Resistance spot welding, (solid state controller SCR)

[Ante]

Hi everyone, i have a resistance welding machine its very old. and automatic (Circuit board) is bad. and i plan to set a PLC for controlling mechanism and control welding. So in my case controlling mechanism is easy, but i dont know
how to control SCR to start welding, what impulse to give to SCR, becouse i have not any drawing connections. i have pictures of SCR, i i dont know what is the part from right side of it. So the main question is, does anyone know wiring diagram of this SCR.
SCR is working before, but now i was planning to set manual switch to start weld throught PLC. but i dont know what is triggering SCR, which connections. And from left side is a large green resistor, for what is it?
this is my first time with SCR'S.

Its water cooled SCR.

kind regards.

 

[eddie49]

Hello and welcome to the Forum !

I think that the component to the right is a pulse transformer. It is used to provide galvanic isolation of the low-voltage timing and trigger circuits from the high-voltage, high-current SCR and welding transformer circuits.
It has one primary ( twin red wires ) and two secondaries ( two sets of twisted red & black wires ). The secondaries go to the gates of the SCRs - which normally require about 5 volts to trigger on, positive to the gate and negative on the cathode.

The primary needs pulses of maybe 5 to 10 volts, synchronised to the zero-voltage crossing point of the AC supply waveform. If in addition to On/Off, you also want to control the welding power, the pulses should start after a delay from that point, to provide variable phase control.

I don't know about the green resistor, perhaps you could show how it is wired.
Instead of re-inventing the faulty control board, maybe it can be fixed? For this option, please describe the original fault and post clear photos of both sides of the board.

 

[Ante]

Thanks for reply, Okay i understand, my Circuit control board of machine isnt it just for welding, but for all parts of machine (pneumatic valves, switches, operation manual automatic...) so Mechanism its start to behave strange and sometimes didnt activate valves like it should, so my plan is to set plc to machine to control mechanism of machine, and also control weld cyclo, I just need parameters like you said (primary voltage, impuls ).

In attachment you may find picture, i draw schematic of welding, as you can see on it, i have just (2x red wires) inputs to welding circuits.
But main problem is that i dont know what voltage goes to that inputs to trigger spot weld.
doesnt have any scheme of machine its problem.

Should inputs be like (0-10V triggering about 100ms), like if i sent 5V on inputs (+5 AC? ) in short amount of time 100ms spot weld is activate, or i just give impuls 5V, then SCR have his on time (his own triggering impuls for weld) or that triggering time i must give throught that 5V and 100ms?
What else i dont understand is, what is part right to SCR? where wires connected it, is it some sort of starting relay of SCR?

 

[eddie49]

Thank you for your drawing of the welder schematic. It shows that the large green wire-wound resistor is in parallel with the welding transformer. This is probably to provide a slight resistive load for the SCR controller, since SCR circuits don't perform well on pure inductive loads.

The two SCRs are packaged in the "Hockey Puck" style, sandwiched between two water-cooled cold plates. The devices are probably rated at 400A or more. I assume the machine runs on 380 volts, what is the supply current draw?

At high currents and high rates of change in the current ( di/dt ) it is important to get the triggering right, otherwise the devices may be damaged. The part numbers of the SCRs may be visible on their sides, between the cold plates, and knowing this will help to select the trigger voltage. I would advise against dismantling the assembly.
The pulses required to drive the SCR gates, via the pulse transformer on the right of the SCRs, are not AC, they are DC. They must be synchronised to the AC supply frequency of 50 Hz. The pulses could be about 20 microseconds in length, and the welding power will be determined by how soon after the start of each half-cycle of the AC waveform the pulse is sent. The SCRs do not have their own on-time, they will turn off at the end of every half-cycle of the AC supply and must be re-triggered.

Even if the control board does not work correctly for the mechanical and pneumatic functions of the welder, the AC power control section may work, and it will be a lot easier to get this working again rather than designing a complete new circuit.

 

[eddie49]

The SCRs are wired "back-to-back" ( anti-parallel ) so that together they can conduct both the positive-going and the negative-going halves of the AC supply. If you search for "Phase Angle Control" you will find descriptions of the method for triggering the two SCRs ( thyristors ) for a variable time during every half-cycle of the waveform, to provide adjustable control of the spot welding heat. This happens 100 times per second.
In addition, another circuit has to enable these trigger pulses for a variable period, perhaps up to 2 seconds, to provide the spot welding time.
The SCR package is labelled "Scofima" - an Italian company. Searching for this shows examples of the type of unit used to control pressure, weld time, and heat:

http://psr-spotweld.ca/images/Controlex5.jpg

http://psr-spotweld.ca/images/Controlex2.jpg

 

[Ante]

Thanky for reply,
the machine runs on 380V yes, 55KVA max as labeled on machine. I agree with you that this is easier to work it again then rather then designing new circuit for trigerring SCR. i understand the work of SCR so its not easy to trigger it, is it possible to buy new SCR with control equipment? the main thing left for me is try to fix motherboard, but last thing is to set new SCR with new control unit, and also get machine new PLC for machanism. I do some examining the right part of the SCR and i discover that these two impulses provide 15V DC constant, so what trigger SCR, is it (negative period of DC so 0V in short time)?. This is an old machine and every scheme is very old and couldnt see whats going on with thoose impulses and voltages.
1. Pressure for welding i can get from pneumatic valve. impuls 24VDC, its set already how much force we need for good nugget weld.
2. so Power and heat i get from transformers through SCR.
3. its always weld with short amount of time, like 200ms every weld when part is positioned.

 

[eddie49]

The supplied voltage is 380v but that is phase-to-phase across only 2 phases, it is not a 3-phase load. So I think the power of 55kVA means a maximum current draw of 55,000 / 380 = 145 Amps.
By looking at the existing ( faulty ) controller it may be possible to determine which parts are responsible for controlling the SCR phase angle, and the 200ms weld "On" time. Those parts may not be faulty, and even if they are, it could be easier to fix them than starting from zero with a new design. Post photos and maybe we can have some ideas.
Alternatively, there are companies that know these controllers and can fix them:

Scofima System Industrial Automation Products | Lektronix

www.lektronix.com

Assuming the existing SCRs are OK, buying a suitable new SCR switching module, with cooling and phase angle controller would seem an unnecessary complication and expense.

I think that the component to the right of the SCRs is a pulse transformer. It's purpose it to safely separate the high-voltage SCR supply and load circuit from the low-voltage control electronics, so that a fault on the 380 volt side does not feed back through the gates and destroy the controller.
The transformer may have one primary and two equal secondaries, for Cathode1 and Gate 1, and Cathode2 and Gate 2. There will be no active components inside, just a ferrite core and copper windings. If you mark and disconnect all 6 wires you may be able to confirm this by checking DC resistances with an Ohmmeter.
I would expect the two sets of Red and Black wires to have the same resistance ( two identical secondaries ). If the primary ( red-to-red ) has the same resistance value, then it is roughly the same turns ratio, and hence voltage. It would be specified as "1:1:1".

An SCR gate needs a pulse of about 5 volts DC, positive on the Gate with respect to the Cathode, lasting a few tens of microseconds. The SCR will conduct Anode-to-Cathode, and stay conducting even after the pulse has gone, until the supply voltage drops to zero. With an AC supply, this will happen at the end of every positive-going and negative-going half-cycle of the AC sine wave, so the gate has to be re-triggered to maintain current flow. Varying the time that the pulse is sent with respect to the start of the AC wave allows power fed to the load to be controlled.
The 15 volts DC that you have measured is rather high, but possible, depending on the pulse transformer ratio. However, only a pulse can be fed through the inductive coupling of a transformer, so it should only be a short-duration positive-going pulse, not a constant high. You could only see and measure these signals with an oscilloscope, not a voltmeter.
Regarding the "old scheme" -- do you have the circuit diagram, can you post it here?

 

[Ante]

Thanks for informations, well i measure betwen input contacts 15V, constant voltage on these two wires. welding is working so SCR is working, just faulty is motherboard, she controls all machine, I found on my supplier of machine motherboard (just one left from 30 years ago, he said that motherboard is good) so first of all i ordered that motherboard and will see if it fix all machine and mechanism control. if wouldnt then i will pick second choice to buying new SCR with control, becouse machine is expensive and its one in order of another machines in production. In my case it would be easier to somehow control welding, and set PLC to control mechanism.

I dont have circuit diagram of machine thats is the main problem.

Motherboard is controling welding time, current, puses. everything is in that board. But will see, whats will happend with "new one", if nothing i'll be back here with new quesitons to find solution for this.
thank you very much for informations, it is very helpful for me to understand how its working.

I will be back when motherboard arrive for few days.

 

[spotweldingac]

The supplied voltage is 380v but that is phase-to-phase across only 2 phases, it is not a 3-phase load. So I think the power of 55kVA means a maximum current draw of 55,000 / 380 = 145 Amps.
By looking at the existing ( faulty ) controller it may be possible to determine which parts are responsible for controlling the SCR phase angle, and the 200ms weld "On" time. Those parts may not be faulty, and even if they are, it could be easier to fix them than starting from zero with a new design. Post photos and maybe we can have some ideas.
Alternatively, there are companies that know these controllers and can fix them:

Scofima System Industrial Automation Products | Lektronix

www.lektronix.com

Assuming the existing SCRs are OK, buying a suitable new SCR switching module, with cooling and phase angle controller would seem an unnecessary complication and expense.

I think that the component to the right of the SCRs is a pulse transformer. It's purpose it to safely separate the high-voltage SCR supply and load circuit from the low-voltage control electronics, so that a fault on the 380 volt side does not feed back through the gates and destroy the controller.
The transformer may have one primary and two equal secondaries, for Cathode1 and Gate 1, and Cathode2 and Gate 2. There will be no active components inside, just a ferrite core and copper windings. If you mark and disconnect all 6 wires you may be able to confirm this by checking DC resistances with an Ohmmeter.
I would expect the two sets of Red and Black wires to have the same resistance ( two identical secondaries ). If the primary ( red-to-red ) has the same resistance value, then it is roughly the same turns ratio, and hence voltage. It would be specified as "1:1:1".

An SCR gate needs a pulse of about 5 volts DC, positive on the Gate with respect to the Cathode, lasting a few tens of microseconds. The SCR will conduct Anode-to-Cathode, and stay conducting even after the pulse has gone, until the supply voltage drops to zero. With an AC supply, this will happen at the end of every positive-going and negative-going half-cycle of the AC sine wave, so the gate has to be re-triggered to maintain current flow. Varying the time that the pulse is sent with respect to the start of the AC wave allows power fed to the load to be controlled.
The 15 volts DC that you have measured is rather high, but possible, depending on the pulse transformer ratio. However, only a pulse can be fed through the inductive coupling of a transformer, so it should only be a short-duration positive-going pulse, not a constant high. You could only see and measure these signals with an oscilloscope, not a voltmeter.
Regarding the "old scheme" -- do you have the circuit diagram, can you post it here?

Hey Eddy
We also have an 80kvA transformer. We want to drive this transformer with a Scr card and thyristor. Can you help us?
You can send an e-mail if you want, thank you very much.
mehmetz518@gmail.com