transistor connected to actuator

Discussion in 'The Projects Forum' started by J.Green, May 5, 2011.

  1. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    I have a 12 VDC actuator. The actuator’s direction is controlled by reversing the polarity in the two leads connected to the actuator. I want to tap into each of these two leads and connect a transistor, to use as a switch, to control an input into a micro processor. When power is applied to the actuator to move in direction X, I want the transistor to act like a closed switch, being closed for the duration power is applied to the actuator. This will ground the input into the processor, bringing it low. I will do this by connecting the base of the transistor to +12V from the actuator, connected via a suitable resistor. The Emitter will be connected to the actuator wire that is grounded. When power is applied to the actuator to move in direction Y, I don’t want anything to be done by the transistor, except act like an open switch, therefore leaving the processor input high.


    At first I didn’t think I need to worry about reverse polarity at the transistor. On a breadboard however, I was getting unexpected current entering/leaving the Collector with the Base was grounded and the Emitter connected to +12VDC. I thought if the polarity of the transistor was reversed, no current would flow into/out of the Collector. I tried connecting a diode upstream of the Base (Option A in sketch), but still measured a small current, 0.3mA, entering/leaving the Collector when the polarity was reversed. Connecting a diode downstream of the Emitter (Option B) seemed to solve the problem. I understand the Option B will drop an additional 0.7V, but I am not worried about that for this application. Or I don’t think I need to be worried about this, as all I am using the transistor for is to pull a processor input low.

    Three questions:

    (1) I would appreciate it if someone could shed some light on why a current is flowing to/from the Collector in Option A when the polarity was reversed.

    (2) do I need to worry about reverse polarity protection here?

    (3) is Option B the best way providing polarity protection here, given the Emitter could be either connected to +12V or ground, depending upon the power supplied to the actuator and that I do not want to use a diode in parallel with a fuse.

    Thank you.

    JG
     
  2. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    I think you need something like an H bridge or a changeover relay.
     
  3. EdGs

    New Member

    Mar 10, 2011
    17
    2
    I would use a resistor to pull the input high when the transistor is "off".

    It would seem that there will be a good bit of noise coming from the actuator power too. Is the microprocessor driven by the same source as the actuator?

    If the processor power is separate from the actuator, I would use an optoisolator such as the PS2501. You can run the emitter of the opto from the actuator power, and the ouput would only be on when the actuator is running whichever direction you want.

    I still would use a resistor to pull the input high when the opto is off.
     
  4. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    The actuator is already being controlled. I am looking to pull a pin on a processor low when the actuator moves in one direction.

    I would like to use only 3 wires, by sharing the ground for the Emitter. I have used opto couplers before to remove noise entering a processor. I could use a opto coupler here, but it would increase the wire count and make installation a bit more complicated (not by much though)

    Anyone have any thoughts on the transistor approach to this application?

    Thank you.
    JG
     
  5. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    Maybe you could explain HOW the actuator is being controlled. Is one end at ground and the other tied to + voltage for one direction and switched, end for end power wise, for the other direction? Is power maintained to hold position, or pulsed to make the move, then disconnected? The more details about what you have and want, the better the answers.
     
  6. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    Yes, one end is tied to + voltage and the other to ground for one direction. The power connections are reversed to move the actuator in the other direction. Power is pulsed to make the actuator move and then disconnected. I have built a model on my bench for testing purposes using two relays and a single momentary SPDT [on -off - on] switch.

    I have attached a diagram similar to the model I built. Ignore the remote control receiver. In the attached I have replaced the 'lift motor' with an actuator. The actuator will be pulsed momentary and then power removed. I want to be able to pull a processor input low (not shown in the attached) when the actuator moves in one direction only.

    Thank you.
    JG
     
  7. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    After looking at your drawing, I have to ask, is there a reason you couldn't use the same signal that drives the relays to drive an opto-isolator which would then feed your processor (one for each directiion)? If you need to monitor actual voltage to the actuator, likewise, you could use opto-isolators tied to the coil.
     
  8. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    Good question. The actuator is part of an existing machine. I have easy access to the actuator and the wires connected to it. I do not have easy access to the device(s) that control the actuator in the machine. I have built the model assuming relays control the actuator, but in the actual machine it could be some other solid state device. Not really sure.

    In short, I only have access to the actuator and wires connected to the actuator. I was hoping to tap into the two wires connected to the actuator to do what I wanted.

    Thank you.
     
  9. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    Looks like Opto-Isolators would be the way to go. By selecting proper series resistors for the internal LED, you can accommodate a wide voltage range. Many varieties of optos are available with different output devices/capabilities, ie transistor, triac, scr, fet,
     
  10. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    Do you see any downside to Option B (see the sketch in the original post)?
     
  11. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    If by "Option B" you mean tapping into the wires on the actuator, I agree it would be the way to go.
     
  12. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    After re-reading you're first and subsequent posts I don't think you need anything more than one of the attached circuits ...
     
  13. J.Green

    Thread Starter Member

    Jul 15, 2008
    26
    0
    I was having some difficulty getting this project working so I picked up an oscilloscope and hooked it up to see what I was dealing with. To my surprise, the voltage in the lines feeding the actuator is much greater than 12 VDC. The voltage is actually peeking at around 150 VDC, settling into around 116 VDC, staying there for 250 ms, and then going to negative 150 VDC for about 1 ms. I ran this a few times and got similar results. I have built a model on my bench and got similar results.

    The opto I am using is an MCT66, max voltage for the LED is around 1.5 volts, with current of 20 mA.

    The voltage in the wires feeding the actuator must be driven high by induction? I'd like some thoughts on this from others please.

    Also, any suggestions on how I can use this 116 VDC spike for 250 ms to the actuator to interface with a PIC micro. Is it as simple as using a resistor to drop 114 VDC, leaving around 2 volts for the opto? 114 VDC / 20 mA = 5.7K resistor?

    I have attached a screen shot from the scope.

    Thank you.
     
  14. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    If you are seeing that high of a voltage, you may want to look into the neon bulb type opto-isolator.
    Here is a DIY, but you may want to substitute a photo-diode for speed.
    http://taweber.powweb.com/store/opto_n_kit.htm

    Obviously, you will want to put a current limit resistor in series with the neon.
     
  15. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Bill, do they make neon optocouplers? Not say'n they don't. Just never seen one. A 4N33 could be used but the limiting resistor would be 15K and dissipate about 1.5W with I= 9.5mA @ 150V input.
     
  16. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    Back in my day, neon optos or incandescent optos were the only ones available. Most of them used LDR or photo-voltaic cell sensors. LED's didn't exist yet. Whether a neon opto is still available or not is a good question. You probably could use a limiting resistor with an LED but the reverse spike would be a concern. A Schottky diode, in reverse direction across the LED to absorb the reverse spike, may work for you.
     
  17. CDRIVE

    Senior Member

    Jul 1, 2008
    2,223
    99
    Gee Bill, back in your day was my day. I just missed them somehow. My first oscillator was an NE2, resistor and cap though. ;)
     
  18. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    A little off topic, but I remember the first digital encoder in my experience. It used very small incandescent lamps and photovoltaic cells mounted on phenolic posts. A glass disk with an optical pattern rotated between them and we had to rotate the phenolic posts to get the 90 degree shift needed for quadrature feedback. The disk also had a single line that would line up with the third lamp/photocell combo to provide a single reference pulse per revolution. To "Home" the machine, it was taken to an extreme limit and when it came off the limit (not very repeatable), the control would wait for the first reference pulse from the encoder and start counting from there. The machine was a Hughes MT-3 milling machine.
     
Loading...