constant current device

Discussion in 'The Projects Forum' started by stirling, Aug 16, 2011.

  1. stirling

    Thread Starter Member

    Mar 11, 2010
    52
    2
    Hi - I'm looking to undertsnad how a constant current device works. Specifically in a plasma cutter.

    Basically, on a plasma cutter you set the current that you want to cut at. Then if the torch moves further away from the metal you're cutting, the plasma supply electronics will increase the voltage to keep the current constant (because the larger distance between the torch and the work presents a greater resistance accross the plasma arc). Similarly if the torch moves closer to the metal, the supply will decrease the voltage to keep the current constant accross the plasma arc.

    The reason I'm asking is that previously I'd assumed that the plasma supply electronics responded to a change in distance via a change in voltage and "instantly" caused an appropriate change in voltage to maintain constant current. However, experimentation suggests that as the torch to metal distance changes, the voltage change actually lags behind "a small amount of time" - i.e. there is a phase angle between the distance changing and the voltage changing. I'm trying to establish if this is in fact the case.

    Any insight welcome.

    Thanks
     
  2. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,770
    970
    Typically hysteresis is built into many electronic circuits.. I suspect that is what you are seeing in this case. Also the more "accurate" a circuit is the more it costs..
     
  3. castley

    Member

    Jul 17, 2011
    31
    0
    Think of your plasma cutter as a varying load with sharp variations. A constant current source with adequate sebsitivity should be able to follow the current demands.
     
  4. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Torch height control is often used in mechanized plasma cutting to improve cut quality. This is typically done by sensing / controlling the torch electrode to plate (work) voltage whilst applying mechanical 'z' axis standoff control. The control strategy can become somewhat complicated when cutting profiles with sharp corners etc.

    Plasma cutting power sources are regulated current drive devices rather than voltage regulated devices. Controlling plasma arcs with voltage control is difficult because of the arc electrical dynamics, which can exhibit negative resistance behavior.

    So you can in fact have (at least) two control strategies going on simultaneously.

    You could Google a plasma cutting system manufacturer to possibly get more info.

    I found this useful link ....

    http://www.centricut.com/Training_Articles.htm

    http://www.centricut.com/New_Lessons/lessons_10.html
     
    Last edited: Aug 18, 2011
  5. stirling

    Thread Starter Member

    Mar 11, 2010
    52
    2
    Thanks for all your replies.

    mcgyvr - agreed - but what I'm after I guess is: any arbitrary current limiting system has to sense a change in current and then calculate the required voltage to change to, and then apply it, it MUST take a finite time - but (and I know I'm asking a lot here) - typically, does anyone here KNOW the order of how long could that take? a nano second - a microsecond - a millisecond?

    castley - again - understood - but I'm looking for figures if possible.

    t_n_k - you're kinda jumping ahead. I've built several THCs with help from a leading cutter manufacturer and I've got the art down to being "very good". BUT they all exhibit a "niggle" in common with many commercial THCs. My theory that I'm trying to prove or discount is that there is a delay between current change and the corresponding voltage change in the plasma unit itself in the order of several milliseconds - which to me seems unexpectedly long. I could be quite wrong but that's why I'm asking.

    As I'm now "winding back" to look at every detail of the process right back to the plasma cutter itself, understandably the manufacturer of the cutter is not keen on divulging too much detail.

    Thanks
     
  6. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,770
    970
    What you are trying to find is the "transient response time" which is dependent on the specific circuit. Could be down to 50 microseconds but doubtful.. Typical constant current power supplies are usually 1-10 milliseconds.
     
  7. stirling

    Thread Starter Member

    Mar 11, 2010
    52
    2
    OK thanks - So it looks possible that I may be barking up the right tree. Much appreciate your help.
     
Loading...