I found this, easy to repair, electronic design for a 10mA Residual Current Device (RCD) online.
All other RCD's, I am aware of, are mechanical @30/40mA and hard to repair.
I used EasyEDA to draw the circuit and got JLCPCB to make a few prototype boards.

I built up 2 boards with the components mentioned in the diagram, yet both boards don't work very well.
The 2 relays are normally closed and conduct mains voltage to output, when a fault develops the relays energize and cut the output.
There are no mistakes, dry solder joints nor loose connections on the boards.

The problem::
1)On both boards the 1 Watt 330R resistor gets extremely hot, It gets hot whether the relays are energized or not --> It was replaced it with a 5W wire-wound version. (which still gets quite warm)

2)The 220uF 35V capacitor was replaced with a 220uF 50V version.

3)The Trip test works sometimes only, eg switching 5/10 times, it switches on the red LED and energizes the both relays .
4)Giving the circuit board or the CT a good 'TAP', causes the circuit to switch on the relays.

The bridge diode gets 32V AC and outputs a VCC of around 30VDC.
The reset switch work fine.

During the last week, testing and trying, I have ruined 4 Op-Amps, So I built a 5V Op-Amp tester for future
reference.
I want it to work, but I just can't find the problem

Can anyone please suggest a solution to the problem(s), I can build anything, but I cannot design a circuit like this.
Many thanks..Jo

 

)Giving the circuit board or the CT a good 'TAP', causes the circuit to switch on the relays.

That sounds like a bad solder connection somewhere on the board.

 

The 'tap/shock" switching happens on both boards. The boards were built up with 7 days inbetween.
As I already mentioned, there is no 'bad' or dry/loose joint.
I think that the relays are on the verge of switching over and the "shock" does it..
The main problems are : WHY the resistor gets so hot.. & The trip test not functioning as one would expect.
There must be a fault in the design, and that is the reason I am on this forum asking for help.

 

If the op-amp is being destroyed for some reason, then voltages outside its spec are being applied.

30Vdc from the bridge rectifier seems high, as it is close to the supply limit for the op-amp, and exceeds the voltage rating of capacitor C5 (25V).

My bet would be that the non-inverting input is exceeding Vcc for some reason. Note that the diode D7 is preventing a negative voltage (with reference to GND being fed to the op-amp). If this component is faulty (open circuit) that protection will be lost.

If the non-inverting op-amp input voltage exceeds Vcc for some reason, the op-amp will likely be toast. You could consider adding protection against this by adding a zener diode of suitable breakdown voltage (less than Vcc) between the non-inverting input and GND.

 

The power supply section is probably being overloaded by the relay current, causing the voltage to drop below the pull-in voltage of the relays.
What is the Vcc voltage when it's triggered?

Capacitive dropper power supplies must be carefully matched to the design load.

 

If the op-amp is being destroyed for some reason, then voltages outside its spec are being applied.

30Vdc from the bridge rectifier seems high, as it is close to the supply limit for the op-amp, and exceeds the voltage rating of capacitor C5 (25V).

My bet would be that the non-inverting input is exceeding Vcc for some reason. Note that the diode D7 is preventing a negative voltage (with reference to GND being fed to the op-amp). If this component is faulty (open circuit) that protection will be lost.

If the non-inverting op-amp input voltage exceeds Vcc for some reason, the op-amp will likely be toast. You could consider adding protection against this by adding a zener diode of suitable breakdown voltage (less than Vcc) between the non-inverting input and GND.

VCC is 30VDC
The opamp is also getting 30VDC between pin 1 & 8, but5 it is still OK after 10 minutes of being powered up.
The relays are off when conducting power to the load, only when a fault is detected are they energized. ( thats the theory anyway).
where do I put a zener between? whais the value you suggest? (I have 24V and 3.3V zeners in stock)

 

The power supply section is probably being overloaded by the relay current, causing the voltage to drop below the pull-in voltage of the relays.
What is the Vcc voltage when it's triggered?

Capacitive dropper power supplies must be carefully matched to the design load.

't
30VDC before and after triggering (voltage at the moment is is triggered , I don't know)
Under normal running, mains power is being conducted to output, eg: The relays are NOT energized, when conducting power to the load, only when a fault is detected are they energized and shut down the output mains voltage.
Quote "Capacitive dropper power supplies must be carefully matched to the design load." I am not sure what you mean by this.

 

VCC is 30VDC
The opamp is also getting 30VDC between pin 1 & 8, but5 it is still OK after 10 minutes of being powered up.
The relays are off when conducting power to the load, only when a fault is detected are they energized. ( thats the theory anyway).
where do I put a zener between? whais the value you suggest? (I have 24V and 3.3V zeners in stock)

VCC is 30VDC
The opamp is also getting 30VDC between pin 1 & 8
Diode D7 is OK
VCC is 30VDC

The opamp is also getting 30VDC between pin 1 & 8, but it is still OK after 10 minutes of being powered up.

The relays are off when conducting power to the load, only when a fault is detected are they energized. ( thats the theory anyway).

where do I put a zener between? what is the value you suggest? (I have 24V and 3.3V zeners in stock)

Diode D7 is OK

I replace Cap C4 with a 220uF 50V version

The board still behaves the same as before. and R1 5W is still getting hot after less than 5 minutes

 

The 24V zener should be ideal, it will prevent the non-inverting input exceeding Vcc (30V) but allowing it to exceed the inverting input and the op-amp output switch.

 

't
30VDC before and after triggering (voltage at the moment is is triggered , I don't know)
Under normal running, mains power is being conducted to output, eg: The relays are NOT energized, when conducting power to the load, only when a fault is detected are they energized and shut down the output mains voltage.
Quote "Capacitive dropper power supplies must be carefully matched to the design load." I am not sure what you mean by this.

It's the voltage when it's triggered you would want to check, I'll bet it drops substantially.
30V is very close to the absolute maximum rating for this part.

 

I got it to switch after numerous times flipping the "Trip test Switch"--> the voltage on the Opamp was 26.8 while the VCC is 30.7.
I tried to replicate the switching , but it won't budge anymore , which has been a problem all along
The opamp, i just tested, is still OK

 

It's meant to trip at 10mA, so why does the test button only pass 1mA?
The current transformer is 5000:1, so for a 10mA input current it will give 2uA output, which will give 13mV across the 6.8k resistor.
I'm not surprised it doesn't work. I think you have the wrong current transformer.

10mA RCDs are available:
https://westbasedirect.com/products...YPrL52ensopcNgC8C5upLROeZRlyxIMUaAnChEALw_wcB
They are also designed so that the contacts can't weld together, and that they remain tripped until mechanically reset.
I can't see any reason why you should ever want to repair it.
They do wear out if repeatedly tripped so should be replaced from time to time.
If it were my safety that it is designed to protect, I'd use a professional tested and approved device every time.

 

yes. i know one can buy anything, BUT The challenge in making it myself. anyway...
The CT is the exact one that the origin eg https://www.aliexpress.com/item/329....order_list.order_list_main.56.38081802VaTgdj
I sent an email to the designer: No help was forthcoming--> But he did tell me that the video and all details were removed.
I have a copy of that video, I;ll attach it here .Nope, it won't accept MP4 . Instead i attach 3 pics of the original

 

I doubt you will be able to make a GFI/RCD using that current transformer. It is designed to give an output of 20mA when a single conductor passed through the core is passing 100A; whereas a current transformer used in a GFI/RCD outputs a current/voltage when it detects an imbalance in current between the two conductors passed through the core – the difference being of the order of a few mA.

It is likely that the current transformer you have will output 20mA when the imbalance between the conductors is 100A – making it totally unsuitable for your project.

 

OK got that ... At the most it will pas 10A 240V 100A is out of the question -->So scaled down it would be a 2mA trip at the most. House power-point circuit breakers are all 10A
According to the video, there needs to be only a couple of mA's difference between the mains conductors, adjustable by the pot meter...??

 

You're missing the point. If you want it to trip at 2mA, the output from the current transformer would be 400nA.
400nA is hardly more than the bias current for an LM358. It would generate a voltage across your 6.8k resistor of 3mV.
If you look at the description of the current transformer on AliExpress, it says it is for an ammeter, not a residual current device.
It doesn't surprise me that the designer is not being helpful. How much help can you give on a circuit that doesn't work?
A current transformer designed for 100A isn't going to maintain its accuracy down to 10mA (or even 2mA) because of the magnetisation current.

 

OK,,,, yes, I guessed there would be more to it .
A Big Thank you Ian0
And also a BIG thank you to all contributors
Much appreciated
Jo.

 

You could experiment with one of these as the front end of your design (the module including an op-amp). Although for a 5A rating, the pot adjustment may give considerable gain for a lower current, say with a 10mA current imbalance, measure the circuit output to determine whether it would be suitable for your project.

Although the seller’s description is not clear, I think it outputs a dc voltage corresponding to an ac current flow.

https://www.ebay.co.uk/itm/401443893349?var=671096078507

 

You could experiment with one of these as the front end of your design (the module including an op-amp). Although for a 5A rating, the pot adjustment may give considerable gain for a lower current, say with a 10mA current imbalance, measure the circuit output to determine whether it would be suitable for your project.

Although the seller’s description is not clear, I think it outputs a dc voltage corresponding to an ac current flow.

https://www.ebay.co.uk/itm/401443893349?var=671096078507

You could, but it is a 1000:1 transformer, so 10uA output for 10mA residual current
https://www.micro-transformer.com/pcb-mounting-current-transformer-ZMCT103C.html
https://5krorwxhmqqirik.leadongcdn.com/ZMCT103C+specification-aidirBqoKomRilSjpimnokp.pdf

 

Hi,
330R 1W resistor get hot ... You can easily improve that, just replace it to a 50-100 Ohm 1W thick film or wirewound resistor to minimize heating.
Opamp is damaged... It's not a surprise, just watch how the opamp's input circuit is made. Replace D7 to a 4-5V Zenner to protect opamp's input . ( It's better to rebuilt this part of circuit- to use the bridge on Shottky's, but it's for future...).
Test resistor R11 to be lovered to 30-50 kOhm.