A better control board

  1. R Kraft Member

    Posts: 791
    Wyoming,USA
    If someone has an old SIP wire feed control board,working or not,I would like to get it.
    If shipping is not too high that is.
    It seems to me that adding and moving a few components, would make the thing work far better.
    I would like to see if a cheap easy fix could be made.
    I could pay shipping with Paypal.
    Robert
     
  2. stuff to fix Member

    Posts: 25
    uk
    I dont have a board but I am interested in the question of designing a better one. Are you willing to say what you have in mind?

    I had a look at the sip website where there is a rather blurry online version of the manual for the newer migmate T155/T135. The control board is significantly more complex than the old ones, though the diagram is too illegible to make out component values or even some components. If anyone has a better copy, that would be very handy.

    They seem to be keeping the same fundamental idea that it is a good idea to drive the motor from the weld output, so that it automatically goes faster if the weld voltage is increased. However, they use a thyristor to switch power into the motor. There is a separate bit of circuitry which seems to have a stabilised power supply (hard to tell, because cant read the component values) though still derived from the weld voltage. A control potentiometer drives a transistor vaguely like the old circuit but more complex, to fire the thyristor. I am half guessing the idea is to use stable power applied to the control section to give a better control of the AC part still powering the motor via the thyristor. They use a relay to switch between the motor connected to power or shorted out through a resistor to provide immediate braking.
     
  3. R Kraft Member

    Posts: 791
    Wyoming,USA
  4. R Kraft Member

    Posts: 791
    Wyoming,USA
    The SCR type circuits are based on an RCA or maybe it was a GE circuit from the 1960s.
    The transistor measures the back EMF from the motor and inhibits the firing signal on the SCR gate.

    Implemented properly ,it is a good motor drive circuit.
     
  5. stuff to fix Member

    Posts: 25
    uk
    The manual is at http://www.sip-group.com/acatalog/04796.pdf

    It is not the same as the one you have posted. In the sip one the motor is connected directly to weld voltage through the thyristor when it has fired. On the face of it that seems a bit odd, but hey. When the thyristor is off, I am not clear what is happening because I cant see what are big capacitors, etc. The circuit might be capable of sensing the back motor emf while it is free wheeling (thyristor off), or the motor might be being fed by charge stored on a capacitor. The thyristor is bypassed by a r and c in series, but I dont know what they are, whether this is supplying any significant current even when the thyristor is off.

    My guess, however, is that when the motor is on and power coming in, the circuit charges up a capacitor which powers the base signal of the transisitor depending on the pot setting. There is a second capacitor immediately on the transistor base with a parallel resistor and this is fed through diodes so it cannot discharge through the pot. This might be to hold the base voltage once charged until the next cycle. The transistor is connected to fire the thyristor when the weld voltage exceeds the voltage set on the transistor base by a sufficient margin.

    However, if the motor is freewheeling and generating a back emf, this would hold off the thyristor until the weld rectified ac exceeds the back emf. Except that it is not clear to me the motor would generate a decent smooth DC signal, so could not reliably hold off the thyristor. If it does, it might mean the motor gets full volts starting from stationary, so fast acceleration, but then succcesively less power each cycle as it gets faster with more back emf. Im not convinced about that.

    There is one of those mov things shorting the transistor, so this is presumably in effect setting a minimum current. When the rectified ac weld voltage exceeds the MOV by a sufficient margin (depending on ratio of resistors I cant read) the thyristor will fire setting things going.
     
  6. R Kraft Member

    Posts: 791
    Wyoming,USA
    Pulsing DC voltage is fed to the control board through connections P2-P5.
    Positive goes through a resistor and a diode to the gate and turns the SCR on.
    When the Voltage on the motor (p6) gets high enough the positive charge turns the transistor on and shunts the SCR gate drive to ground.
    Since the capacitors, diodes and resistors provide a varying phase shift it enables the SCR to fire for differing number of degrees of the 120hz waveform.
    Each time the voltage from the main transformer drops low enough,the charge carriers drop out and the circuit resets until the next cycle.
    Robert
     
  7. stuff to fix Member

    Posts: 25
    uk
    Right thanks, I had misremembered which way round the gate on the scr works and was thinking it had to go down wrt anode to fire, rather than go up wrt cathode. Lets try again. Let me know what you think of this one.

    So from cold start, voltage from P2 gets to the scr gate via a resistor and diode. However, the circuit has a resistor and capacitor in parallel connecting the gate and cathode. This locks them at the same voltage. The stalled motor is shorting the cathode to ground, so current flows through this and this last resistor, charging the capacitor until the difference between gate and anode is enough to fire the SCR.

    I can not see any resistors in this path, so full voltage from P2 is applied to one side of the motor through the SCR. The other terminal P5 seems to connect to the motor through a fuse? There are relay contacts marked which in the position shown short out the motor and disconnect P5, so to turn on the motor they must have swapped around by presumably the torch button press working the second relay shown higher up in the diagram. So I reckon full voltage immediately appears across the motor when the scr fires.

    we need a diagram at least with part numbers so we know which item we are talking about.

    With the relay which is marked as closed now open, there is just one other resistor connected to the SCR cathode/P6. This looks to me to be feeding three things drawn in parallel, a small capacitor, an electrolytic capacitor and a reversed diode. I dont know what the diode is? a zener perhaps to limit maximum voltage, or a protection diode to short out reverse voltage from somewhere? I think this set of components is simply charging up to weld voltage limited only by the conduction on the SCR and its own feed resistor. I would say the back voltage of the motor would be insufficient to block anything (since its not moving yet/moving slow). The full weld voltage is effectively being applied across the resistance of the motor windings. If the diode thing is a zener, it might be limiting the maximum voltage at this point, so that effectively this is a mini regulated power supply for the adjustment section. Each cycle, the big capacitor will charge to this zener voltage and then go into operation.

    The adjuster pot feeds a capacitor on the base of the transistor through diodes. So whatever maximum voltage is reached each cycle, a fixed proportion of it gets transferred onto that capacitor and is held there. I would guess the time constant is enough so it will discharge slowly over a few mains cycles, but enough to hold it fairly steady from cycle to cycle, being topped up through the diodes each cycle. I think this part of the circuit is not getting any feedback from the motor because it is fed at full weld voltage limited only by the SCR resistance (I would think negligible?)

    The transistor base has now been set to some voltage depending on the pot position. The weld voltage goes to zero as the mains cycle ends, current stops flowing and the SCR turns off. We start a new cycle. The capacitor between gate and cathode of the scr discharged after the scr came on through the r and d connecting the gate to P2 so it is now holding zero volts g-K and the scr is not being triggered.

    Input voltage starts to rise. This time the transistor has a base voltage already and starts to conduct. This holds down the voltage appearing at the gate of the SCR. The cathode is still shorted to ground through the motor and the R/C combination between gate and cathode starts to charge and build up a voltage to fire the SCR. However, it will take longer than last time because the voltage at the gate is also lower, thanks to the regulating transistor conducting. Power will not be connected until later in the cycle, so in effect we have a pulse width modulated speed control.

    The rate of firing of the SCR will also be affected by the motor speed. If it generates a back voltage because it is already running, this will increase the voltage showing at the SCR cathode, and slow the charging of the firing capacitor. If for some reason the motor slows down (wire rubbing?), then there will be some automatic feedback triggering it a bit earlier and trying to speed it up again. I dont know if this is significant or negligible, but in the earlier simple circuit if the motor slowed the feedback tended to further reduce its feed current, not increase it.

    The motor final speed will depend upon the total applied voltage as well as the proportion of time it is connected each cycle. So changing weld range bumps up the voltage and makes it feed faster, automatically.

    I have not accounted for two components which short the SCR. A resistor and possibly a capacitor in series.
    [edit: found the explanation: a scr can be triggered incorrectly by a sudden voltage change between K and A. Therefore a 'snubber' is placed across it to limit transient voltage changes and prevent mis-triggers. These two components would not therefore affect the basic explanation of how the circuit works]
     
    Last edited: Jul 7, 2011
  8. R Kraft Member

    Posts: 791
    Wyoming,USA
    The resistor and capacitor across the gate would provide a filter against false triggering.
    The resistor and capacitor in series are a snubber circuit, to protect the SCR from spikes,and prevent damaging triggering conditions.
    Snubbers and varistors are important in welders,the inductive kick from the relay coils and the output reactor choke will kill the board without them.
    Robert
     
  9. stuff to fix Member

    Posts: 25
    uk
    To get back to the original board and trying to mod it without having to build a giant new circuit. I think one of the things wrong with it is that there is no stabilisation of the voltage applied to the transistor base. The + supply comes from the weld voltage, but the - is all connected to the input to the motor. The circuits you posted and this new one from SIP all connect the transistor drive directly to the 0V terminal from the weld supply. I think the transistor should connect to the motor but the preset which sets the minimum speed ought to be connected to 0V. This would no doubt mean changing some values around to get the voltages right.

    But I dont really understand how the circuit is intended to work! The MOV on the board might just be protection against over voltage, but I am still not sure if it could be intended as part of the power stabilisation. If the charteristics were right, it might be stabilising the voltage across the transistor. If the transistor conducts more, the MOV would take less current and voltage drop across it would fall, so more voltage would be applied to the motor. Does anyone know what values it is?

    The Sip circuit also uses relays to switch the motor in and out of circuit, which some have suggested here as a good mod. This allows you to add some smoothing capacitors on the board. I think that although the new sip circuit does not have big smoothing caps on the motor, it does have much better smoothing for the control circuit, and this will result in a smoother supply to the motor. In one of your circuits, the time constant feeding the transistor base is 220 ohm feeding a 100uF capacitor, which is 22mS, whilst the rectified ac has 10ms period. The other circuit uses 110 ohm, so 11 ms.
     
  10. R Kraft Member

    Posts: 791
    Wyoming,USA
    The Cemont type board should work far better than the old sip board.
    I suppose if you wanted the voltage to the circuit be more uniform,you could add a clipper type regulator to lead P2.:http://www.mig-welding.co.uk/forum/album.php?albumid=918&pictureid=4178
    But I have not heard any comments on the new board,so I really can't say whether it needs work or not.
    Robert
     
  11. stuff to fix Member

    Posts: 25
    uk
    I have found another controller circuit here http://www.wizardanswers.com/hfwelder.html which is supposedly for a harbour freight 90A welder. It is similar to the one you posted except it does not derive the power for the transistor base from that going into the motor, rather from a steady +ve supply which is also powering the motor. This is derived from a small bridge, on the weld transformer ac output.
     
  12. R Kraft Member

    Posts: 791
    Wyoming,USA
  13. For what it's worth, I did a quick search for "PWM" ( pulse width modulated ) on eBay.uk, and came up with Item 320683792011. This is a 12v - 24v, 3 Amp PWM motor speed controller, complete with potentiometer, providing 0-100% speed control, costs £5.90 and £2.50 shipping from China. At that price, and assuming PWM would be the best way of controlling a wirefeed motor, it hardly seems worth going down the home-made route.
     
  14. stuff to fix Member

    Posts: 25
    uk
    Does anyone know what the rating of the MOV on the controller board is? (usually a round yellow thing in the bottom left corner on the pics I have seen)
     
Recent Posts