Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Adding a 2V DC Bias to AC Mains

  1. Apr 6, 2013 #1
    Need to do something which requires adding a 2V DC bias to power coming out of the wall (120V, 60Hz), going to a motor.

    Basically need the 2V DC bias across the motor windings. Motor windings have about a 10 ohm DC resistance. I have a DC power supply that has a max voltage of about 50V. Pic of my circuit posted below, but I think it was frying the power supply because the impedance on the outputs are really low. I'm wondering if there is any way I can do a similar circuit but with some high resistance before the power supply. If not, any idea how I can get the DC bias across the motor windings?

    Motor represented by the 10 ohm resistor.

    Attached Files:

  2. jcsd
  3. Apr 6, 2013 #2
    Why do you use an air core transformer for mains ac? Also your dc supply is connected in parallel to the transformer and the motor. It should be in series.
    Can you give more details? What is the output voltage and resistance of the transformer and what current can your dc supply output?
  4. Apr 6, 2013 #3
    Its not actually an air core transformer-- I believe its actually an iron core, EI isolating transformer. I don't have details on me right now on the output resistance of the transformer. I just put together that model quickly. Its a 1-1 transformer, so output voltage is also 120V.

    The supply should be connected in series? I'll try that next time, but would that mean all the power is going through the supply?

    DC supply can output ~3A, and from I see, I just need 200mA to get 2V across the windings.
  5. Apr 6, 2013 #4


    User Avatar

    Staff: Mentor

    The transformer needs to have a secondary center tap. Add your 2V offset in there.

    Fair warning -- this thread may be closed as dangerous if you keep posting stuff like air core power transformer at 60Hz...
    Last edited: Apr 7, 2013
  6. Apr 6, 2013 #5
    smk037, do not connect your DC power supply in series! It will damage the power supply. You can use a parallel connection, but you must block the AC from the power supply. You can do this with an additional isolation transformer connected in series with the DC power supply. Leave the other winding of the second isolation transformer open. Connected this way the second isolation transformer will act as a choke, blocking the AC from the power supply. Also, you should add a 10 ohm resistor directly across the output of the DC power supply. This will shunt what little ac makes it through the choke to a negligible level (for ac the choke will be a much greater resistance than the 10 ohm resistor). Also, you will need to eliminate R3 and R4.

    I will try to get back later and do a sketch (if you want to use this method). It will require you to get another transformer. It doesn't have to be an isolation transformer. You can use the primary of most any power transformer (it must be able to pass 200ma of current). It should not be so small that it saturates at 200ma. Probably a 25 watt or greater power transformer would do.

    If it's not too much trouble, it would be interesting to know why you want to do this.
    Last edited: Apr 6, 2013
  7. Apr 7, 2013 #6
    As I was doing my sketch I noticed a problem with my idea. Although it protects the DC power supply quite well it will not work, because there is no way to prevent the DC current from passing through T2. And I'm pretty sure the DC resistance of T2 will be lower than 10 ohms. So I guess we'll have to wait for Berkeman to elaborate on his idea, as I have no idea how a center tap on T2 will give us what we want. This is a more difficult problem than I realized.

    As an alternative I will give you this sketch, which is not exactly the same thing as a DC bias but it should have the same effect, depending on what you're trying to do. It also has the advantage of being very simple.


    Since each diode will have a 0.7v drop, the effect will be that the motor will see a 2.1 volt difference form one half of the cycle to the other. So more current will flow in one direction than the other.
  8. Apr 7, 2013 #7

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Now you're getting someplace, Turtle !.
    Asymmetric sinewave will have a DC content.

    Do not be surprised if the motor and transformer get hot. Feel them after a few minutes , and after an hour.

    If you can monitor current, with say one of these and a 'scope, be watchful for high peak current .



    Asymmetry in relay drivers used to burn up relay coils in my plant. It was a subtle failure. We learned to spot it by measuring DC value of the AC output, if more than a volt we replaced the thyristor.
  9. Apr 7, 2013 #8
    After receiving more info from smk037, it appears that an asymmetric siinewave test is actually what the he is looking for. So the circuit shown in post #6 would be the way to go. A motor might be subject to this type of condition (asymmetric sinewave) if it is used with a thyristor type speed control. And the op wants to test the motor for this condition.

    I once worked for an ac induction motor manufacturer. We never had to test our motors for this. However, that was over 30 years ago. Even though our motors were not designed to be run with thyristor speed controls, there was no doubt that some people did it anyway, especially with the 18 pole ceiling fan motors. So it's a good idea to test them for it.
  10. Apr 8, 2013 #9
    Thanks guys, I got the circuit working.

    Now, I'm not entirely sure I understand the reasoning behind this test (it is a UL requirement which is why I'm doing it).

    It seems to me that 2V across a winding with ~10ohm resistance would not make much of a difference as far as heat goes. Could someone explain why this might cause some heat issues? I'm guessing it has more to do with the assymetry of the AC signal, but I'm not sure I quite understand why (am fairly new to working with AC motors).

  11. Apr 8, 2013 #10

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    My relays had coil resistance of ~100 ohms also.
    Our asymmetric thyristor firing caused them to draw high current on one side of the sinewave .

    I first found it by observing coil current with a clamp on current probe connected to an oscilloscope. One side of sinewave was nice and smooth like you expect, other had a sharp peak as if it were approaching an asymptote.
    That current peak gave the wave more heating value than a sinewave of same average value.
    Think about what RMS does - you work it backwards - Square , take Mean(average), then square Root of that mean.
    Squaring a narrow peak gives it tremendous weight in the Mean calculation that follows.

    The reason it draws high current is average flux in core is not zero, so it runs up close to saturation on one side of the sinewave. So it takes more current to maintain [itex]d\Phi/dt[/itex], hence that sharp current peak with high RMS content.
    That's the answer to your question about how it can cause harm.

    A volt of DC offset did not heat my relay coils much. But four or five volts would burn a coil up in a week, and that tripped the plant several times. So we examined every single relay's coil current . The clamp around current probe allowed us to do it non-intrusively, otherwise it would have been an impossible task.
    This was a sneaky problem as the thyristors worked fine on the bench at room temperatue, but when the relay driver was placed in the equipment rack and warmed up it would develop that DC offset. Take it back to shop and it worked fine again.
    Needless to say the operators were happy when we found that ! They'd all come over to see the 'scope trace whenever we found one in the plant.

    I urge you to experiment with a 'scope and current probe.
    If you don't have a current probe you can use a 1 ohm(or less) resistorin the neutral leg. I have used a meter shunt for bench testing.

    But be darn sure you are in the neutral!


    Sorry for the boring anecdote. Maybe it'll help somebody .

    old jim
  12. Apr 8, 2013 #11

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Since you have a test setup you could cause more DC offset and find the point where your motors complain.
    I'd be curious whether a motor is less affected because it has more airgap than a relay.

    It's just an interesting piece of trivia that might someday come in handy. That's my problem - my alleged brain is overloaded with them.

    old jim
  13. Apr 8, 2013 #12
    I am not entirely sure here, but I think with ac equipment and coils the problem is saturating the core. A fairly low dc current seems to be enough for the core to saturate. Usually the current drawn is limited mostly by the inductance of some coil with a core. If you get into a region where the core starts to saturate, the inductance will go down, the current will go up and the core saturation goes up even more. Maybe the problem lies somewhere there.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook