 1/6 the output power in the transistor rather than the load.

    Whether the power transistor is PNP or NPN, it will still need at least 80mA base current to saturate, and many common optoisolators can't handle that much of a load without a very high Vce; a problem that our OP was initially running into. Limiting the current through the optoisolator also reduced their power transistor's base current, and it came out of saturation.

    I added a PNP driver (2N2907) to handle the power transistor's base current requirement. This cut down the optoisolator's current requirement to roughly 8mA, the PNP sources roughly 90mA, and the NPN power transistor sinks 800mA+. Everything's within specification limitations, and Vce(sat) of the power transistor will be far lower than that of a Darlington.
     
  14. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Touché on all counts Sgt. I was to lazy to run the numbers at that hour of the morning. Come to think of it, I don't think the Nyquil helped either. :)
     
  15. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    After looking at it again, I realized that I posted a schematic that had old values for R2 and R4. Those values result in about 25mA current in the optoisolator's output when it's turned on.

    Increasing R2 to 560 Ohms and R4 to 2.2k Ohms will have the optoisolator sinking roughly 9mA current, and roughly 8.7mA current through Q2's base.
     
  16. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Just as a side note here: Mechatronic's optoisolator is not any known opto model. It's a discrete LED and two terminal photo-transistor that he drew a rectangle around. Tina will let him use the LED as the control element for the photo-transistor and defaults to a coupling factor of 1 and Beta of 200.
    Rc=1
    Re=1

    EDIT: Second look tells me that the photo-transistor is a standard NPN. I guess like Mechatronic's said "It's only a schematic and not spiced". Sort of defeating the whole purpose of spice though.
     
    Last edited: Oct 17, 2009
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Yeah, I noticed that. I installed TI-Tina a couple years back, but it seemed awkward to use, and didn't have certain components in the library (like optocouplers) that would've been very nice to have.

    I'm not complaining at all that Mechatronic didn't simulate it. The presence of a schematic helped a great deal to get down to a problem resolution quickly. Too bad I didn't notice that the load was 7.2 Ohms instead of 72 Ohms; that would've saved some time.
     
  18. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Honestly, I also saw 72 Ohms. As far as Tina is concerned, the pro versions do provide two models of opto-isolators, but both models are darlingtons. I have not yet learned to use the Import Wizard to add my own models. It ticks me off that something as common as a 2N3055 is not in the library. However, Tina isn't a US based company. On the other hand, Macro Models are fairly easy to make.
     
  19. Mechatronic

    Thread Starter New Member

    Oct 15, 2009
    9
    0
    Really appericiate the help Gents. It was extreamly helpfull
     
  20. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Glad that helped. Did you get it working with values similar to what I had in the simulation? Or did you wind up doing something else?
     
Loading...