Hi! I need some help designing an RC snubber circuit to go across my SCRs. I have 2 SCRs connected antiparallel to each other which conduct one after the other. I also have one RC snubber circuit connected parallel to these SCRs.

I already went through the design guidelines found online where you are supposed to find the intrinsic inductance and capacitance of the circuit using the intrinsic ringing frequency. The results I got were my intrinsic inductance is around 47 micro H and intrinsic capacitance is around 37 nano F.

Using the values of this capacitance and inductance i found a range in which my RC components should lie. R<35 ohms and 3.5 pico F < C < 1 micro F.

I have tried experimenting with different values of RC in this range .
I have 2 capacitors available of 225 nano F, using them i make 112 nano F, 450 nano F, and just 225 nano F.

There is still some voltage overshoot and ringing when the SCRs turn off, I can't seem to find the RC values that will snub the ringing completely. Some combinations help reduce the voltage overshoot, and some combinations reduce the damping but increase the overshoot. I am confused as to how to select an RC from the range I obtained.


max switching frequency: 100Hz
max voltage : 1400V
SCR holding current: 0.4A

I'll attach some photos of the ringing that i observe (the 2 pictures show ringing at different snubber circuit combinations).

 

An interesting question, with quite a bit unknown. Where in the circuit, which we do not see, are the snubber elements connected?? and what is the load?? Is the load reactive or resistive??
Snubber requirements depend on the load, and without knowing that there is no way to know.

 

See Fairchild AN-3006, AN-3003

Also:-

 

Given that we are not told what the load is, or what the load reactance might be, nor what the source impedance is, nor the voltage, all of the advice is a guess at best. Is this an SCR inverter? Or a light dimmer? or something else???

 

A simple way that I have found is to make the Resistor value the same as
the DC-Resistance value of the Load,
then,
while watching the waveform with an Oscilloscope,
( start the design process with a Circuit-Simulator-Program ),
keep increasing the Capacitance value until the spike, and or, anomalies are minimized.

Too much Capacitance is just as bad as not enough Capacitance.
The Capacitance must balance the actual operating Inductance-characteristics of the the Load,

Math-calculations of the Inductance, and Switching characteristics,
quite often, do not reflect the "real-world", "actual" behavior of the Circuit.
.
.
.

 

At one point I resorted to this: https://www.st.com/en/thyristors-scr-and-ac-switches/standard-and-snubberless-triacs.html

Life became much simpler a

 

Certainly a device with adequate ratings for the application does not require a snubber. Likewise, in a resistive device control installation, there are not a lot of inductive spikes, nor capacitor issues. But preventing electrical noise is still required.

 

A simple way that I have found is to make the Resistor value the same as
the DC-Resistance value of the Load,
then,
while watching the waveform with an Oscilloscope,
( start the design process with a Circuit-Simulator-Program ),
keep increasing the Capacitance value until the spike, and or, anomalies are minimized.

Too much Capacitance is just as bad as not enough Capacitance.
The Capacitance must balance the actual operating Inductance-characteristics of the the Load,

Math-calculations of the Inductance, and Switching characteristics,
quite often, do not reflect the "real-world", "actual" behavior of the Circuit.
.
.
.

According to my calculations C = 74nF and R = 18ohm so experimenting at 800nF I probably might have overestimated the capacitance i need, I'll try following what you have suggested and see what happens. Thanks!

 

At one point I resorted to this: https://www.st.com/en/thyristors-scr-and-ac-switches/standard-and-snubberless-triacs.html

Life became much simpler a

Thanks for mentioning these!

 

Your Schematic does not show any Power-Source.
.
.
.

 

Your Schematic does not show any Power-Source.
.
.
.

The power source is in parallel to the setup. It is disabled when the thyristors are operating. You can assume the capacitor to be charged to 1400V already.

 

I started from 70nF and went in increments of 100nF up till around 600nF (resistance being fixed at 20ohms) still had no luck snubbing the spike. The best results i got were around 600nF, so around the higher end. However, after a point adding more capacitance doesn't seem to help my case

 

What does your Circuit do ?,
it's impossible to tell what it does,
especially without a Power-Supply shown.
.
.
.

 

Why not to add a MOV varistor in parallel to RC snubber? It will limit the overshoot easilly

 

The snubber should limit the rate of voltage change as well as limit the maximum voltage a MOV cannot to that.

 

What does your Circuit do ?,
it's impossible to tell what it does,
especially without a Power-Supply shown.
.
.
.

Sorry to give you guys information in pieces.

So, the circuit first charges my output capacitor (56uF) through a coil (51uH), then the power supply is disabled and the thyristors are fired one after the other. When the first thyristor is fired the capacitor discharges through the coil (LC oscillations start to happen). After the capacitor is done discharging through the first thyristor, the current through the LC reverses and the capacitor starts charging back up through the anti-parallel thyristor. Then when the capacitor is now done recharging, the 2nd thyristor turns off as well and this is when I get a turn-off spike.

It's a magnetic stimulator.

 

The snubber should limit the rate of voltage change as well as limit the maximum voltage a MOV cannot to that.

I saw several times how solid state contactors are made, the RC snubber plus MOV in parallel to the thyristors/triacs

 

I am unable to open the simulation file without that simulator program, which I do not have.
And after looking at the circuit fragment in the earlier posting and then reading the description of how the circuit is intended to function, it seems to me that you do not need a snubber to achieve the magnetic field formation.
Without a compete circuit, at least the power portion, there is no way to do an adequate analysis.
I am not able to discern what is in other folks minds and so can not provide any more advice.

 

Where is the Load ???
Please give a fully detailed description of the Load.
.
.
.

 

Where is the Load ???
Please give a fully detailed description of the Load.
.
.
.

L1 and C1 are the loads. I was wondering whether using an RCD clamp would be more beneficial inorder to reduce the voltage overshoot.