Triac in spot welder controller issue

[Nicky3087]

Hello everyone, this is my first topic, I am very new in the electronic world, meaning that my knowledge is almost in existence.
I built a spot welder for batteries out of an used MOT and this controller (https://s.click.aliexpress.com/e/pB899WTW). The first controller that I bought was apparently not working properly as it delivered to much power and te pulse at the electrodes was rather not constant (kind of that when it reached the max amperage it whent to zero power many times during the duration of the pulse). I ordered another one and it was working fantastic up until yesterday when it started to do the same as the previous controller. My questions: is there a chance that a back inductive spike could have damaged the triac (bta100 800b)? How can I know if the controller includes a protection against this? Thanks everyone!

[autotel]

I don't have any experience at all regarding this, but since there are no answers yet, perhaps it will be better than nothing.

One time I was doing a driver for a rather big motor (~8 A), which was working with a lot of stutter. I was using a PWM controller, and I wanted to check how important was induction in this problem. I added a diode from the driver chip to the pump, making it impossible for the pump to add voltage "back" to the controller. As you would expect, the diode got very, very hot; but the pump started running smoothly as expected. On the anode side of the diode I could see a very noisy pwm signal, and on the other side, a clean pwm.

You could try the same, add a diode (one that wouldn't blow up right away) that could prevent inductance to affect the driver circuit and see how important inductance is. Please bear in mind that I have no idea about the necessary security measures, so research a bit on whether there is anything dangerous about this.

[Nicky3087]

This is an AC circuit, would a diode work?

[autotel]

I am sorry, I am out of knowledge on this...

[MagicSmoker]

...My questions: is there a chance that a back inductive spike could have damaged the triac (bta100 800b)? How can I know if the controller includes a protection against this? Thanks everyone!

There does appear to be some nominal attempt at snubbing on the board (a series RC network across the triac main terminals), but who knows if that is sufficient.

Basically, the capacitor in such a snubber absorbs the energy stored in the transformer leakage inductance* when it is turned off, charging up to a voltage dictated by the energy balance between them (0.5 * LI² = 0.5 * CV²).

The series resistor is selected to limit the peak voltage experienced by the triac when the energy from the transformer transfers to the resistor as well as limit the discharge current into the triac when it next turns on, so there is a bit of a conflicting requirement here, but the value is usually not terribly critical. The BTA100-800B triac is supposed to be rated for 800V and 100A (I'm very skeptical of both specs) so a resistor value somewhere between 4.7R and 47R will be fine.

For example, if the leakage inductance is 100uH and the primary current is 20A when turned off then the energy stored is 0.02J. If we want to limit the peak voltage across the capacitor to 400V then it has to be at least LI² / V² = 250nF (0.25uF), and if we want to limit the peak discharge current to 40A then the resistor needs to be 400V / 40A = 10R.**

So look at the red blob next to the gray metal oxide resistor with the color code orange - white - black - gold to see what it's value is.


EDIT - no, a diode won't work. EDIT 2 - edited for clarity.

* - leakage inductance is that which remains across any one winding when all the other windings are directly shorted. It arises from not all the magnetic flux in one winding linking to another winding, hence it has "leaked" out of the winding, so to speak.

** - I've made some oversimplications here; in reality, the capacitor voltage will follow the mains voltage with some lag and the energy stored in it from the leakage inductance will discharge back into the mains while the triac is off, but worst case for discharge would be immediately turning on the triac again before the capacitor has any chance to discharge so that is what the calculation assumes. The RC snubber also limits voltage rate of rise which helps the triac turn off inductive loads, but, again, I'm oversimplifying here because it's the beginners section.

[Nicky3087]

Hi, thank you for the answer. The capacitor is a CBB81  1259V 103J. For what I found is rated to 0.15uf meaning that I'll have to replace it with a higher value one?

[MagicSmoker]

Hi, thank you for the answer. The capacitor is a CBB81  1259V 103J. For what I found is rated to 0.15uf meaning that I'll have to replace it with a higher value one?

What's the current and leakage inductance? What dV/dt and dI/dt can the triac tolerate? If you don't know all of these things then you can't really calculate the RC values, just guess at them.

That said, 103J is the value/tolerance code and it means 0.01uF (10nF) and +/-5% tolerance, which is way too small for a mains frequency snubber for a triac supposedly rated at 100A. You want at least 0.1uF (100nF) there (needs a 3W resistor), up to 0.33uF (needs a 7W resistor), and the resistor value should be dropped to something between 10R to 22R and of the metal oxide (currently used) or vitreous enamel type; don't use a wirewound.

[Nicky3087]

Hello guys I wanted to update, the triac controller is working a charm, I upgraded the snnuber circuit as per the instructions of MagicSmoker. I also added a varistor in pararel to the triac. I am not sure if the values that I choose for the capacitor are OK but it seems to be doing the job. Thanks!

[maxmad]

What are the new values of R C that you used?