Relay Module KY-019 5V

[tlfong01]

Now I am thinking of using HC14 to do the level conversion.

First test by hand looks very good.

• 0V ~ 0.8V converts to ~ 5.0V (Vcc = 5.06V)

• 3V ~3.3V converts to ~ 0.05V

[tlfong01]

Now I am thinking of using HC14 to do the level conversion.

First test by hand looks very good.

• 0V ~ 0.8V converts to ~ 5.0V (Vcc = 5.06V)

• 3V ~3.3V converts to ~ 0.05V

I input DS3231 32KHz signal to HC14 2A input, and found 2Y output OK. Next step is use Rpi GPIO pin17 to input to HC14 pin 1A, and use pint 1Y output to control relay.

[tlfong01]

I input DS3231 32KHz signal to HC14 2A input, and found 2Y output OK. Next step is use Rpi GPIO pin17 to input to HC14 pin 1A, and use pint 1Y output to control relay.

32KHz is too quick for the relay. So now I used DS3231's programmed 1Hz signal for the HC14 to control the relay. Everything goes well.

[tlfong01]

I input DS3231 32KHz signal to HC14 2A input, and found 2Y output OK. Next step is use Rpi GPIO pin17 to input to HC14 pin 1A, and use pint 1Y output to control relay.

32KHz is too quick for the relay. So now I used DS3231's programmed 1Hz signal for the HC14 to control the relay. Everything goes well.

1Hz is too slow to display on the scope nicely. So I programmed DS3231 to 1kHz, then use CD4040 to count down to 8Hz, 125mS period. I can now remove Rpi, and used DS3231 + CD4040 + HC14 to simulate converted Rpi GPIO signal for testing relays.

The 8Hz signal has only 1Vpp noise, and therefore much better than that of Rpi GPIO signal.

[tlfong01]

The 8Hz signal has only 1Vpp noise, and therefore much better than that of Rpi GPIO signal.

Now I came back to test Rpi. I used GPIO pin 17 to toggle relay at the same frequency of about 8Hz, and used the same HC14 1A/1Y to convert the GPIO signal and apply it to the relay IN. The relay also toggle happily. However, the scope displays a very noisy signal.

I think this noisy signal is a potential big problem.

I am thinking of forgetting Rpi GPIO all together, and use signals form I2C MCP23017 IO expander, plus DS3231 to control the relays.

[tlfong01]

I think this noisy signal is a potential big problem.

Now I am tidying up the messy wiring, hoping that noise might reduce a bit.
The DS3231 + CD4040 + HC14 is now a standalone, Rpi programmable, relay signal generator (3.3V 8Hz and more) and 6 channel logical level converter.

[tlfong01]

Now I am tidying up the messy wiring, hoping that noise might reduce a bit.

I am a bit surprised to find that the Rpi signal is not so dirty now (Yellow waveform is DS3231, blue is Rpi).

[tlfong01]

I am a bit surprised to find that the Rpi signal is not so dirty now, ...

Now that I have a reliable standalone relay signal generator, I will take a break on this relay part and come back to study the 200VAC mains control.

Earlier I thought the MCB in my cheap tiny old flat could detect earth leakage. On second thought it shouldn't be so, because earth leakage protection is only for expensive, upper class souls, not for "low end population" like me.

Anyway, I thought the time has come for me to confirm what type of MCB do I have. Just now I checked out and was surprised to find that my MCB is an Earth Leakage Breaker. So my life is not that cheap after all!

I summarized the spec below.

Fuji Electric Earth Leakage Breaker Type EG52A F53567774
Rating 200V 50Hz 40A
Rated tripping current = 30mA
Rated non tripping current = 15mA
Rated interrupting current = 2.5kA
Rated tripping time = below 0.1S

One thing I don't understand is the following instruction.

• Push the test button before driving appliance

Does it mean to push the test button (to make sure it works) before using a new appliance? Or after a black out?

Another thing I need to check out is the following.

• What is the life of the MCB (10, 20, 30 years, or longer?)

• Any MTBF (Mean Time Between Failure)?

[davidcoton]

This is off topic but important.

One thing I don't understand is the following instruction.

Push the test button before driving appliance?

Does it mean to push the test button (to make sure it works) before using a new appliance? Or after a black out?

Another thing I need to check out is the following.

What is the life of the MCB (10, 20, 30 years, or longer?)

Any MTBF (Mean Time Between Failure)?

The usual instructions these days in the UK are that you should test the RCD (Residual Current Device, AKA Earth Leakage Breaker) every 3 months. Stop all computers that run on the supply, make sure you have light independent of the mains, and push the test button. If it doesn't trip it needs replacing. If it does trip (which will switch everything off), reset and reboot your computers.
The RCD is not usually called an MCB (Miniture Circuit Breaker). That is the other (one per circuit) trips, which are resettable replacementas for conventional fuses.
MCBs and RCDs don't usually have a quoted lifetime, which is why the RCD test is important. MCBs usually fail safe, but in any case there is no simple and safe test.
These days my professional advice in the UK is to fit RCBOs (combined RCD and MCB) to each circuit. This avoids an earth fault on one circuit switching off other circuits unnecessarily.
The rating of your RCD (100mS trip time) does not meet current UK regulations, but that is a bit technical since the trip time depends on the leakage current, and the trip time on your RCD does not have a current specified. (UK requires 30mS at 5 times rated leakage current).
From your photo, it is time to get your installation professionally inspected (in the UK, this is recommended every ten years for owner-occupied houses, five years for rental property). I would not be surprised if replacement was recommended, but I am not familiar with your local regulations.

Hope that helps, I can answer further questions but it probably needs an "Off Topic" thread and the Mods' indulgence.

[tlfong01]

This is off topic but important.

1. > ... you should test the RCD every 3 months.
I see. In the past 10 years or so, I only pressed the little red button 3 or 4 times, usually after all lights went out accidentally or similar bad events. I never knew what exactly the button meant. But I do know how to flip the big black switch back after a bad event.

2. > ... Stop all computers that run on the supply, make sure you have light independent of the mains, and push the test button.
Yes, so I never experiment with 200VAC things during the night. I always wait till the morning after. And I always have a torch under my pillow.

3. > ... The RCD is not usually called an MCB
I see. I always thought MCB meant "Main" Circuit Breaker. Now I know my swtich box is actually a RCBO, with a RCD and MCB (5P1N 6P).

4, > ... why the RCD test is important. MCBs usually fail safe, but in any case there is no simple and safe test.
I am surprised to hear that there is no simple and safe test. But of course human is still stupid and even with the so called "AI", does not either know how to safely "test" and avoid earth quake, volcano eruptions etc.

5. > ... my advice ... to fit RCBOs each circuit. This avoids an earth fault on one circuit switching off other circuits unnecessarily.
This is interesting. Yesterday I spent about one hour reading Wikipedia's very long, but very useful article on RCD. It is only after the long reading then I appreciate you suggestion above. I should read more and ask you again about this later.

6. > ... your RCD (100mS trip time) does not meet current UK regulations, but that is a bit technical.
Yes, I need to google and wiki more to get enough knowledge to understand the problem. The sad thing is that the more I google, the more I know I don't know, and the more I know I don't know that I don't know. Life is indeed too short for an electronics hobbyist.

7 > ... is time to get your installation professionally inspected. I would not be surprised if replacement was recommended, but I am not familiar with your local regulations.
Last year, my landlord told me to find an aluminium window guy to reinstall all the rusty and worn out windows hinges (to make sure no loosen window drops 24 floors and hit some unlucky dog passing by below), as required by the regulation. Perhaps next year she would ask me to replace the RCBO, IF regulation tells her so.

8 > ... Hope that helps
Many thanks for your help lengthening my life expectancy.

Yesterday I went back to the drawing board and made the very first draft of my safe relay box, as attached.

[tlfong01]

Yesterday I went back to the drawing board and made the very first draft of my safe relay box, as attached.

The first version of the schematic is a bit messy. So I did some refactoring, as list below.

[tlfong01]

The first version of the schematic is a bit messy. So I did some refactoring, as list below.

Anyway, I started the soldering work of my relay box. It was only when I found it awkward to reach the inner places squeezed with components, then I realized that the box's front and back panel can be easily removed and let your screwdriver and soldering iron comfortably reach the innermost places.

Also when I was screwing the wires to the mains terminal block, then I appreciated the use of anodization of the box surface. The outside and inside anodized surfaces are insulated. So even if the terminal wires were loosen and fallen, touching the chassis , there is no risk of any electric shock.

No wonder nobody answered the forum question of how to earth an anodized chassis.

The answer is that anodized aluminum is like plastic. It is stupid to do any earthing to the box!

Errata 2018jun28hkt2122
Actually it is stupid NOT to earth the seemingly plastic like anodised aluminium box. See the following two replies.

[davidcoton]

The answer is that anodized aluminum is like plastic. It is stupid to do any earthing to the box!

Anodizing can get scratched. You may have unanodized metal bolts or other components penetrating the anodisation.
It is important that a metal case is earthed, even if anodized.
Without knowing your particular case, I would use self-tapping screws or nuts and bolts to make earth studs.
If the panels and frame are not in good electrical contact, you will need to earth each part.

[tlfong01]

The answer is that anodized aluminum is like plastic. It is stupid to do any earthing to the box!

Anodizing can get scratched. You may have unanodized metal bolts or other components penetrating the anodisation.
It is important that a metal case is earthed, even if anodized.
Without knowing your particular case, I would use self-tapping screws or nuts and bolts to make earth studs.
If the panels and frame are not in good electrical contact, you will need to earth each part.

Many thanks for your advice and detailed instruction. So it is stupid NOT to earth anodized chassis!.

Now I should first earth both the front and back panel, before wiring the rest of the circuit.

[tlfong01]

Anodizing can get scratched. You may have unanodized metal bolts or other components penetrating the anodisation.
Without knowing your particular case, I would use self-tapping screws or nuts and bolts to make earth studs.
If the panels and frame are not in good electrical contact, you will need to earth each part.

I guess I have watched too many TV ads on scratch proof things like watches, glasses etc and thought anodization should be a very advanced technology and so make the surface as hard as diamond. I should have mixed up with the graphite thing.

New Graphene Material Turns Harder Than Diamond When It's Hit By A Bullet
http://www.iflscience.com/technology/th ... -a-bullet/

To check out how my chassis is scratch proof, I used a paper cutter to scratch the surface, I surprisingly found that it is as soft as plastic, and after a couple of cuts, the surface becomes conducting!

Anyway, I am earthing the back panel. It is very easy to use an electric drill machine and a grinding bit to grind away the anodized surface and make it conducting. I am using 4mm stainless screw and nut, plus a nickel plated self locking nut at the top.

I also googled again, harder this time, and found more discussions on how to earth anodized things. Now I know stainless steel and aluminium contact does corrode, and some sort of corrosion inhibitor should be used under the steel washer, but of course that is too over kill for a hobbyist's Micky mouse project.

[tlfong01]

Anyway, I am earthing the back panel.

Now I have also earthed the main chassis body. The green wire stripped off from a mains cable is a bit too thick and stiff to handle. Next time I might use 22AWG wire instead.

[davidcoton]

Next time I might use 22AWG wire instead.

The wire used for earthing must be at least as thick as the line and neutral supply wires.

Nice wiring, btw.

[tlfong01]

Next time I might use 22AWG wire instead.

The wire used for earthing must be at least as thick as the line and neutral supply wires.
Nice wiring, btw.

> 1. The wire used for earthing must be at least as thick as the line and neutral supply wires.

I see. So the life saving rope should be stronger, or at least as strong, and never be the weakest link of my life, however cheap it is.

> 2. Nice wiring, btw.

Thank you. I do have a nice metal work bench.

[tlfong01]

The wire used for earthing must be at least as thick as the line and neutral supply wires.

Now I have earthed main body, front and back panel. The earth wiring squeezed inside the small box is tedious and boring [next time I will use a bigger box.], so I don't feel like doing the remaining top cover. I will now go back to do the more interesting relay related wiring.

[tlfong01]

I will now go back to do the more interesting relay related wiring.

I still have not wired the relays and the mains and transformer output part. But the wiring is getting quite messy, and I forgot which wire goes to where. I have no confidence that no smoke comes out from the Rpi when I flip the 220VAC mains switch.

So I think I would go very slowly, and eat the elephant by 3 bites:

• 1. Test the circuit without the relay board, using 3 manual switches (green block in the schematic attached below).

• 2. Test the circuit with the relay board without the Rpi, but with DS3231 Real Time Clock's 5V square wave signal (8 Hz/cycles per second, or 0.125 second per cycle, to 4 seconds per cycle (2 seconds on, 2 seconds off).

• 3. Test the the circuit with Rpi, first loading 12VAC, then 24VAC, and finally 220VAC.

[tlfong01]

So I think I would go very slowly, and eat the elephant by 3 bites:

• 1. Test the circuit without the relay board, using 3 manual switches.

[tlfong01]

Now I am earthing the top panel, connecting the earth of Mains in and Mains out together. I found the following bugs:

• 1. One earth wire was dangling. I extended three wires L, N, E for the two 12V/24V transformers. I forgot that the transformers need no earth lines. So I removed the redundant earth wire.

• 2. I found two dry joints, and one Live wire has chipping surface, though not exposing its brass strings. I used insulated sleeve to fix the problem.

As There’s Never Just One Cockroach in the Kitchen, I begin to worry that I could not sleep well to night, ...

[Mortimer]

When we construct metal encased systems we are required to create a single main earth stud, typically a larger diameter than all the others. To this is tied the incoming mains earth, or in the case of industrial or marine systems a separate earth strap to an earth bar or the hull. All other panels of the chassis are then individually connected to this main stud. One "No! No!" is to daisy chain the individual metal parts. I.e. the smaller earth stud will have a single cable coming from it, going to the main earth stud.

The no daisy chain rule is because this means individual panels can be completely removed without affecting the integrity of the earthing to the rest of the chassis.

We are also not permitted to use he chassis as a conductive path for earthing. All earthing current must be able to flow through the earth wires.

[tlfong01]

... required to create a single main earth stud, typically a larger diameter than ...
... tied the incoming mains earth, or ... a separate earth strap to an earth bar ...
... All other panels of the chassis individually connected to this main stud.
... One "No! No!" is to daisy chain the individual metal parts.
... smaller earth stud ...single cable coming from it, going to the main earth stud.
... no daisy chain ... individual panels ... remove ... integrity earthing rest of the chassis
... not permitted to use he chassis as a conductive path for earthing.

Thanks a lot for your detailed explanation and advice on earth stud. I only heard of the term "stud" a day ago. I googled and learnt that it is a metal rod with screw thread on both ends (Of course I have seen such thing before, but I did not know its name is "stud"). I guess one stud end is screwed to a heavy metal block buried 6 feet undergound, and the other end to the earthing wires.

I was daisy chaining the mains lines and terminal blocks, because I did not have any overall wiring plan in the beginning of the relay box project. I just added things as I went along, with the result that I often forgot which wire goes where, and not sure if any lonely wires dangling somewhere.

Now I think my life is not that cheap to deserve a sloppy earth stud design. So I proudly made a miniature, horizontally placed stud, as shown below. [Errata - The two wires labelled L in, Lout, should read E in, E out.]

Now I can sleep well.

[Mortimer]

Looks good.