I am trying to integrate an auto switch-OFF when the FAULT pin on U1 is pulled low other switch-ON. Would the schematic be the best approach and considering this is there a need to add another resistor in series to the FAULT pin. Here R6 acts as a Base Resistor for T1 and that would also lower the current to the FAULT pin on U1.

Datasheets of components if needed:
FODM121B
BT137-600E
KST56MTF
ACS722LLCTR
ACS711KEXLT



Appreciate any reply

 

I'm confused. You show ACS722 but that doesn't have a /FAULT output. The ACS711 does, but isnt certified for >100v and shoudn't be used for live AC control.

You should use the ACS722 and compare its output with a reference voltage to create a FAULT signal. If this remains present for a sufficient period of time then switch off.

If you did use the ACS711 be aware that it has no time delay so /FAULT can be triggered and latched by transient conditions, so it is potentially unsuitable for noisy supplies or applications involving motors for example - it is intended for audio line solutions not live AC and you cannot use it above 100v without other isolating techniques.

Your use of the optoisolator is plain wrong and dangerous. The collector voltage must be independant of your logic side - either derive from the live AC (eg series capacitor/resistor/diode plus smoothing capacitor) or isolated DC-DC converter from your 5v supply. An opto-diac, eg MOC3021 etc, would be a better/safer option. Better still, use a purpose-made SSR instead of optoisolater & triac.

 

Sorry about the confusion. I was initially thinking of the ACS711 but did not notice >100 in the datasheet. Thank you for pointing it out. The load is going be a motor. How did you notice the the FAULT pin would fluctuate due to noise ? Is it assumed that AC signals are noisy by nature ?

The ACS722LLCTR and ACS710KLATR seems a good replacement even more the later is cheaper and has fault pin as well. Would that have the same problem like the previous? The former has a BW_SEL pin instead of the FAULT.

 

I am trying to integrate an auto switch-OFF when the FAULT pin on U1 is pulled low other switch-ON. Would the schematic be the best approach and considering this is there a need to add another resistor in series to the FAULT pin. Here R6 acts as a Base Resistor for T1 and that would also lower the current to the FAULT pin on U1.

Datasheets of components if needed:
FODM121B
BT137-600E
KST56MTF
ACS722LLCTR
ACS711KEXLT

View attachment 246571

Appreciate any reply

Do you need that Opto?

 

How did you notice the the FAULT pin would fluctuate due to noise ? Is it assumed that AC signals are noisy by nature ?

I didn't say that. If you look at the internal circuit for the ACS711 and read the info you will see that /FAULT is latched low when the current = 100% rated value. The only way to unlatch it is to remove power from the chip. So any AC transient that hits 100% will generate a fault and render the system immediately inoperative. Thats virtually gauranteed with any form of inductive load or an AC supply that has inductive loads elsewhere generating voltage spikes. Hence that part is useless for your requirements on several fronts.

The ACS722LLCTR and ACS710KLATR seems a good replacement even more the later is cheaper and has fault pin as well. Would that have the same problem like the previous? The former has a BW_SEL pin instead of the FAULT.

The ACS710 has a FAULT ENABLE pin that can turn the /FAULT pin on and off, AND has the ability with a capacitor to hold off fault indication for a duration so avoiding some level of transient triggering. However, without any capacitor it still triggers within 1uS or so. For me that's just too fast for an AC circuit - it should stand a 100% current draw for milliseconds or more without tripping. Fortunately it also comes in a version with no latching circuit, so you can do your own decision making externally, though you have no control over what level is a fault current. Looking at the spec for the 10A part for example, the maximum current measured is +/-24A = +/-2.04v (from Vcc/2) but the 'fault' current is spec'd as between 0.25x and 0.4x Vcc, ie 1.25 to 2v, or +/- 14A to +/-24A which is a huge range.

For me the FAULT circuitry is unnecessary. The Vout pin gives a voltage that represents the current and its simply a matter of detecting when that is above a certain level for longer than some defined time interval that represents a fault condition, ie you define it rather than being slave to a solution that's not suitable for your situation. The way you reference it & handle the timing can be done with discrete analog parts though my prefered approach would be to use a small microcontroller to simply measure the voltage against time and generate the fault output, as well as drive an LED to indicate the fault (or maybe an OLED display to show voltage & current - its a simple add-on).

 

Do you need that Opto?

No he doesn't - at least not the way he's shown it. Though the ACS710 is isolated to 1000v+ he wll need isolation to drive a triac directly if any part of the low-voltage circuit has the potential to interact with a user. In theory, if its all inside an insulated box or a grounded metal case the triac could be fed directly but using a MOC3021 or similar is a safer option.