OPamp comparator on ac mains

[anishkgt]

Hi All,

I am trying to use a comparator to monitor the difference in voltage when the secondary of a transformer are shorted when spot welding. This will be like an added feature to trigger the Thyristors when then electrodes touch the nickel strips to be welded.

I am just beginning to learn to use opamps (as comparators) and need some advise on how to get this working. The principle is simple, When the thyristors are not fire and when the transformer does not have any power to it, there is a small leakage voltage of 1.5vac at the primary of the trasnformer and when the secondary are shorted the voltage drops down to 0.5VAC with this voltage drop am hoping to use an opamp as a comparator to monitor the voltage drop and send its out to an arduino to trigger the Thyristors.

My concerns, would it be ok to have the opamp exposed while the actual weld happens. The current at the primary is 20A in short pulse.

[anishkgt]

A second option with an opto isolator

[PartialDischarge]

You need to rectify and filter the voltage you are trying to compare, so you compare the "mean" not the instantaneous voltage.
The speed (step response) of the filter must be matched to the frequency of the line.

Also always use a limiting voltage device at the input pins after the high impedance resistors (like a low current zener). Although you seem to expect that always less that 1.5V will be present I doubt that. There could be switching spikes at turn-off or turn-off of the thyristors where you'll get the full line voltage at the primary.

Make sure the high inrush current in the primary does not reflect on the voltage powering the opamps by using filtering capacitors or blocking inductors in series if necessary.

[eblc1388]

... there is a small leakage voltage of 1.5vac at the primary of the trasnformer and when the secondary are shorted the voltage drops down to 0.5VAC with this voltage drop am hoping to use an opamp as a comparator to monitor the voltage drop and send its out to an arduino to trigger the Thyristors.

When the thyristors triggered, full mains voltage will be seen by the opamp input. This is not good. Therefore you have to have a limiter circuit to condition the input signal before passing it to the Opamp input. Many Opamp input don't like voltage below its negative supply voltage so there is another point to consider.

Based on your requirement, I have come up with the following circuit, along with the simulation results. If you set the opamp non inverting input voltage to 0.5V dc, then the opamp's output will provide a low to high transition when the secondary of the transformer is shorted.

 

[anishkgt]

When the thyristors triggered, full mains voltage will be seen by the opamp input.

That voltage would not reach the opamp as the opto-coupler will not conduct during this period.

The conduction of the optocoupler is controlled by the micro controller hence there will be completed isolation when the thyristors are triggered and will only conduct when the thyristors are not conducting.

[anishkgt]

You need to rectify and filter the voltage you are trying to compare, so you compare the "mean" not the instantaneous voltage.
The speed (step response) of the filter must be matched to the frequency of the line.

We just need to monitor the voltage variation, is a rectifier necessary ?

Also always use a limiting voltage device at the input pins after the high impedance resistors (like a low current zener). Although you seem to expect that always less that 1.5V will be present I doubt that. There could be switching spikes at turn-off or turn-off of the thyristors where you'll get the full line voltage at the primary.

I was thinking of a 1M Ohm resistor in series with the input of the opamp. Would the zener also be in series? When you say zener is it those dc type one or protection diodes ?


Www.Georgehobby.wordpress.com

Equipments: DSO104z, Hakko FX888D, Brymen BM869s

[anishkgt]

here is a simpler approach to get a rectified dc from a ac signal.

[anishkgt]

So normally with a DMM we probe the HOT and NEUTRAL to check the voltage. Here we just prone one terminal on the AC power which is connected to the inverting pin of the opamp via 1M resistor, so my question is, shouldn't there be another for the NEUTRAL side also to be connected from the load ? or since the Opamp is connected to GND it would use it and the neutral for the AC is not required. Would really appreciate if somebody could advise.

[anishkgt]

I've added an opto isoloator and used a precision rectifier. But i don't see any change on the output when the AC voltage is changed.
Would a comparator be a better choice here ?

Thanks in advance.