💬 Homini AC Powered Relay (2) Module

[openhardware.io]

https://www.openhardware.io/view/345/Homini-AC-Powered-Relay-2-Module

[tonnerre33]

Hello,
very nice work !
I don't know if you read another post writed by @hugch, but you have all good with this project.

Using a 5 V Arduino and the NRF without a level switcher with 3.3 V it will be break in the future.

The NRF on your board is powered by 3.3v same on the pins datas

The big relay routes have no protection (I don't know if solder resist is the correct word).

You put 1 fuse by relay

Maybe as pulldown resistor to the relay input could be useful.

You have added pullup resistor

The three GND and Vcc Pins are not connected to each other.

They are all connected

It is not a good idea to drive current intensive peripheral directly with an micro controller

You added MOSFET for command the relays

Just 3 things :

• Why you didn't put the thermalfuse in touch of the HILINK ? If heat increase inside the box, maybe is too late, that why i was thinking we need to place this protection nearby the power module.

• I didn't understand the current mesure. The ACS mesure the current only in the SSR1 ?

• Why you didn't use NRF24L01+ SMD component for gain some place ?

[Samuel235]

@tonnerre33 - Thank you for showing me these comments and pointing out the fact that my board is ticking those boxes. Is there anything else you would like to see developed onto this board?

Why you didn't put the thermalfuse in touch of the HILINK ? If heat increase inside the box, maybe is too late, that why i was thinking we need to place this protection nearby the power module.

I have put it here because, I thought with the fuse being so sensitive, it would be surfice to pick up any heat build-up from where it is. Do we have any proof that it is a much better idea to have it against the HLK module? I know people do it for that reason, but I couldn't find any solid test reasoning for this. But if this is a proven idea, I will mod my design to reflect this.

I didn't understand the current mesure. The ACS mesure the current only in the SSR1 ?

No, that current sensor will pick up the current that flows to both relays. The two sets of pins on the sensing IC are internally joined, so the current flows through. So its essentially placed between live input and the two relays. I understand your confusion though, I must have double checked that section 10 times! XD

Why you didn't use NRF24L01+ SMD component for gain some place ?

I didn't see any need for this on this application. It wouldn't have saved me any place for other components, in fact it would have wasted some room on the board here unless i had it sticking out off the side of the board, which i didn't want at all. My design rules for this was to keep the whole board within the 50mmx50mm that you get as 'prototyping boards' from PCB houses. To be honest, i wanted it much smaller, but that was not happening here for me. I'm toying with the idea of ditching the HLK and converting the 240v to DC with my own circuitry, but that is not yet. I'm an electronic engineering
student so it may come at some point tbh. After having some thinking time, i'm not actually sure how useful it would be having this behind a light switch instead of behind a light in the UK. At the moment, as we live in a house that i don't want to be ripping out the lighting circuits from the existing routes, my light switches have to be installed with terminated cables behind in case we sold the house, i could then just remove all the mysensors and revert the old wiring back.

I also remember reading somewhere that we can't use FTDI or SPI (One of them) while a radio is attached as it uses the same pins. I think its the FTDI function. Not sure if this is 100% true but I just threw the radio on a socket in case. Originally, I hard soldered the radio in place, but then realised this so i went for the socket route.

Now that i'm definitely not designing this to go inside of the light switch, it will allow me to create a module like this but with more relays on. Not too sure of the use case for this though. Its not something that i will be spending my time on for now. I may make a smaller version for 1 relay, and i do want to look at possibly doing a dimming module with triacs rather than relays.

[tonnerre33]

@Samuel235

Do we have any proof that it is a much better idea to have it against the HLK module?

I haven't any proof, it's just that in my job, we place a PTC on the winding of the transformer for mesuring the temperature.
But we do that only in the big transformers (630KVA - 1000 KVA). For the littles transformers that we use, i think it is not necessary if we protect the primary of the transformer. In french industry, we protect also the secondary of the transformer for protecting him against overload (the primary fuse protecting only against short circuit).

Then, if we don't place the thermalfuse for the transformer i don't know why to put him in the circuit. For example @scalz didn't place this protection in his projects, because he thinks the same thing. If we are protecting with fuse, why protecting with termalfuse ? I think it's a big debate

No, that current sensor will pick up the current that flows to both relays.

But if i look the schema, the current in SSR2 will never pass by the ACS712 and for the Hall effect sensor a current need to go across a magnetic field, isn't it ?

[Samuel235]

@tonnerre33 - Aww damn, you're right there about my layout. I've edited that into my new revision. Seriously, thank you! I can't believe i spent so long looking and double checking that area and still made the mistake. I suppose its something we all do at some point

As for the thermal fuse. So what are you saying, from reading what you have put in the last message, i get the feeling you wouldn't have bothered with it at all considering its a small transformer compared to some of your larger ones that you're dealing with in industry.

Do you know of anyone using the clamp style current sensing? I'm tempted to make a node that attaches to the main electric line into the house and then do away with all the current sensing I design into the nodes. Would make certain things much easier that way. Then i could do away with this issue here too

[tonnerre33]

@Samuel235 said in Homini AC Powered Relay (2) Module:

I suppose its something we all do at some point

Sure ! and i' m not the last for doing mistakes

Do you know of anyone using the clamp style current sensing?

I didn't see that for now

[Samuel235]

@tonnerre33 - I have updated this in my openhardware.io page and added it to the to-do list. Thank you

I may have a look at how accurate those types of sensors are... Could make for a nice little node!

[hugch]

Hi,
I think this is the best and nice one of all the similar projects for 2 relay switching here

Just three things I don't understand too.

• Why your are using 5 V for the external switchs?
• And why your are using the ac/dc module with 5 V and not 3,3 V (HLK-PM03) directly? With this you can remove the voltage regulator.
• Which smd fuses do you use? I think there is no 230 V AC fuse with SMD 1206 package available, is it?

And @tonnerre33:
The HLK-PM03 have output short circuit protection in it: http://www.hlktech.net/product_detail.php?ProId=59

[Samuel235]

@hugch - Thanks for the interest in this product, its nice to see someone interested in something i've made.

Why your are using 5 V for the external switchs?

The 5V for the switches was simply because i needed to use 5V for the relays it made the 5V line easier to route to the switch terminals. I could just route 3V3 but it was more convenient at the time to use the 5V line.

And why your are using the ac/dc module with 5 V and not 3,3 V (HLK-PM03) directly? With this you can remove the voltage regulator.

I needed to get a stable 5V to switch the relays, in prototyping i couldn't get the relays to be stable on 3V.

Which smd fuses do you use? I think there is no 230 V AC fuse with SMD 1206 package available, is it?

You're right, there isn't any advertised anywhere to be 240V rated. I also found that there are alot of SMD fuses that do not have a voltage rating and so i took a risk knowing that i could remove this feature easily by just bridging the gap with solder. Even though i have ran this module without the fuse blowing at 240v, i'm pretty sure this is not a stable method of running this module. In Rev1.2 i have removed this feature. I'm looking at possibly putting in glass fuses on holders that you can remove via a panel in the 3D printed enclosure. Not too confident about it right now though. The routing on the board would have to be messed with, rather a lot.

So please look at this board without the fuses on the relay lines.

In Rev1 this was implemented to protect the relay lines because i couldn't get the two fuses on board to protect the relays at the time. However, in Rev1.2 we're re-routing the AC Lines and should be able to get the relays behind another 3A or 10A fuse. If not, they remain unprotected. - PLEASE BE AWARE OF THIS. I've not added this warning on the OpenHardware page yet because as it stands, they're behind a SMD fuse (whether it be working or not, no risk if the fuse doesn't work anyway).

[hugch]

@Samuel235 said in Homini AC Powered Relay (2) Module:

@hugch - Thanks for the interest in this product, its nice to see someone interested in something i've made.

Why your are using 5 V for the external switchs?

The 5V for the switches was simply because i needed to use 5V for the relays it made the 5V line easier to route to the switch terminals. I could just route 3V3 but it was more convenient at the time to use the 5V line.

But if you use 5 V for the switch, you have to use a voltage divider or so to protect the atmega328p input, because the controller works with 3,3 V.

And why your are using the ac/dc module with 5 V and not 3,3 V (HLK-PM03) directly? With this you can remove the voltage regulator.

I needed to get a stable 5V to switch the relays, in prototyping i couldn't get the relays to be stable on 3V.

Okay, sorry I thought there are such relays with 3V available.
Than forget it and use the 5 V one

Which smd fuses do you use? I think there is no 230 V AC fuse with SMD 1206 package available, is it?

You're right, there isn't any advertised anywhere to be 240V rated. I also found that there are alot of SMD fuses that do not have a voltage rating and so i took a risk knowing that i could remove this feature easily by just bridging the gap with solder. Even though i have ran this module without the fuse blowing at 240v, i'm pretty sure this is not a stable method of running this module. In Rev1.2 i have removed this feature. I'm looking at possibly putting in glass fuses on holders that you can remove via a panel in the 3D printed enclosure. Not too confident about it right now though. The routing on the board would have to be messed with, rather a lot.

Okay. I will wait for the next revision. But I hope it will be possible to add any type of fuse for 230 V for the relay. Because protection is important for this type of relais. In Germany the most common protection for private houses is 16 A and the relay could not switch 16 A.

I am very excited about the next version. You did a great job.

[Samuel235]

@hugch

But if you use 5 V for the switch, you have to use a voltage divider or so to protect the atmega328p input, because the controller works with 3,3 V.

That is best practice however, because we are detecting high or low and not the actual voltage then we are okay here. From what i have done in the past and also read again recently. I can't remember where i found that read otherwise i would link.

Okay. I will wait for the next revision. But I hope it will be possible to add any type of fuse for 230 V for the relay. Because protection is important for this type of relais. In Germany the most common protection for private houses is 16 A and the relay could not switch 16 A.

I too really do want protection on the relay. I'm going to be doing a little in-depth reading for this condition as you have to use fuses that allow the relay to work in its operating allowances in terms of what the manufacturer deems acceptable for that relay. They have specific fuse requirements that allow it to remain working, without meeting these you can trigger false blowing of the fuse. It isn't quite as simple as just putting a fuse there, i wished it was. I also need to research how a relay fails, does it fail open or shut and therefore flowing electric or not. I think it welds itself shut and allows electricity to flow. Not good as the board will not be rated for such current up to that 16A that is back in the electric cupboard.

You're more than welcome to through articles and documents my way and i will read through them and get the next revision sorted soon as possible for all of us. Its because of these issues in question that i'm only using this in testing purposes still, not inside the ceiling or walls.

Thank you once again

[hugch]

@Samuel235

But if you use 5 V for the switch, you have to use a voltage divider or so to protect the atmega328p input, because the controller works with 3,3 V.

That is best practice however, because we are detecting high or low and not the actual voltage then we are okay here. From > what i have done in the past and also read again recently. I can't remember where i found that read otherwise i would link.

Sorry that is not correct. Pleas take a look at the atmega328p datasheet on page 365:

The maxim voltage on any pin (except RESET) by 3.3 V as Vcc is 3.8 V and not 5 V.
You can use an easy voltage divider like:

And for the fuses:
I know it is not easy. I worked for a view month by a fuse manufacturer as student. The problem here is the missing information from the relay manufacturer right?

[Samuel235]

@hugch

You, and the datasheet, are correct. That is my mistake and will add to the update list for next revision. I will either do as you suggested with a voltage divider or if possible, I will just route 3V3 to the switches. Thank you for pointing this out. I'm even more interested to find that article out that i read to question that now.

Lesson here is to always read the datasheet carefully. I think its working fine to be honest but because it is beyond the datasheet it is not supported to work.

Fuses: I'm not sure if they have documented it yet, the relays are OMRON G3MB-202P. If you find anything regarding fuses please post in here. I haven't had much of a luck recently, I did when i designed the board spend a little time looking for this but with no success. It maybe information that is easily available now

[hugch]

@Samuel235 said in Homini AC Powered Relay (2) Module:

OMRON G3MB-202P

Okay I'm trying to help you with the fuse component.
I found a datasheet for the OMRON G3MB-202P. And there are enoght information to be known for fuse selection.
I try to calculate it here (and i will try it with my bad English ).

The most important information is the melting integral.
The Melting integral has A²s as unit. So this means the maximum current for a time can exists without damaging the device. For further information look at wikipedia.

So we need any further information about:

• the protection which is present before (the typical circuit protection in private houses)
• the melting integral from the device which we want to protect
• the maximum voltage
• the maximum switching current
• the breaking capacity
• Inrush current for the switched devices (we can't know)

Typical values for an automatic circuit breaker in private houses are:

• from 25 to 100 A²s
• 230 V AC
• 16 A

So this protection isn't good enough for our relay. The relay have these values:

• 230 V AC
• 2 A maximum switching current
• I²t value (melting integral): 4 A²s
• the allowed inrush current over a small time is higher than the melting integral. It is a bit confusing i think, but if we calculate the protection for the given melting integral and it's fine. So we can define the parameters for the client (switching site of this application here)

The fuse have to be:

• I²s value higher or equal than the I²s value from the existing protection
• rating voltage over 230 V AC
• rating current 2 A or lower (because 2 A is the maximum at 25 °C, for 40 °C it is about 1.6 A)
• a maximum switching time of 1 second at 2 A or lower time with higher current but below 4A²s! To calculate use the switching time from the fuse datasheet an multiply it two times with the given current for this switching time.

And i think a fast blow fuse would be the best. There are SMD fuses with 10.1 x 3 mm and 250 V AC available.
If there is no fuse available with the values above, we could combine multiple fuses. A possible solution could be one bigger fuse for both relays and the ac/dc component. And a smaller fuse for the switching site of the relay and the HLK-PM01. But for this it is important that the circuit have only one input for the hot one (L) of 230 V and one output for each relay. In this case we could reduce the big connectors from 6 to 4. Like (L, N, Relay1, Relay2).
And the temperature fuse can work for all components too

So the protection for the primary site of the HLK-PM01 is a problem i think. Because the are no information available about the I²s value from HLK-PM01.

I hope you can understand my English and could follow my explanation?